1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
15 TL/02 The BSD's have an arc4random API. One of the functions to induce
16 adding randomness was arc4random_stir(), but it has been removed in
17 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
18 function when detected.
20 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
21 cause callback expansion.
23 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
24 syntax errors in an expansion can be treated as a string instead of
25 logging or causing an error, due to the internal use of bool_lax
26 instead of bool when processing it.
28 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
29 server certificates when making smtp deliveries.
31 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
33 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
35 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
37 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
38 Merged patch from Sebastian Wiedenroth.
40 JH/05 Fix results-pipe from transport process. Several recipients, combined
41 with certificate use, exposed issues where response data items split
42 over buffer boundaries were not parsed properly. This eventually
43 resulted in duplicates being sent. This issue only became common enough
44 to notice due to the introduction of conection certificate information,
45 the item size being so much larger. Found and fixed by Wolfgang Breyha.
47 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
48 size buffer was used, resulting in syntax errors when an expansion
51 JH/07 Add support for directories of certificates when compiled with a GnuTLS
52 version 3.3.6 or later.
54 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
55 is EXPERIMENTAL_EVENT, the main-configuration and transport options
56 both become "event_action", the variables become $event_name, $event_data
57 and $event_defer_errno. There is a new variable $verify_mode, usable in
58 routers, transports and related events. The tls:cert event is now also
59 raised for inbound connections, if the main configuration event_action
62 TL/06 In test suite, disable OCSP for old versions of openssl which contained
63 early OCSP support, but no stapling (appears to be less than 1.0.0).
65 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
66 server certificate names available under the smtp transport option
67 "tls_verify_cert_hostname" now do not permit multi-component wildcard
70 JH/10 Time-related extraction expansions from certificates now use the main
71 option "timezone" setting for output formatting, and are consistent
72 between OpenSSL and GnuTLS compilations. Bug 1541.
74 JH/11 Fix a crash in mime ACL when meeting a zero-length parameter in the
77 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
78 include certificate info, eximon was claiming there were spoolfile
81 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
83 JH/14 Log delivery-related information more consistently, using the sequence
84 "H=<name> [<ip>]" wherever possible.
86 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
87 are problematic for Debian distribution, omit them from the release
93 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
94 checkers that were complaining about end of non-void function with no
97 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
98 This was a regression intruduced in 4.83 by another bugfix.
100 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
102 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
103 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
109 TF/01 Correctly close the server side of TLS when forking for delivery.
111 When a message was received over SMTP with TLS, Exim failed to clear up
112 the incoming connection properly after forking off the child process to
113 deliver the message. In some situations the subsequent outgoing
114 delivery connection happened to have the same fd number as the incoming
115 connection previously had. Exim would try to use TLS and fail, logging
116 a "Bad file descriptor" error.
118 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
119 utilities have not been installed.
121 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
122 temporary space as the ACL may create new global variables.
124 TL/01 LDAP support uses per connection or global context settings, depending
125 upon the detected version of the libraries at build time.
127 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
128 to extract and use the src ip:port in logging and expansions as if it
129 were a direct connection from the outside internet. PPv2 support was
130 updated based on HAProxy spec change in May 2014.
132 JH/02 Add ${listextract {number}{list}{success}{fail}}.
134 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
135 Properly escape header and check for NULL return.
137 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
140 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
142 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
143 characters in header names, implemented as a verify condition.
144 Contributed by Michael Fischer v. Mollard.
146 TL/05 Rename SPF condition results err_perm and err_temp to standardized
147 results permerror and temperror. Previous values are deprecated but
148 still accepted. In a future release, err_perm and err_temp will be
149 completely removed, which will be a backward incompatibility if the
150 ACL tests for either of these two old results. Patch contributed by
151 user bes-internal on the mailing list.
153 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
155 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
156 selectors, in both main and reject logs.
158 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
161 JH/07 Add malware type "sock" for talking to simple daemon.
163 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
165 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
166 routers/transports under cutthrough routing.
168 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
169 numbers. Touch up "bool" conditional to keep the same definition.
171 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
173 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
175 JH/12 Expand items in router/transport headers_add or headers_remove lists
176 individually rather than the list as a whole. Bug 1452.
178 Required for reasonable handling of multiple headers_ options when
179 they may be empty; requires that headers_remove items with embedded
180 colons must have them doubled (or the list-separator changed).
182 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
183 view the policy declared in the DMARC record. Currently, $dmarc_status
184 is a combined value of both the record presence and the result of the
187 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
189 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
190 dnslookup router and the smtp transport (applying to the forward
193 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
194 of ldap servers used for a specific lookup. Patch provided by Heiko
197 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
198 New variable $lookup_dnssec_authenticated for observability.
200 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
201 Patch submitted by Lars Timman.
203 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
205 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
206 Requires trusted mode and valid format message id, aborts otherwise.
207 Patch contributed by Heiko Schlichting.
209 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
210 certextract with support for various fields. Bug 1358.
212 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
213 is requested by default, modifiable by smtp transport option
216 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
217 operate on certificate variables to give certificate fingerprints
218 Also new ${sha256:cert_variable}.
220 JH/23 The PRDR feature is moved from being Experimental into the mainline.
222 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
223 Christian Aistleitner.
225 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
227 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
228 file. Patch from Wolfgang Breyha.
230 JH/25 Expand the coverage of the delivery $host and $host_address to
231 client authenticators run in verify callout. Bug 1476.
233 JH/26 Port service names are now accepted for tls_on_connect_ports, to
234 align with daemon_smtp_ports. Bug 72.
236 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
237 support and error reporting did not work properly.
239 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
240 and is readable. Patch from Andrew Colin Kissa.
242 TL/14 Enhance documentation of ${run expansion and how it parses the
243 commandline after expansion, particularly in the case when an
244 unquoted variable expansion results in an empty value.
246 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
248 PP/02 Fix internal collision of T_APL on systems which support RFC3123
249 by renaming away from it. Addresses GH issue 15, reported by
252 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
254 TL/15 SECURITY: prevent double expansion in math comparison functions
255 (can expand unsanitized data). Not remotely exploitable.
262 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
264 PP/02 Make -n do something, by making it not do something.
265 When combined with -bP, the name of an option is not output.
267 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
270 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
271 $sender_host_name and config options to manage this, and basic check
274 PP/05 DSCP support for outbound connections and control modifier for inbound.
276 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
277 (Only plugin which currently uses this is kerberos4, which nobody should
278 be using, but we should make it available and other future plugins might
279 conceivably use it, even though it would break NAT; stuff *should* be
280 using channel bindings instead).
282 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
283 name; added for Sendmail compatibility; requires admin caller.
284 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
285 just ignore it); requires trusted caller.
286 Also parse but ignore: -Ac -Am -X<logfile>
289 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
291 TL/02 Add +smtp_confirmation as a default logging option.
293 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
294 Patch by Magnus Holmgren from 2007-02-20.
296 TL/04 Bugzilla 1281 - Spec typo.
297 Bugzilla 1283 - Spec typo.
298 Bugzilla 1290 - Spec grammar fixes.
300 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
302 TL/06 Add Experimental DMARC support using libopendmarc libraries.
304 TL/07 Fix an out of order global option causing a segfault. Reported to dev
305 mailing list by by Dmitry Isaikin.
307 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
309 JH/02 Support "G" suffix to numbers in ${if comparisons.
311 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
313 NM/01 Bugzilla 1197 - Spec typo
314 Bugzilla 1196 - Spec examples corrections
316 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
318 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
319 gnutls_enable_pkcs11, but renamed to more accurately indicate its
322 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
323 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
325 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
326 "acl {{name}{arg}...}", and optional args on acl condition
329 JH/05 Permit multiple router/transport headers_add/remove lines.
331 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
333 JH/07 Avoid using a waiting database for a single-message-only transport.
334 Performance patch from Paul Fisher. Bugzilla 1262.
336 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
339 JH/09 Add $headers_added variable, with content from use of ACL modifier
340 add_header (but not yet added to the message). Bugzilla 199.
342 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
343 Pulled from Bugzilla 817 by Wolfgang Breyha.
345 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
347 (nb: this is the same fix as in Exim 4.80.1)
349 JH/11 Add A= logging on delivery lines, and a client_set_id option on
352 JH/12 Add optional authenticated_sender logging to A= and a log_selector
355 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
357 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
358 advertise SMTP AUTH mechanism to us, instead of a generic
359 protocol violation error. Also, make Exim more robust to bad
360 data from the Dovecot auth socket.
362 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
364 When a queue runner is handling a message, Exim first routes the
365 recipient addresses, during which it prunes them based on the retry
366 hints database. After that it attempts to deliver the message to
367 any remaining recipients. It then updates the hints database using
370 So if a recipient address works intermittently, it can get repeatedly
371 deferred at routing time. The retry hints record remains fresh so the
372 address never reaches the final cutoff time.
374 This is a fairly common occurrence when a user is bumping up against
375 their storage quota. Exim had some logic in its local delivery code
376 to deal with this. However it did not apply to per-recipient defers
377 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
379 This change adds a proper retry rule check during routing so that the
380 final cutoff time is checked against the message's age. We only do
381 this check if there is an address retry record and there is not a
382 domain retry record; this implies that previous attempts to handle
383 the address had the retry_use_local_parts option turned on. We use
384 this as an approximation for the destination being like a local
385 delivery, as in LMTP.
387 I suspect this new check makes the old local delivery cutoff check
388 redundant, but I have not verified this so I left the code in place.
390 TF/02 Correct gecos expansion when From: is a prefix of the username.
392 Test 0254 submits a message to Exim with the header
396 When I ran the test suite under the user fanf2, Exim expanded
397 the header to contain my full name, whereas it should have added
398 a Resent-Sender: header. It erroneously treats any prefix of the
399 username as equal to the username.
401 This change corrects that bug.
403 GF/01 DCC debug and logging tidyup
404 Error conditions log to paniclog rather than rejectlog.
405 Debug lines prefixed by "DCC: " to remove any ambiguity.
407 TF/03 Avoid unnecessary rebuilds of lookup-related code.
409 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
410 Bug spotted by Jeremy Harris; was flawed since initial commit.
411 Would have resulted in OCSP responses post-SNI triggering an Exim
412 NULL dereference and crash.
414 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
416 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
417 Bug detection, analysis and fix by Samuel Thibault.
418 Bugzilla 1331, Debian bug #698092.
420 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
422 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
423 Server implementation by Todd Lyons, client by JH.
424 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
425 config variable "prdr_enable" controls whether the server
426 advertises the facility. If the client requests PRDR a new
427 acl_data_smtp_prdr ACL is called once for each recipient, after
428 the body content is received and before the acl_smtp_data ACL.
429 The client is controlled by bolth of: a hosts_try_prdr option
430 on the smtp transport, and the server advertisement.
431 Default client logging of deliveries and rejections involving
432 PRDR are flagged with the string "PRDR".
434 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
435 fclose(). Diagnosis by Todd Lyons.
437 PP/17 Update configure.default to handle IPv6 localhost better.
438 Patch by Alain Williams (plus minor tweaks).
441 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
442 This is now consistent with GnuTLS, and is now documented: the
443 previous undocumented portable approach to treating the option as
444 unset was to force an expansion failure. That still works, and
445 an empty string is now equivalent.
447 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
448 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
449 not performing validation itself.
451 PP/20 Added force_command boolean option to pipe transport.
452 Patch from Nick Koston, of cPanel Inc.
454 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
457 TF/04 Added udpsend ACL modifer and hexquote expansion operator
459 PP/21 Fix eximon continuous updating with timestamped log-files.
460 Broken in a format-string cleanup in 4.80, missed when I repaired the
461 other false fix of the same issue.
462 Report and fix from Heiko Schlichting.
465 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
466 Report from Prashanth Katuri.
468 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
469 It's SecureTransport, so affects any MacOS clients which use the
470 system-integrated TLS libraries, including email clients.
472 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
473 using a MIME ACL for non-SMTP local injection.
474 Report and assistance in diagnosis by Warren Baker.
476 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
478 JH/16 Fix comparisons for 64b. Bugzilla 1385.
480 TL/09 Add expansion variable $authenticated_fail_id to keep track of
481 last id that failed so it may be referenced in subsequent ACL's.
483 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
486 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
487 ldap library initialization, allowing self-signed CA's to be
488 used. Also properly sets require_cert option later in code by
489 using NULL (global ldap config) instead of ldap handle (per
490 session). Bug diagnosis and testing by alxgomz.
492 TL/12 Enhanced documentation in the ratelimit.pl script provided in
493 the src/util/ subdirectory.
495 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
496 renamed to Transport Post Delivery Action by Jeremy Harris, as
499 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
500 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
501 redis_servers = needs to be configured which will be used by the redis
502 lookup. Patch from Warren Baker, of The Packet Hub.
504 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
506 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
507 hostname or reverse DNS when processing a host list. Used suggestions
508 from multiple comments on this bug.
510 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
512 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
513 Missed a few lines, added it to make the runtest require no keyboard
516 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
517 contains upper case chars. Make router use caseful_local_part.
519 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
520 support when GnuTLS has been built with p11-kit.
526 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
528 This, or similar/improved, will also be change PP/11 of 4.82.
534 PP/01 Handle short writes when writing local log-files.
535 In practice, only affects FreeBSD (8 onwards).
536 Bugzilla 1053, with thanks to Dmitry Isaikin.
538 NM/01 Bugzilla 949 - Documentation tweak
540 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
543 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
545 PP/02 Implemented gsasl authenticator.
547 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
549 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
550 `pkg-config foo` for cflags/libs.
552 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
553 with rest of GSASL and with heimdal_gssapi.
555 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
556 `pkg-config foo` for cflags/libs for the TLS implementation.
558 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
559 properties get this fed in as external SSF. A number of robustness
560 and debugging improvements to the cyrus_sasl authenticator.
562 PP/08 cyrus_sasl server now expands the server_realm option.
564 PP/09 Bugzilla 1214 - Log authentication information in reject log.
565 Patch by Jeremy Harris.
567 PP/10 Added dbmjz lookup type.
569 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
571 PP/12 MAIL args handles TAB as well as SP, for better interop with
572 non-compliant senders.
573 Analysis and variant patch by Todd Lyons.
575 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
576 Bug report from Lars Müller <lars@samba.org> (via SUSE),
577 Patch from Dirk Mueller <dmueller@suse.com>
579 PP/13 tls_peerdn now print-escaped for spool files.
580 Observed some $tls_peerdn in wild which contained \n, which resulted
581 in spool file corruption.
583 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
584 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
585 or write after TLS renegotiation, which otherwise led to messages
588 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
589 as a tracking header (ie: a signed header comes before the signature).
590 Patch from Wolfgang Breyha.
592 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
593 comma-sep list; embedded commas doubled.
595 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
597 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
599 Report and patch from Dmitry Banschikov.
601 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
602 Removed SSL_clear() after SSL_new() which led to protocol negotiation
603 failures. We appear to now support TLS1.1+ with Exim.
605 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
606 lets Exim select keys and certificates based upon TLS SNI from client.
607 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
608 before an outbound SMTP session. New log_selector, +tls_sni.
610 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
611 NULL dereference. Report and patch from Alun Jones.
613 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
614 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
615 Not seeing resolver debug output on NetBSD, but suspect this is a
616 resolver implementation change.
618 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
619 Left warnings. Added "eximon gdb" invocation mode.
621 PP/21 Defaulting "accept_8bitmime" to true, not false.
623 PP/22 Added -bw for inetd wait mode support.
625 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
626 locate the relevant includes and libraries. Made this the default.
628 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
629 Bugzilla 1246, report and most of solution from Tomasz Kusy.
631 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
632 This may cause build issues on older platforms.
634 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
635 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
636 gnutls_require_mac & gnutls_require_protocols (no longer supported).
637 Added SNI support via GnuTLS too.
638 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
640 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
642 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
643 Applied second patch from Janne, implementing suggestion to default
644 multiple-strings-in-record handling to match SPF spec.
646 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
648 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
649 read-only, out of scope).
650 Patch from Wolfgang Breyha, report from Stuart Northfield.
652 PP/29 Fix three issues highlighted by clang analyser static analysis.
653 Only crash-plausible issue would require the Cambridge-specific
654 iplookup router and a misconfiguration.
655 Report from Marcin Mirosław.
657 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
659 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
660 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
661 As part of this, removing so much warning spew let me fix some minor
662 real issues in debug logging.
664 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
665 assignment on my part. Fixed.
667 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
668 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
669 Janne Snabb (who went above and beyond: thank you).
671 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
672 string otherwise requires a connection and a bunch more work and it's
673 relatively easy to get wrong. Should also expose TLS library linkage
676 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
677 64-bit ${eval} (JH/03).
679 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
680 GNU libc to support some of the 64-bit stuff, should not lead to
681 conflicts. Defined before os.h is pulled in, so if a given platform
682 needs to override this, it can.
684 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
685 protection layer was required, which is not implemented.
686 Bugzilla 1254, patch from Wolfgang Breyha.
688 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
689 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
690 tls_dhparam take prime identifiers. Also unbreak combination of
691 OpenSSL+DH_params+TLSSNI.
693 PP/39 Disable SSLv2 by default in OpenSSL support.
699 PP/01 Solaris build fix for Oracle's LDAP libraries.
700 Bugzilla 1109, patch from Stephen Usher.
702 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
704 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
707 TF/02 Fix a couple more cases where we did not log the error message
708 when unlink() failed. See also change 4.74-TF/03.
710 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
711 lock up or crash if it happened to be inside a call to libc when it
712 got a SIGUSR1 from exiwhat.
714 The SIGUSR1 handler appends the current process status to the process
715 log which is later printed by exiwhat. It used to use the general
716 purpose logging code to do this, but several functions it calls are
717 not safe for signals.
719 The new output code in the SIGUSR1 handler is specific to the process
720 log, and simple enough that it's easy to inspect for signal safety.
721 Removing some special cases also simplifies the general logging code.
722 Removing the spurious timestamps from the process log simplifies
725 TF/04 Improved ratelimit ACL condition.
727 The /noupdate option has been deprecated in favour of /readonly which
728 has clearer semantics. The /leaky, /strict, and /readonly update modes
729 are mutually exclusive. The update mode is no longer included in the
730 database key; it just determines when the database is updated. (This
731 means that when you upgrde Exim will forget old rate measurements.)
733 Exim now checks that the per_* options are used with an update mode that
734 makes sense for the current ACL. For example, when Exim is processing a
735 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
736 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
737 must specify per_mail/readonly. If you omit the update mode it defaults to
738 /leaky where that makes sense (as before) or /readonly where required.
740 The /noupdate option is now undocumented but still supported for
741 backwards compatibility. It is equivalent to /readonly except that in
742 ACLs where /readonly is required you may specify /leaky/noupdate or
743 /strict/noupdate which are treated the same as /readonly.
745 A useful new feature is the /count= option. This is a generalization
746 of the per_byte option, so that you can measure the throughput of other
747 aggregate values. For example, the per_byte option is now equivalent
748 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
750 The per_rcpt option has been generalized using the /count= mechanism
751 (though it's more complicated than the per_byte equivalence). When it is
752 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
753 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
754 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
755 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
756 Note that using per_rcpt with a non-readonly update mode in more than
757 one ACL will cause the recipients to be double-counted. (The per_mail
758 and per_byte options don't have this problem.)
760 The handling of very low rates has changed slightly. If the computed rate
761 is less than the event's count (usually one) then this event is the first
762 after a long gap. In this case the rate is set to the same as this event's
763 count, so that the first message of a spam run is counted properly.
765 The major new feature is a mechanism for counting the rate of unique
766 events. The new per_addr option counts the number of different
767 recipients that someone has sent messages to in the last time period. It
768 behaves like per_rcpt if all the recipient addresses are different, but
769 duplicate recipient addresses do not increase the measured rate. Like
770 the /count= option this is a general mechanism, so the per_addr option
771 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
772 example, measure the rate that a client uses different sender addresses
773 with the options per_mail/unique=$sender_address. There are further
774 details in the main documentation.
776 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
778 TF/06 Removed a few PCRE remnants.
780 TF/07 Automatically extract Exim's version number from tags in the git
781 repository when doing development or release builds.
783 PP/02 Raise smtp_cmd_buffer_size to 16kB.
784 Bugzilla 879. Patch from Paul Fisher.
786 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
787 Heavily based on revision 40f9a89a from Simon Arlott's tree.
790 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
792 PP/05 Variable $av_failed, true if the AV scanner deferred.
793 Bugzilla 1078. Patch from John Horne.
795 PP/06 Stop make process more reliably on build failure.
796 Bugzilla 1087. Patch from Heiko Schlittermann.
798 PP/07 Make maildir_use_size_file an _expandable_ boolean.
799 Bugzilla 1089. Patch from Heiko Schlittermann.
801 PP/08 Handle ${run} returning more data than OS pipe buffer size.
802 Bugzilla 1131. Patch from Holger Weiß.
804 PP/09 Handle IPv6 addresses with SPF.
805 Bugzilla 860. Patch from Wolfgang Breyha.
807 PP/10 GnuTLS: support TLS 1.2 & 1.1.
809 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
812 PP/11 match_* no longer expand right-hand-side by default.
813 New compile-time build option, EXPAND_LISTMATCH_RHS.
814 New expansion conditions, "inlist", "inlisti".
816 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
818 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
820 PP/14 fix log_write() format string regression from TF/03.
821 Bugzilla 1152. Patch from Dmitry Isaikin.
827 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
829 PP/02 Harmonised TLS library version reporting; only show if debugging.
830 Layout now matches that introduced for other libraries in 4.74 PP/03.
832 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
834 PP/04 New "dns_use_edns0" global option.
836 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
839 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
840 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
842 TK/01 Updated PolarSSL code to 0.14.2.
843 Bugzilla 1097. Patch from Andreas Metzler.
845 PP/07 Catch divide-by-zero in ${eval:...}.
848 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
851 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
852 format-string attack -- SECURITY: remote arbitrary code execution.
854 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
855 time unintentionally subject to list matching rules, letting the header
856 cause arbitrary Exim lookups (of items which can occur in lists, *not*
857 arbitrary string expansion). This allowed for information disclosure.
859 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
860 INT_MIN/-1 -- value coerced to INT_MAX.
866 NM/01 Workround for PCRE version dependency in version reporting
869 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
870 This fixes portability to compilers other than gcc, notably
871 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
873 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
874 makefiles for portability to HP-UX and POSIX correctness.
876 PP/01 Permit LOOKUP_foo enabling on the make command-line.
877 Also via indented variable definition in the Makefile.
878 (Debugging by Oliver Heesakkers).
880 PP/02 Restore caching of spamd results with expanded spamd_address.
881 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
883 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
884 Improves build reliability. Fix from: Frank Elsner
886 NM/02 Fix wide character breakage in the rfc2047 coding
887 Fixes bug 1064. Patch from Andrey N. Oktyabrski
889 NM/03 Allow underscore in dnslist lookups
890 Fixes bug 1026. Patch from Graeme Fowler
892 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
893 Code patches from Adam Ciarcinski of NetBSD.
895 NM/04 Fixed exiqgrep to cope with mailq missing size issue
898 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
899 is logged, to avoid truncation. Patch from John Horne.
901 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
902 Patch from Jakob Hirsch.
904 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
905 SQL string expansion failure details.
906 Patch from Andrey Oktyabrski.
908 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
909 Patch from Simon Arlott.
911 PP/09 New lookups functionality failed to compile on old gcc which rejects
912 extern declarations in function scope.
913 Patch from Oliver Fleischmann
915 PP/10 Use sig_atomic_t for flags set from signal handlers.
916 Check getgroups() return and improve debugging.
917 Fixed developed for diagnosis in bug 927 (which turned out to be
920 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
921 Patch from Mark Zealey.
923 PP/12 Bugzilla 1056: Improved spamd server selection.
924 Patch from Mark Zealey.
926 PP/13 Bugzilla 1086: Deal with maildir quota file races.
927 Based on patch from Heiko Schlittermann.
929 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
930 Patch from Uwe Doering, sign-off by Michael Haardt.
932 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
933 variable declaration deep within a block. Bug and patch from
936 PP/15 lookups-Makefile IRIX compatibilty coercion.
938 PP/16 Make DISABLE_DKIM build knob functional.
940 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
941 Patch by Simon Arlott
943 TF/03 Fix valgrind.h portability to C89 compilers that do not support
944 variable argument macros. Our copy now differs from upstream.
950 TF/01 Failure to get a lock on a hints database can have serious
951 consequences so log it to the panic log.
953 TF/02 Log LMTP confirmation messages in the same way as SMTP,
954 controlled using the smtp_confirmation log selector.
956 TF/03 Include the error message when we fail to unlink a spool file.
958 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
959 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
960 for maintaining out-of-tree patches for some time.
962 PP/01 Bugzilla 139: Documentation and portability issues.
963 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
964 Handle per-OS dynamic-module compilation flags.
966 PP/02 Let /dev/null have normal permissions.
967 The 4.73 fixes were a little too stringent and complained about the
968 permissions on /dev/null. Exempt it from some checks.
969 Reported by Andreas M. Kirchwitz.
971 PP/03 Report version information for many libraries, including
972 Exim version information for dynamically loaded libraries. Created
973 version.h, now support a version extension string for distributors
974 who patch heavily. Dynamic module ABI change.
976 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
977 privilege escalation vulnerability whereby the Exim run-time user
978 can cause root to append content of the attacker's choosing to
981 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
984 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
985 If dropping privileges for untrusted macros, we disabled normal logging
986 on the basis that it would fail; for the Exim run-time user, this is not
987 the case, and it resulted in successful deliveries going unlogged.
988 Fixed. Reported by Andreas Metzler.
994 PP/01 Date: & Message-Id: revert to normally being appended to a message,
995 only prepend for the Resent-* case. Fixes regression introduced in
996 Exim 4.70 by NM/22 for Bugzilla 607.
998 PP/02 Include check_rfc2047_length in configure.default because we're seeing
999 increasing numbers of administrators be bitten by this.
1001 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1003 PP/03 Bugzilla 994: added openssl_options main configuration option.
1005 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1007 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1009 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1011 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1012 without a peer certificate, leading to a segfault because of an
1013 assumption that peers always have certificates. Be a little more
1014 paranoid. Problem reported by Martin Tscholak.
1016 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1017 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1018 NB: ClamAV planning to remove STREAM in "middle of 2010".
1019 CL also introduces -bmalware, various -d+acl logging additions and
1020 more caution in buffer sizes.
1022 PP/09 Implemented reverse_ip expansion operator.
1024 PP/10 Bugzilla 937: provide a "debug" ACL control.
1026 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1028 PP/12 Bugzilla 973: Implement --version.
1030 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1032 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1034 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1036 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1037 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1038 ignore trailing whitespace.
1040 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1042 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1045 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1046 Notification from Dr Andrew Aitchison.
1048 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1049 ExtendedDetectionInfo response format.
1050 Notification from John Horne.
1052 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1055 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1056 XSL and documented dependency on system catalogs, with examples of how
1059 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1062 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1063 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1064 configuration file which is writeable by the Exim user or group.
1066 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1067 of configuration files to cover files specified with the -C option if
1068 they are going to be used with root privileges, not just the default
1071 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1072 option (effectively making it always true).
1074 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1075 files to be used while preserving root privileges.
1077 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1078 that rogue child processes cannot use them.
1080 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1081 run-time user, instead of root.
1083 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1084 Exim run-time user without dropping privileges.
1086 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1087 result string, instead of calling string_vformat() twice with the same
1090 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1091 for other users. Others should always drop root privileges if they use
1092 -C on the command line, even for a whitelisted configure file.
1094 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1096 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1102 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1103 $data_path, and $header_path variables; fixed documentation bugs and
1106 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1107 exipick to access non-standard spools, including the "frozen" queue
1110 NM/01 Bugzilla 965: Support mysql stored procedures.
1111 Patch from Alain Williams
1113 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1115 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1116 Patch from Andreas Metzler
1118 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1119 Patch from Kirill Miazine
1121 NM/05 Bugzilla 671: Added umask to procmail example.
1123 JJ/03 installed exipick 20100323.0, fixing doc bug
1125 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1126 directory. Notification and patch from Dan Rosenberg.
1128 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1130 TK/02 Improve log output when DKIM signing operation fails.
1132 MH/01 Treat the transport option dkim_domain as a colon separated
1133 list, not as a single string, and sign the message with each element,
1134 omitting multiple occurences of the same signer.
1136 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1137 Bugzilla 985, 986. Patch by Simon Arlott
1139 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1140 Patch by Simon Arlott
1142 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1143 MBX locking. Notification from Dan Rosenberg.
1149 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1151 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1153 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1155 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1157 NM/04 Bugzilla 915: Flag broken perl installation during build.
1163 TK/01 Added patch by Johannes Berg that expands the main option
1164 "spamd_address" if it starts with a dollar sign.
1166 TK/02 Write list of recipients to X-Envelope-Sender header when building
1167 the mbox-format spool file for content scanning (suggested by Jakob
1170 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1171 (http://www.dcc-servers.net/) support via dccifd. Activated by
1172 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1174 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1175 by Mark Daniel Reidel <mr@df.eu>.
1177 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1178 When building exim an external PCRE library is now needed -
1179 PCRE is a system library on the majority of modern systems.
1180 See entry on PCRE_LIBS in EDITME file.
1182 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1183 conversation. Added nologin parameter to request.
1184 Patch contributed by Kirill Miazine.
1186 TF/01 Do not log submission mode rewrites if they do not change the address.
1188 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1190 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1191 log files in place. Contributed by Roberto Lima.
1193 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1195 TF/03 Bugzilla 615: When checking the local_parts router precondition
1196 after a local_part_suffix or local_part_prefix option, Exim now
1197 does not use the address's named list lookup cache, since this
1198 contains cached lookups for the whole local part.
1200 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1201 Robert Millan. Documentation is in experimental-spec.txt.
1203 TF/04 Bugzilla 668: Fix parallel build (make -j).
1205 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1207 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1208 Patch provided by Jan Srzednicki.
1210 TF/05 Leading white space used to be stripped from $spam_report which
1211 wrecked the formatting. Now it is preserved.
1213 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1214 that they are available at delivery time.
1216 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1218 TF/08 TLS error reporting now respects the incoming_interface and
1219 incoming_port log selectors.
1221 TF/09 Produce a more useful error message if an SMTP transport's hosts
1222 setting expands to an empty string.
1224 NM/06 Bugzilla 744: EXPN did not work under TLS.
1225 Patch provided by Phil Pennock.
1227 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1228 Patch provided by Richard Godbee.
1230 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1231 acl_smtp_notquit, added index entry.
1233 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1234 Patch provided by Eugene Bujak.
1236 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1237 accept(). Patch provided by Maxim Dounin.
1239 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1240 Patch provided by Phil Pennock.
1242 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1244 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1245 Patch provided by Brad "anomie" Jorsch.
1247 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1248 Patch provided by Dean Brooks.
1250 TK/05 Add native DKIM support (does not depend on external libraries).
1252 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1253 Patch provided by Graeme Fowler.
1255 NM/16 Bugzilla 851: Documentation example syntax fix.
1257 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1259 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1262 NM/19 Bugzilla 745: TLS version reporting.
1263 Patch provided by Phil Pennock.
1265 NM/20 Bugzilla 167: bool: condition support.
1266 Patch provided by Phil Pennock.
1268 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1269 clients. Patch provided by Phil Pennock.
1271 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1272 Patch provided by Brad "anomie" Jorsch.
1274 NM/23 Bugzilla 687: Fix misparses in eximstats.
1275 Patch provided by Heiko Schlittermann.
1277 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1278 Patch provided by Heiko Schlittermann.
1280 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1281 plus update to original patch.
1283 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1285 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1286 Patch provided by David Brownlee.
1288 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1290 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1292 NM/30 Bugzilla 888: TLS documentation bugfixes.
1294 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1296 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1297 Unlike the original bugzilla I have changed all shell scripts in src tree.
1299 NM/33 Bugzilla 898: Transport filter timeout fix.
1300 Patch by Todd Rinaldo.
1302 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1303 Patch by Serge Demonchaux.
1305 NM/35 Bugzilla 39: Base64 decode bug fixes.
1306 Patch by Jakob Hirsch.
1308 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1310 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1312 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1314 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1320 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1321 ALT-N's libdkim that I have put here:
1322 http://duncanthrax.net/exim-experimental/
1324 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1325 were called 'true' and it seems that C99 defines that as a reserved
1326 keyword to be used with 'bool' variable types. That means you could
1327 not include C99-style headers which use bools without triggering
1328 build errors in sieve.c.
1330 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1331 as mailq or other aliases. Changed the --help handling significantly
1332 to do whats expected. exim_usage() emits usage/help information.
1334 SC/01 Added the -bylocaldomain option to eximstats.
1336 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1338 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1340 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1346 PH/01 Another patch from the Sieve maintainer.
1348 PH/02 When an IPv6 address is converted to a string for single-key lookup
1349 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1350 dots are used instead of colons so that keys in lsearch files need not
1351 contain colons. This was done some time before quoting was made available
1352 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1353 (notated using the quote facility) so as to distinguish them from IPv4
1354 keys. This meant that lookups for IP addresses in host lists did not work
1355 for iplsearch lookups.
1357 This has been fixed by arranging for IPv6 addresses to be expressed with
1358 colons if the lookup type is iplsearch. This is not incompatible, because
1359 previously such lookups could never work.
1361 The situation is now rather anomolous, since one *can* have colons in
1362 ordinary lsearch keys. However, making the change in all cases is
1363 incompatible and would probably break a number of configurations.
1365 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1368 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1369 conversion specification without a maximum field width, thereby enabling
1370 a rogue spamd server to cause a buffer overflow. While nobody in their
1371 right mind would setup Exim to query an untrusted spamd server, an
1372 attacker that gains access to a server running spamd could potentially
1373 exploit this vulnerability to run arbitrary code as the Exim user.
1375 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1376 $primary_hostname instead of what libspf2 thinks the hosts name is.
1378 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1379 a directory entry by the name of the lookup key. Previously, if a
1380 symlink pointed to a non-existing file or a file in a directory that
1381 Exim lacked permissions to read, a lookup for a key matching that
1382 symlink would fail. Now it is enough that a matching directory entry
1383 exists, symlink or not. (Bugzilla 503.)
1385 PH/03 The body_linecount and body_zerocount variables are now exported in the
1388 PH/04 Added the $dnslist_matched variable.
1390 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1391 This means they are set thereafter only if the connection becomes
1394 PH/06 Added the client_condition to authenticators so that some can be skipped
1395 by clients under certain conditions.
1397 PH/07 The error message for a badly-placed control=no_multiline_responses left
1398 "_responses" off the end of the name.
1400 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1402 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1403 (without spaces) instead of just copying the configuration text.
1405 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1407 PH/11 Added $max_received_linelength.
1409 PH/12 Added +ignore_defer and +include_defer to host lists.
1411 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1412 way in which PCRE > 7.0 is built.
1414 PH/14 Implemented queue_only_load_latch.
1416 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1417 MAIL command. The effect was to mangle the value on 64-bit systems.
1419 PH/16 Another patch from the Sieve maintainer.
1421 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1423 PH/18 If a system quota error occurred while trying to create the file for
1424 a maildir delivery, the message "Mailbox is full" was not appended to the
1425 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1426 only to a quota excession during the actual writing of the file.
1428 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1429 characters?) which causes problems in log lines. The DN values are now
1430 passed through string_printing() before being added to log lines.
1432 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1433 and InterBase are left for another time.)
1435 PH/21 Added message_body_newlines option.
1437 PH/22 Guard against possible overflow in moan_check_errorcopy().
1439 PH/23 POSIX allows open() to be a macro; guard against that.
1441 PH/24 If the recipient of an error message contained an @ in the local part
1442 (suitably quoted, of course), incorrect values were put in $domain and
1443 $local_part during the evaluation of errors_copy.
1449 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1450 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1453 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1454 issue a MAIL command.
1456 PH/02 In an ACL statement such as
1458 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1460 if a client was not listed at all, or was listed with a value other than
1461 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1462 the condition was not true (as it should be), so access was not denied.
1463 The bug was that the ! inversion was incorrectly passed on to the second
1464 item. This has been fixed.
1466 PH/03 Added additional dnslists conditions == and =& which are different from
1467 = and & when the dns lookup returns more than one IP address.
1469 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1470 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1472 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1473 FSYNC, which compiles an option called disable_fsync that allows for
1474 bypassing fsync(). The documentation is heavily laced with warnings.
1476 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1478 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1479 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1480 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1481 including adding "make clean"; (3) Added -fPIC when compiling the test
1482 dynamically loaded module, to get rid of a warning.
1484 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1485 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1486 = 0s. The bug is otherwise harmless.
1488 PH/07 There was a bug in the dovecot authenticator such that the value of
1489 $auth1 could be overwritten, and so not correctly preserved, after a
1490 successful authentication. This usually meant that the value preserved by
1491 the server_setid option was incorrect.
1493 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1495 PH/09 Installed PCRE release 7.0.
1497 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1498 run for batched SMTP input. It is now run at the start of every message
1499 in the batch. While fixing this I discovered that the process information
1500 (output by running exiwhat) was not always getting set for -bs and -bS
1501 input. This is fixed, and it now also says "batched" for BSMTP.
1503 PH/11 Added control=no_pipelining.
1505 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1506 patch, slightly modified), and move the expansion of helo_data till after
1507 the connection is made in the smtp transport (so it can use these
1510 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1512 PH/14 Added log_selector = +pid.
1514 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1516 PH/16 Add ${if forany and ${if forall.
1518 PH/17 Added dsn_from option to vary the From: line in DSNs.
1520 PH/18 Flush SMTP output before performing a callout, unless control =
1521 no_callout_flush is set.
1523 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1524 was true (the default) a successful delivery failed to delete the retry
1525 item, thus causing premature timeout of the address. The bug is now
1528 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1530 PH/21 Long custom messages for fakedefer and fakereject are now split up
1531 into multiline reponses in the same way that messages for "deny" and
1532 other ACL rejections are.
1534 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1535 with slight modification.
1537 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1538 draft, changing the syntax and factoring some duplicate code".
1540 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1541 for deliveries of the second and subsequent messages over the same SMTP
1544 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1545 ${reduce, with only minor "tidies".
1547 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1549 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1550 expansion side effects.
1552 PH/27 When a message times out after an over-quota error from an Exim-imposed
1553 quota, the bounce message says "mailbox is full". This message was not
1554 being given when it was a system quota that was exceeded. It now should
1557 MH/03 Made $recipients available in local_scan(). local_scan() already has
1558 better access to the recipient list through recipients_list[], but
1559 $recipients can be useful in postmaster-provided expansion strings.
1561 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1562 in the case of a MAIL command with additional options following the
1563 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1564 were accidentally chopped off.
1566 PH/29 SMTP synchronization checks are implemented when a command is read -
1567 there is a check that no more input is waiting when there shouldn't be
1568 any. However, for some commands, a delay in an ACL can mean that it is
1569 some time before the response is written. In this time, more input might
1570 arrive, invalidly. So now there are extra checks after an ACL has run for
1571 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1572 pipelining has not been advertised.
1574 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1576 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1577 returned in the SMTP response when smtp_return_error_details was false.
1578 This has been fixed.
1580 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1581 instead of the C I/O that was originally supplied, because problems were
1582 reported on Solaris.
1584 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1585 Exim which did not show up earlier: it was assuming that a call to
1586 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1587 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1588 was a macro that became an assignment, so it seemed to work. This has
1589 changed to a proper function call with a void return, hence the compile
1590 error. Exim's code has been fixed.
1592 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1595 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1597 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1599 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1600 because it turns out to be too big or there is a timeout) there is no
1601 "Completed" line in the log. When some messages of this type were
1602 selected by exigrep, they were listed as "not completed". Others were
1603 picked up by some special patterns. I have improved the selection
1604 criteria to be more general.
1606 PH/38 The host_find_failed option in the manualroute router can now be set
1607 to "ignore", to completely ignore a host whose IP address cannot be
1608 found. If all hosts are ignored, the behaviour is controlled by the new
1609 host_all_ignored option.
1611 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1612 homing or because of multiple MX records with /mx) generated more than
1613 one IP address, and the following item turned out to be the local host,
1614 all the secondary addresses of the first item were incorrectly removed
1615 from the list, along with the local host and any following hosts (which
1616 is what is supposed to happen).
1618 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1619 whoever called Exim into the -H file. In the case of the daemon it was
1620 behaving confusingly. When first started, it used values for whoever
1621 started the daemon, but after a SIGHUP it used the Exim user (because it
1622 calls itself on a restart). I have changed the code so that it now always
1625 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1626 message are rejected with the same error (e.g. no authentication or bad
1627 sender address), and a DATA command is nevertheless sent (as can happen
1628 with PIPELINING or a stupid MUA), the error message that was given to the
1629 RCPT commands is included in the rejection of the DATA command. This is
1630 intended to be helpful for MUAs that show only the final error to their
1633 PH/42 Another patch from the Sieve maintainer.
1635 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1636 Eximstats - Fixed some broken HTML links and added missing column headers
1638 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1639 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1641 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1643 PH/43 Yet another patch from the Sieve maintainer.
1645 PH/44 I found a way to check for a TCP/IP connection going away before sending
1646 the response to the final '.' that terminates a message, but only in the
1647 case where the client has not sent further data following the '.'
1648 (unfortunately, this is allowed). However, in many cases there won't be
1649 any further data because there won't be any more messages to send. A call
1650 to select() can be used: if it shows that the input is "ready", there is
1651 either input waiting, or the socket has been closed. An attempt to read
1652 the next input character can distinguish the two cases. Previously, Exim
1653 would have sent an OK response which the client would never have see.
1654 This could lead to message repetition. This fix should cure that, at
1655 least in a lot of common cases.
1657 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1658 advertised in response to EHLO.
1664 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1665 fixed by 4.65/MH/01 (is this a record?) are fixed:
1667 (i) An empty string was always treated as zero by the numeric comparison
1668 operators. This behaviour has been restored.
1670 (ii) It is documented that the numeric comparison operators always treat
1671 their arguments as decimal numbers. This was broken in that numbers
1672 starting with 0 were being interpreted as octal.
1674 While fixing these problems I realized that there was another issue that
1675 hadn't been noticed. Values of message_size_limit (both the global option
1676 and the transport option) were treated as octal if they started with 0.
1677 The documentation was vague. These values are now always treated as
1678 decimal, and I will make that clear in the documentation.
1684 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1685 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1688 MH/01 Don't check that the operands of numeric comparison operators are
1689 integers when their expansion is in "skipping" mode (fixes bug
1690 introduced by 4.64-PH/07).
1692 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1693 child addresses, Exim now panics and dies. Previously, because the count
1694 is held in a short int, deliveries were likely to be lost. As such a
1695 large number of recipients for a single message is ridiculous
1696 (performance will be very, very poor), I have chosen to impose a limit
1697 rather than extend the field.
1703 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1704 leftover -K file (the existence of which was triggered by #402).
1705 While we were at it, introduced process PID as part of the -K
1706 filename. This should rule out race conditions when creating
1709 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1710 processing considerably. Previous code took too long for large mails,
1711 triggering a timeout which in turn triggers #401.
1713 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1714 in the DK code in transports.c. sendfile() is not really portable,
1715 hence the _LINUX specificness.
1717 TF/01 In the add_headers option to the mail command in an Exim filter,
1718 there was a bug that Exim would claim a syntax error in any
1719 header after the first one which had an odd number of characters
1722 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1723 callout verification, Exim cached a "reject" for the entire domain. This
1724 is correct for most verifications, but it is not correct for a recipient
1725 verification with use_sender or use_postmaster set, because in that case
1726 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1727 case of MAIL FROM:<> rejection from other early rejections (e.g.
1728 rejection of HELO). When verifying a recipient using a non-null MAIL
1729 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1730 Whatever the result of the callout, the value of the domain cache is
1731 left unchanged (for any other kind of callout, getting as far as trying
1732 RCPT means that the domain itself is ok).
1734 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1737 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1738 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1739 session, the daemon ignores ECONNECT errors and logs others; it now
1740 ignores EPIPE as well.
1742 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1743 (quoted-printable decoding).
1745 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1746 later the small subsequent patch to fix an introduced bug.
1748 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1750 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1752 PH/08 An error is now given if message_size_limit is specified negative.
1754 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1755 to be given (somewhat) arbitrary names.
1757 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1760 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1761 miscellaneous code fixes
1763 PH/10 Added the log_reject_target ACL modifier to specify where to log
1766 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1767 hostname. This is wrong, because it relates to the incoming message (and
1768 probably the interface on which it is arriving) and not to the outgoing
1769 callout (which could be using a different interface). This has been
1770 changed to use the value of the helo_data option from the smtp transport
1771 instead - this is what is used when a message is actually being sent. If
1772 there is no remote transport (possible with a router that sets up host
1773 addresses), $smtp_active_hostname is used.
1775 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1776 tweaks were necessary in order to get it to work (see also 21 below):
1777 (a) The code assumed that strncpy() returns a negative number on buffer
1778 overflow, which isn't the case. Replaced with Exim's string_format()
1780 (b) There were several signed/unsigned issues. I just did the minimum
1781 hacking in of casts. There is scope for a larger refactoring.
1782 (c) The code used strcasecmp() which is not a standard C function.
1783 Replaced with Exim's strcmpic() function.
1784 (d) The code set only $1; it now sets $auth1 as well.
1785 (e) A simple test gave the error "authentication client didn't specify
1786 service in request". It would seem that Dovecot has changed its
1787 interface. Fortunately there's a specification; I followed it and
1788 changed what the client sends and it appears to be working now.
1790 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1793 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1794 address A is aliased to B and C, where B exists and C does not. Without
1795 -v the output is "A verified" because verification stops after a
1796 successful redirection if more than one address is generated. However,
1797 with -v the child addresses are also verified. Exim was outputting "A
1798 failed to verify" and then showing the successful verification for C,
1799 with its parentage. It now outputs "B failed to verify", showing B's
1800 parentage before showing the successful verification of C.
1802 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1803 look up a TXT record in a specific list after matching in a combined
1806 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1807 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1808 they consult the DNS. I had assumed they would set it the way they
1809 wanted; and indeed my experiments on Linux seem to show that in some
1810 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1811 To be on the safe side, however, I have now made the interface to
1812 host_find_byname() similar to host_find_bydns(), with an argument
1813 containing the DNS resolver options. The host_find_byname() function now
1814 sets these options at its start, just as host_find_bydns() does. The smtp
1815 transport options dns_qualify_single and dns_search_parents are passed to
1816 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1817 of host_find_byname() use the default settings of RES_DEFNAMES
1818 (qualify_single) but not RES_DNSRCH (search_parents).
1820 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1821 spool_read_header() do less string testing, by means of a preliminary
1822 switch on the second character of optional "-foo" lines. (This is
1823 overdue, caused by the large number of possibilities that now exist.
1824 Originally there were few.) While I was there, I also converted the
1825 str(n)cmp tests so they don't re-test the leading "-" and the first
1826 character, in the hope this might squeeze out yet more improvement.
1828 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1829 flag allowing group syntax was set by the header_syntax check but not
1830 turned off, possible causing trouble later; (2) The flag was not being
1831 set at all for the header_verify test, causing "group"-style headers to
1832 be rejected. I have now set it in this case, and also caused header_
1833 verify to ignore an empty address taken from a group. While doing this, I
1834 came across some other cases where the code for allowing group syntax
1835 while scanning a header line wasn't quite right (mostly, not resetting
1836 the flag correctly in the right place). These bugs could have caused
1837 trouble for malformed header lines. I hope it is now all correct.
1839 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1840 with the "reply" argument non-NULL. The code, however (which originally
1841 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1842 but it didn't always do it. This confused somebody who was copying the
1843 code for some other use. I have removed all the tests.
1845 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1846 feature that was used to support insecure browsers during the U.S. crypto
1847 embargo. It requires special client support, and Exim is probably the
1848 only MTA that supported it -- and would never use it because real RSA is
1849 always available. This code has been removed, because it had the bad
1850 effect of slowing Exim down by computing (never used) parameters for the
1851 RSA_EXPORT functionality.
1853 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1854 authenticator to fail if there's a tab character in the incoming data
1855 (there should never be unless someone is messing about, as it's supposed
1856 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1857 if the connection is using TLS or if the remote IP is the same as the
1858 local IP, and the "valid-client-cert option" if a client certificate has
1861 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1862 authenticators. This can be used for authorization after authentication
1863 succeeds. (In the case of plaintext, it servers for both authentication
1866 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1867 if any retry times were supplied.
1869 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1870 connection, where there is no client IP address to check. In this
1871 situation, the verify now always succeeds.
1873 PH/25 Applied John Jetmore's -Mset patch.
1875 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1877 PH/27 In a string expansion for a processed (not raw) header when multiple
1878 headers of the same name were present, leading whitespace was being
1879 removed from all of them, but trailing whitespace was being removed only
1880 from the last one. Now trailing whitespace is removed from each header
1881 before concatenation. Completely empty headers in a concatenation (as
1882 before) are ignored.
1884 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1885 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1887 PH/29 [Removed. This was a change that I later backed out, and forgot to
1888 correct the ChangeLog entry (that I had efficiently created) before
1889 committing the later change.]
1891 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1892 address errors (4xx response to RCPT) over the same connection as other
1893 messages routed to the same hosts. Such deliveries are always "forced",
1894 so retry times are not inspected. This resulted in far too many retries
1895 for the affected addresses. The effect occurred only when there were more
1896 hosts than the hosts_max_try setting in the smtp transport when it had
1897 the 4xx errors. Those hosts that it had tried were not added to the list
1898 of hosts for which the message was waiting, so if all were tried, there
1899 was no problem. Two fixes have been applied:
1901 (i) If there are any address or message errors in an SMTP delivery, none
1902 of the hosts (tried or untried) are now added to the list of hosts
1903 for which the message is waiting, so the message should not be a
1904 candidate for sending over the same connection that was used for a
1905 successful delivery of some other message. This seems entirely
1906 reasonable: after all the message is NOT "waiting for some host".
1907 This is so "obvious" that I'm not sure why it wasn't done
1908 previously. Hope I haven't missed anything, but it can't do any
1909 harm, as the worst effect is to miss an optimization.
1911 (ii) If, despite (i), such a delivery is accidentally attempted, the
1912 routing retry time is respected, so at least it doesn't keep
1913 hammering the server.
1915 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1916 in ${readsocket because some servers need this prod.
1918 PH/32 Added some extra debug output when updating a wait-xxx database.
1920 PH/33 The hint "could be header name not terminated by colon", which has been
1921 given for certain expansion errors for a long time, was not being given
1922 for the ${if def:h_colon_omitted{... case.
1924 PH/34 The spec says: "With one important exception, whenever a domain list is
1925 being scanned, $domain contains the subject domain." There was at least
1926 one case where this was not true.
1928 PH/35 The error "getsockname() failed: connection reset by peer" was being
1929 written to the panic log as well as the main log, but it isn't really
1930 panic-worthy as it just means the connection died rather early on. I have
1931 removed the panic log writing for the ECONNRESET error when getsockname()
1934 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1935 runs only) independently of the message's sender address. This meant
1936 that, if the 4xx error was in fact related to the sender, a different
1937 message to the same recipient with a different sender could confuse
1938 things. In particualar, this can happen when sending to a greylisting
1939 server, but other circumstances could also provoke similar problems.
1940 I have changed the default so that the retry time for these errors is now
1941 based a combination of the sender and recipient addresses. This change
1942 can be overridden by setting address_retry_include_sender=false in the
1945 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1946 remote server are returned as part of bounce messages. This was not
1947 happening for LMTP over a pipe (the lmtp transport), but now it is the
1948 same for both kinds of LMTP.
1950 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1951 in header lines that were in fact CNAMEs. This is no longer the case.
1953 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1954 and queue runs started by the daemon processed all messages. This has
1955 been fixed so that -R and -S can now usefully be given with -q<time>.
1957 PH/40 Import PCRE release 6.7 (fixes some bugs).
1959 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1961 PH/42 Give an error if -q is specified more than once.
1963 PH/43 Renamed the variables $interface_address and $interface_port as
1964 $received_ip_address and $received_port, to make it clear that these
1965 values apply to message reception, and not to the outgoing interface when
1966 a message is delivered. (The old names remain recognized, of course.)
1968 PH/44 There was no timeout on the connect() call when using a Unix domain
1969 socket in the ${readsocket expansion. There now is.
1971 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1972 be meaningful with "accept".
1974 SC/01 Eximstats V1.43
1975 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1977 SC/02 Eximstats V1.44
1978 Use a glob alias rather than an array ref in the generated
1979 parser. This improves both readability and performance.
1981 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1982 Collect SpamAssassin and rejection statistics.
1983 Don't display local sender or destination tables unless
1984 there is data to show.
1985 Added average volumes into the top table text output.
1987 SC/04 Eximstats V1.46
1988 Collect data on the number of addresses (recipients)
1989 as well as the number of messages.
1991 SC/05 Eximstats V1.47
1992 Added 'Message too big' to the list of mail rejection
1993 reasons (thanks to Marco Gaiarin).
1995 SC/06 Eximstats V1.48
1996 Mainlog lines which have GMT offsets and are too short to
1997 have a flag are now skipped.
1999 SC/07 Eximstats V1.49 (Alain Williams)
2000 Added the -emptyok flag.
2002 SC/08 Eximstats V1.50
2003 Fixes for obtaining the IP address from reject messages.
2005 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2006 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2007 whitesspace changes from 4.64-PH/27
2009 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2012 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2013 are found, allow negative numbers in numeric criteria)
2015 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2017 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2018 to match changes made in 4.64-PH/43
2020 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2022 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2023 "domain" is actually an IP address, and force a failure. This locks out
2024 those revolvers/nameservers that support "A-for-A" lookups, in
2025 contravention of the specifications.
2027 PH/48 When a host name was looked up from an IP address, and the subsequent
2028 forward lookup of the name timed out, the host name was left in
2029 $sender_host_name, contrary to the specification.
2031 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2032 restricted to single-key lookups, Exim was not diagnosing an error if
2033 * or *@ was used with a query-style lookup.
2035 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2037 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2038 long ago, but noone interested enough thought of it. Let's just say that
2039 the "1.1" means that there are some new functions that weren't there at
2040 some point in the past.
2042 PH/51 Error processing for expansion failure of helo_data from an smtp
2043 transport during callout processing was broken.
2045 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2046 tested/used via the -bh/-bhc/-bs options.
2048 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2049 bug, fixed in subsequent PCRE releases).
2051 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2052 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2054 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2060 SC/01 Use a glob alias rather than an array ref in eximstats generated
2061 parser. This improves both readability and performance.
2063 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2064 Don't display local sender or destination tables in eximstats unless
2065 there is data to show.
2066 Added average volumes into the eximstats top table text output.
2068 SC/03 Collect data on the number of addresses (recipients) as well
2069 as the number of messages in eximstats.
2071 TF/01 Correct an error in the documentation for the redirect router. Exim
2072 does (usually) call initgroups() when daemonizing.
2074 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2075 with consistent privilege compared to when running as a daemon.
2077 TF/03 Note in the spec that $authenticated_id is not set for local
2078 submissions from trusted users.
2080 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2081 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2083 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2084 by adding some example configuration directives to the default
2085 configuration file. A little bit of work is required to uncomment the
2086 directives and define how usernames and passwords are checked, but
2087 there is now a framework to start from.
2089 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2090 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2091 without this. I don't know how relevant this is to other LDAP libraries.
2093 PH/02 Add the verb name to the "unknown ACL verb" error.
2095 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2097 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2099 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2100 directory not expanded when it should be if an expanded home directory
2101 was set for the address (which is overridden by the transport).
2103 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2106 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2107 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2108 because it is too late at that time, and has no effect.
2110 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2111 security issue with \' (bugzilla #107). I could not use the
2112 PQescapeStringConn() function, because it needs a PGconn value as one of
2115 PH/08 When testing addresses using -bt, indicate those final addresses that
2116 are duplicates that would not cause an additional delivery. At least one
2117 person was confused, thinking that -bt output corresponded to deliveries.
2118 (Suppressing duplicates isn't a good idea as you lose the information
2119 about possibly different redirections that led to the duplicates.)
2121 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2122 systems where poll() doesn't work, in particular OS X.
2124 PH/10 Added more information to debugging output for retry time not reached.
2126 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2127 operations in malware.c.
2129 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2132 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2133 syslog_duplication set false, log lines that would normally be written
2134 both the the main log and to the reject log were not written to syslog at
2137 PH/14 In the default configuration, change the use of "message" in ACL warn
2138 statements to "add_header".
2140 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2141 not followed by a command (e.g. "seen endif").
2143 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2144 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2147 PH/17 Added extra conditions to the default value of delay_warning_condition
2151 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2152 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2153 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2156 The Auto-Submitted: and various List- headers are standardised, whereas I
2157 don't think Precedence: ever was.
2159 PH/18 Refactored debugging code in route_finduser() to show more information,
2160 in particular, the error code if getpwnam() issues one.
2162 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2163 This is apparently needed in addition to the PH/07 change above to avoid
2164 any possible encoding problems.
2166 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2167 but not after initializing Perl.
2169 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2170 output them only if debugging. By default they are written stderr,
2171 apparently, which is not desirable.
2173 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2176 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2179 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2181 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2182 authenticated or an ident call has been made. Suppress the default
2183 values for $authenticated_id and $authenticated_sender (but permit -oMai
2184 and -oMas) when testing with -bh.
2186 PH/24 Re-jigged the order of the tests in the default configuration so that the
2187 tests for valid domains and recipients precede the DNS black list and CSA
2188 tests, on the grounds that those ones are more expensive.
2190 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2191 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2192 This bug exists in every version of Exim that I still have, right back to
2195 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2196 However, an attempt to turn on case-sensitivity in a regex key by
2197 including (?-i) didn't work because the subject string was already
2198 lowercased, and the effects were non-intuitive. It turns out that a
2199 one-line patch can be used to allow (?-i) to work as expected.
2205 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2206 other effects) broke the use of negated acl sub-conditions.
2208 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2211 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2212 "Deny" causes Exim to reject the incoming connection with a 554 error.
2213 Unfortunately, if there is a major crisis, such as a disk failure,
2214 tcp-wrappers gives "deny", whereas what one would like would be some
2215 kind of temporary error. A kludge has been added to help with this.
2216 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2217 554 error is used if errno is still zero or contains ENOENT (which occurs
2218 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2221 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2223 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2224 errors. Otherwise a message that provokes a temporary error (when other
2225 messages do not) can cause a whole host to time out.
2227 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2228 addresses were routed directly to files or pipes from a redirect router.
2229 File deliveries just didn't batch; pipe deliveries might have suffered
2232 PH/06 A failure to get a lock for a hints database would erroneously always say
2233 "Failed to get write lock", even when it was really a read lock.
2235 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2236 of 0600. This has been changed to use the value of the lockfile_mode
2237 option (which defaults to 0600).
2239 PH/08 Applied small patch from the Sieve maintainer.
2241 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2242 folder from quota calculations, a direct delivery into this folder messed
2243 up the contents of the maildirsize file. This was because the regex was
2244 used only to exclude .Trash (or whatever) when the size of the mailbox
2245 was calculated. There was no check that a delivery was happening into an
2246 excluded directory. This bug has been fixed by ignoring all quota
2247 processing for deliveries into excluded directories.
2249 PH/10 Added the maildirfolder_create_regex option to appendfile.
2255 PH/01 The code for finding all the local interface addresses on a FreeBSD
2256 system running IPv6 was broken. This may well have applied to all BSD
2257 systems, as well as to others that have similar system calls. The broken
2258 code found IPv4 interfaces correctly, but gave incorrect values for the
2259 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2260 that it would not match correctly against @[] and not recognize the IPv6
2263 PH/02 The ipliteral router was not recognizing addresses of the form user@
2264 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2266 PH/03 Added disable_ipv6.
2268 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2269 decoded form, because it is most often used to construct To: headers
2270 lines in autoreplies, and the decoded form may well be syntactically
2271 invalid. However, $reply_address has leading white space removed, and all
2272 newlines turned into spaces so that the autoreply transport does not
2275 PH/05 If group was specified without a user on a router, and no group or user
2276 was specified on a transport, the group from the router was ignored.
2278 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2279 for visible compile-time settings that can be used to change these
2280 numbers, for those that want even more. Backwards compatibility with old
2281 spool files has been maintained. However, going back to a previous Exim
2282 release will lost any variables that are in spool files.
2284 PH/07 Two small changes when running in the test harness: increase delay when
2285 passing a TCP/IP connection to a new process, in case the original
2286 process has to generate a bounce, and remove special handling of
2287 127.0.0.2 (sic), which is no longer necessary.
2289 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2290 be the same on different OS.
2292 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2295 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2296 whether --show-vars was specified or not
2298 JJ/02 exipick: Added support for new ACL variable spool format introduced
2301 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2302 syntactically invalid From: or Reply-to: line, and a filter used this to
2303 generate an autoreply, and therefore failed to obtain an address for the
2304 autoreply, Exim could try to deliver to a non-existent relative file
2305 name, causing unrelated and misleading errors. What now happens is that
2306 it logs this as a hard delivery error, but does not attempt to create a
2309 PH/11 The exinext utility has a -C option for testing purposes, but although
2310 the given file was scanned by exinext itself; it wasn't being passed on
2311 when Exim was called.
2313 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2314 an end-of-file indication when reading a command response.
2316 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2317 compiled. In many other places in Exim, IPv6 addresses are always
2318 recognized, so I have changed this. It also means that IPv4 domain
2319 literals of the form [IPV4:n.n.n.n] are now always recognized.
2321 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2322 used if the router is not running as root, for example, when verifying at
2323 ACL time, or when using -bh. The debugging output from this situation was
2324 non-existent - all you got was a failure to exec. I have made two
2327 (a) Failures to set uid/gid, the current directory, or a process leader
2328 in a subprocess such as that created by queryprogram now generate
2329 suitable debugging ouput when -d is set.
2331 (b) The queryprogram router detects when it is not running as root,
2332 outputs suitable debugging information if -d is set, and then runs
2333 the subprocess without attempting to change uid/gid.
2335 PH/15 Minor change to Makefile for building test_host (undocumented testing
2338 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2339 additional section of a DNS packet that returns MX or SRV records.
2340 Instead, it always explicitly searches for A/AAAA records. This avoids
2341 major problems that occur when a DNS server includes only records of one
2342 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2343 fixed another bug: if SRV records were looked up and the corresponding
2344 address records were *not* found in the additional section, the port
2345 values from the SRV records were lost.
2347 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2348 using the correct key (the original address) when searching the retry
2349 rules in order to find which one to use for generating the retry hint.
2351 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2352 adding the default one. Similarly, if it contains a Reply-To: header, the
2353 errors_reply_to option, if set, is not used.
2355 PH/19 When calculating a retry time, Exim used to measure the "time since
2356 failure" by looking at the "first failed" field in the retry record. Now
2357 it does not use this if it is later than than the arrival time of the
2358 message. Instead it uses the arrival time. This makes for better
2359 behaviour in cases where some deliveries succeed, thus re-setting the
2360 "first failed" field. An example is a quota failure for a huge message
2361 when small messages continue to be delivered. Without this change, the
2362 "time since failure" will always be short, possible causing more frequent
2363 delivery attempts for the huge message than are intended.
2364 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2366 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2367 $1, $2, $3) because the numerical variables can be reset during some
2368 expansion items (e.g. "match"), thereby losing the authentication data.
2370 PH/21 Make -bV show the size of off_t variables so that the test suite can
2371 decide whether to run tests for quotas > 2G.
2373 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2374 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2375 filecount value is greater than 2G or if a quota value is greater than 2G
2376 on a system where the size of off_t is not greater than 4, a panic error
2379 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2380 never match. The debug and -bh output now contains an explicit error
2381 message indicating a malformed IPv4 address or mask.
2383 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2384 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2385 PH/23 above applies.
2387 PH/25 Do not write to syslog when running in the test harness. The only
2388 occasion when this arises is a failure to open the main or panic logs
2389 (for which there is an explicit test).
2391 PH/26 Added the /no_tell option to "control=freeze".
2393 PH/27 If a host name lookup failed very early in a connection, for example, if
2394 the IP address matched host_lookup and the reverse lookup yielded a name
2395 that did not have a forward lookup, an error message of the form "no IP
2396 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2397 could be logged. Now it outputs the IP address instead of "NULL".
2399 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2400 allows the sender and the authenticated sender to be set when
2401 submitting a message from within Exim. Since child_open_exim() is
2402 documented for local_scan(), the new function should be too.
2404 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2405 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2406 results in an empty string is now treated as unset.
2408 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2410 PH/31 Added sender_verify_fail logging option.
2412 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2413 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2414 tidied the source and removed it altogether.
2416 PH/33 When a queue run was abandoned because the load average was too high, a
2417 log line was always written; now it is written only if the queue_run log
2418 selector is set. In addition, the log line for abandonment now contains
2419 information about the queue run such as the pid. This is always present
2420 in "start" and "stop" lines but was omitted from the "abandon" line.
2422 PH/34 Omit spaces between a header name and the colon in the error message that
2423 is given when verify = headers_syntax fails (if there are lots of them,
2424 the message gets confusing).
2426 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2427 names, as there are now some PTR records that contain slashes. This check
2428 is only to protect against broken name servers that fall over on strange
2429 characters, so the fact that it applies to all lookups doesn't matter.
2431 PH/36 Now that the new test suite is complete, we can remove some of the
2432 special code in Exim that was needed for the old test suite. For example,
2433 sorting DNS records because real resolvers return them in an arbitrary
2434 order. The new test suite's fake resolver always returns records in the
2437 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2438 bounces) except when queue_only is set, to avoid logging races between
2439 the different processes.
2441 PH/38 Panic-die if .include specifies a non-absolute path.
2443 PH/39 A tweak to the "H" retry rule from its user.
2445 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2446 a label. They prevented compilation on older perls.
2448 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2449 a warning to be raised on newish perls.
2451 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2452 on queue. Changes to match documented behaviour of showing count of
2453 messages matching specified criteria.
2455 PH/40 Changed the default ident timeout from 30s to 5s.
2457 PH/41 Added support for the use of login_cap features, on those BSD systems
2458 that have them, for controlling the resources used by pipe deliveries.
2460 PH/42 The content-scanning code uses fopen() to create files in which to put
2461 message data. Previously it was not paying any attention to the mode of
2462 the files. Exim runs with umask(0) because the rest of the code creates
2463 files with open(), and sets the required mode explicitly. Thus, these
2464 files were ending up world-writeable. This was not a big issue, because,
2465 being within the spool directory, they were not world-accessible. I have
2466 created a function called modefopen, which takes an additional mode
2467 argument. It sets umask(777), creates the file, chmods it to the required
2468 mode, then resets the umask. All the relevant calls to fopen() in the
2469 content scanning code have been changed to use this function.
2471 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2472 to 24 hours. This avoids potential overflow problems when processing G
2473 and H retry rules. I suspect nobody ever tinkers with this value.
2475 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2477 PH/45 When the plaintext authenticator is running as a client, the server's
2478 challenges are checked to ensure they are valid base64 strings. By
2479 default, the authentication attempt is cancelled if an invalid string is
2480 received. Setting client_ignore_invalid_base64 true ignores these errors.
2481 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2482 they are received. Thus, the responses can be made to depend on the
2483 challenges. If an invalid string is ignored, an empty string is placed in
2486 PH/46 Messages that are created by the autoreply transport now contains a
2487 References: header, in accordance with RFCs 2822 and 3834.
2489 PH/47 Added authenticated_sender_force to the smtp transport.
2491 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2493 PH/49 Installed latest patch from the Sieve maintainer.
2495 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2496 was also set, the appendfile transport was unnecessarily scanning a
2497 directory of message files (e.g. for maildir delivery) to find the count
2498 of files (along with the size), even though it did not need this
2499 information. It now does the scan only if it needs to find either the
2500 size of the count of files.
2502 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2504 PH/52 Two bugs concerned with error handling when the smtp transport is
2507 (i) Exim was not creating retry information for temporary errors given
2508 for individual recipients after the DATA command when the smtp transport
2509 was used in LMTP mode. This meant that they could be retried too
2510 frequently, and not timed out correctly.
2512 (ii) Exim was setting the flag that allows error details to be returned
2513 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2514 recipients that were returned after the DATA command.
2516 PH/53 This is related to PH/52, but is more general: for any failing address,
2517 when detailed error information was permitted to be returned to the
2518 sender, but the error was temporary, then after the final timeout, only
2519 "retry timeout exceeded" was returned. Now it returns the full error as
2520 well as "retry timeout exceeded".
2522 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2523 do this, and (what is worse) MTAs that accept it.
2525 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2526 will now be deprecated.
2528 PH/56 New os.c-cygwin from the Cygwin maintainer.
2530 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2531 formats (previously only available in exim formats via -bpr, -bpru,
2532 and -bpra. Now also available in native and exiqgrep formats)
2534 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2535 with very large, slow to parse queues
2537 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2539 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2541 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2542 responses to authentication challenges, though it was showing the
2543 challenges; (ii) I've removed the CR characters from the debug output for
2546 PH/58 Allow for the insertion of a newline as well as a space when a string
2547 is turned into more than one encoded-word during RFC 2047 encoding. The
2548 Sieve code now uses this.
2550 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2551 data_4xx, lost_connection, tls_required.
2553 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2554 message was being sent as an SMTP response.
2556 PH/61 Add -l and -k options to exicyclog.
2558 PH/62 When verifying, if an address was redirected to one new address, so that
2559 verification continued, and the new address failed or deferred after
2560 having set something in $address_data, the value of $address_data was not
2561 passed back to the ACL. This was different to the case when no
2562 redirection occurred. The value is now passed back in both cases.
2564 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2565 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2566 use login_cap.h, so on its own it isn't the distinguishing feature. The
2567 new name refers directly to the setclassresources() function.
2569 PH/65 Added configuration files for NetBSD3.
2571 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2573 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2574 is preferred over IPv4.
2576 PH/68 The bounce_return_message and bounce_return_body options were not being
2577 honoured for bounces generated during the reception of non-SMTP messages.
2578 In particular, this applied to messages rejected by the ACL. This bug has
2579 been fixed. However, if bounce_return_message is true and bounce_return_
2580 body is false, the headers that are returned for a non-SMTP message
2581 include only those that have been read before the error was detected.
2582 (In the case of an ACL rejection, they have all been read.)
2584 PH/69 The HTML version of the specification is now built in a directory called
2585 spec_html instead of spec.html, because the latter looks like a path with
2586 a MIME-type, and this confuses some software.
2588 PH/70 Catch two compiler warnings in sieve.c.
2590 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2591 function verify_get_ident() calls ip_connect() to connect a socket, but
2592 if the "connect()" function timed out, ip_connect() used to close the
2593 socket. However, verify_get_ident() also closes the socket later, and in
2594 between Exim writes to the log, which may get opened at this point. When
2595 the socket was closed in ip_connect(), the log could get the same file
2596 descriptor number as the socket. This naturally causes chaos. The fix is
2597 not to close the socket in ip_connect(); the socket should be closed by
2598 the function that creates it. There was only one place in the code where
2599 this was missing, in the iplookup router, which I don't think anybody now
2600 uses, but I've fixed it anyway.
2602 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2603 well as to direct DNS lookups. Otherwise the handling of names in host
2604 lists is inconsistent and therefore confusing.
2610 PH/01 Two changes to the default runtime configuration:
2612 (1) Move the checks for relay_from_hosts and authenticated clients from
2613 after to before the (commented out) DNS black list checks.
2615 (2) Add control=submission to the relay_from_hosts and authenticated
2616 clients checks, on the grounds that messages accepted by these
2617 statements are most likely to be submissions.
2619 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2621 (1) Generate an error if the third argument for the ${prvs expansion is
2624 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2627 (3) Reset the variables that are obtained from the first argument of
2628 ${prvscheck and used in the second argument before leaving the code,
2629 because their memory is reclaimed, so using them afterwards may do
2632 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2633 one (it's much easier than Tom thought :-).
2635 (5) Because of (4), we can now allow for the use of $prvscheck_result
2636 inside the third argument.
2638 PH/03 For some reason, the default setting of PATH when running a command from
2639 a pipe transport was just "/usr/bin". I have changed it to
2642 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2643 anything to be listed in the output from -bV.
2645 PH/05 When a filter generated an autoreply, the entire To: header line was
2646 quoted in the delivery log line, like this:
2648 => >A.N.Other <ano@some.domain> <original@ddress> ...
2650 This has been changed so that it extracts the operative address. There
2651 may be more than one such address. If so, they are comma-separated, like
2654 => >ano@some.domain,ona@other.domain <original@ddress> ...
2656 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2657 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2658 not being looked up in the rDNS to get a host name, Exim was showing the
2659 IP address twice in Received: lines, even though the IP addresses were
2660 identical. For example:
2662 Received: from [1.2.3.4] (helo=[1.2.3.4])
2664 However, if the real host name was known, it was omitting the HELO data
2665 if it matched the actual IP address. This has been tidied up so that it
2666 doesn't show the same IP address twice.
2668 PH/07 When both +timestamp and +memory debugging was on, the value given by
2669 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2670 called by the debug printing, thereby overwriting the timestamp buffer.
2671 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2673 PH/08 When the original message was included in an autoreply transport, it
2674 always said "this is a copy of the message, including all the headers",
2675 even if body_only or headers_only was set. It now gives an appropriate
2678 PH/09 Applied a patch from the Sieve maintainer which:
2680 o fixes some comments
2681 o adds the (disabled) notify extension core
2682 o adds some debug output for the result of if/elsif tests
2683 o points to the current vacation draft in the documentation
2684 and documents the missing references header update
2688 o fixes a bug in processing the envelope test (when testing
2689 multiple envelope elements, the last element determinted the
2692 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2693 Electronic Mail") by including:
2695 Auto-submitted: auto-generated
2697 in the messages that it generates (bounce messages and others, such as
2698 warnings). In the case of bounce messages for non-SMTP mesages, there was
2699 also a typo: it was using "Auto_submitted" (underscore instead of
2700 hyphen). Since every message generated by Exim is necessarily in response
2701 to another message, thes have all been changed to:
2703 Auto-Submitted: auto-replied
2705 in accordance with these statements in the RFC:
2707 The auto-replied keyword:
2709 - SHOULD be used on messages sent in direct response to another
2710 message by an automatic process,
2712 - MUST NOT be used on manually-generated messages,
2714 - MAY be used on Delivery Status Notifications (DSNs) and Message
2715 Disposition Notifications (MDNs),
2717 - MUST NOT be used on messages generated by automatic or periodic
2718 processes, except for messages which are automatic responses to
2721 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2722 to the default Received: header definition.
2724 PH/12 Added log selector acl_warn_skipped (default on).
2726 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2727 variables because (a) they are in the wrong storage pool and (b) even if
2728 they were copied, it wouldn't work properly because of the caching.
2730 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2731 checking when decoding. Apparently there are clients that generate
2732 overlong encoded strings. Why am I not surprised?
2734 PH/15 If the first argument of "${if match_address" was not empty, but did not
2735 contain an "@" character, Exim crashed. Now it writes a panic log message
2736 and treats the condition as false.
2738 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2740 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2741 extension "envelope-auth". The code is finished and in agreement with
2742 other implementations, but there is no documentation so far and in fact,
2743 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2744 not changing the active code.
2746 Print executed "if" and "elsif" statements when debugging is used. This
2747 helps a great deal to understand what a filter does.
2749 Document more things not specified clearly in RFC3028. I had all this
2750 sorted out, when out of a sudden new issues came to my mind. Oops."
2752 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2755 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2756 canonical form (as documented). However, after a host name lookup from
2757 the IP address, check_host() was doing a simple string comparison with
2758 addresses acquired from the DNS when checking that the found name did
2759 have the original IP as one of its addresses. Since any found IPv6
2760 addresses are likely to be in abbreviated form, the comparison could
2761 fail. Luckily, there already exists a function for doing the comparison
2762 by converting both addresses to binary, so now that is used instead of
2763 the text comparison.
2765 PH/20 There was another similar case to PH/19, when a complete host name was
2766 given in a host list; looking up its IP address could give an abbreviated
2767 form, whereas the current host's name might or might not be abbreviated.
2768 The same fix has been applied.
2774 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2775 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2778 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2779 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2781 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2783 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2784 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2785 Also, for "Auto-Submitted", treat anything other than "no" as
2786 non-personal, in accordance with RFC 3834. (Previously it treated
2787 anything starting "auto-" as non-personal.)
2789 TF/01 The control=submission/name=... option had a problem with syntax
2790 errors if the name included a slash character. The /name= option
2791 now slurps the rest of the string, so it can include any characters
2792 but it must come last in the list of options (after /sender_retain
2795 PH/05 Some modifications to the interface to the fake nameserver for the new
2803 TK/01 Added the "success_on_redirect" address verification option. See
2804 NewStuff for rationale and an example.
2806 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2808 PH/02 Patch to exigrep to allow it to work on syslog lines.
2810 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2811 fread() to skip over the body file's header line, because in Cygwin the
2812 header line is locked and is inaccessible.
2814 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2815 co-exist for some time) to make it clear that it is the Exim ID that is
2816 referenced, not the Message-ID: header line.
2818 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2819 string_format() function, because snprintf() does not exist on all
2822 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2823 ${stat: expansion item.
2825 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2826 protocol synchronization error", to keep the pedants happy.
2828 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2829 well as for IRIX systems, when gcc is being used. See the host.c source
2832 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2834 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2837 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2838 transport and to the smtp transport in LMTP mode.
2840 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2842 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2843 run a filter in a subprocess. This could lead to confusion in subsequent
2844 lookups in the parent process. There should also be a search_tidyup() at
2845 the end of the subprocess.
2847 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2848 only if the host matched helo_try_verify_hosts, which caused the
2849 verification to occur when the EHLO/HELO command was issued. The ACL just
2850 tested the remembered result. Now, if a previous verification attempt has
2851 not happened, "verify = helo" does it there and then.
2853 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2855 TK/03 Fix log output including CR from clamd.
2857 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2858 exist provoked a memory error which could cause a segfault.
2860 PH/15 Installed PCRE 6.2
2862 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2864 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2865 of the problem. Specifically, suggested +O2 rather than +O1 for the
2868 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2870 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2871 record, the port was not correctly logged when the outgoing_port log
2872 selector was set (it logged the transort's default port).
2874 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2875 fallback_hosts, and "hosts" in the smtp transport.
2877 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2878 host errors such as "Connection refused".
2880 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2881 authentication with radiusclient 0.4.9:
2883 - Error returned from rc_read_config was caught wrongly
2884 - Username/password not passed on to radius server due to wrong length.
2886 The presumption is that some radiusclient API changes for 4.51/PH/17
2887 were not taken care of correctly. The code is still untested by me (my
2888 Linux distribution still has 0.3.2 of radiusclient), but it was
2889 contributed by a Radius user.
2891 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2892 expanding the "port" option of the smtp transport.
2894 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2895 while reading a MIME header. Thanks to Tom Hughes for a patch.
2897 PH/24 Include config.h inside local_scan.h so that configuration settings are
2900 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2901 that in an ACL for RCPT (for example), you can examine exactly what was
2904 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2905 commands, but it was not correctly comparing the address with the actual
2906 client host address. Thus, it would show the EHLO address in Received:
2907 header lines when this was not necessary.
2909 PH/27 Added the % operator to ${eval:}.
2911 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2912 it should be ignoring failures (because with -C, for example, it has lost
2913 privilege). It wasn't ignoring creation failures other than "already
2916 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2919 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2920 because an input line was too long, either on its own, or by virtue of
2921 too many continuations, the temporary file was not being removed, and the
2922 return code was incorrect.
2924 PH/31 Missing "BOOL" in function definition in filtertest.c.
2926 PH/32 Applied Sieve patches from the maintainer.
2928 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2930 PH/33 Added "verify = not_blind".
2932 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2933 Local/Makefile (with some defaults set). These are used in built scripts
2934 such as exicyclog, but they have never been used in the exim_install
2935 script (though there are many overriding facilities there). I have
2936 arranged that the exim_install script now takes note of these two
2939 PH/35 Installed configuration files for Dragonfly.
2941 PH/36 When a locally submitted message by a trusted user did not contain a
2942 From: header, and the sender address was obtained from -f or from an SMTP
2943 MAIL command, and the trusted user did not use -F to supply a sender
2944 name, $originator_name was incorrectly used when constructing a From:
2945 header. Furthermore, $originator_name was used for submission mode
2946 messages from external hosts without From: headers in a similar way,
2947 which is clearly wrong.
2949 PH/37 Added control=suppress_local_fixups.
2951 PH/38 When log_selector = +received_sender was set, and the addition of the
2952 sender made the log line's construction buffer exactly full, or one byte
2953 less than full, an overflow happened when the terminating "\n" was
2956 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2957 when the result of a list match is failure because a DNS lookup failed.
2959 PH/40 RM_COMMAND is now used in the building process.
2961 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2962 the "build-* directories that it finds.
2964 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2965 address in a domain literal was a prefix of an interface address.
2967 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2968 when verifying a sender address, unless rewrite_headers is false.
2970 PH/44 Wrote a long comment about why errors_to addresses are verified as
2971 recipients, not senders.
2973 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2974 the ratelimit ACL was added.
2976 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2978 PH/46 Added extra information about PostgreSQL errors to the error string.
2980 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2981 avoiding the need to install special zones in a real server. This is
2982 backwards compatible; if it can't find the fake resolver, it drops back.
2983 Thus, both old and new test suites can be run.
2985 TF/02 Added util/ratelimit.pl
2987 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2988 clock is set back in time.
2990 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2991 Candler <B.Candler@pobox.com>.
2993 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2994 OK for addresses that are the result of redirections.
2996 PH/48 A number of further additions for the benefit of the new test suite,
2997 including a fake gethostbyname() that interfaces to the fake DNS resolver
3000 TF/06 The fix for widen_domains has also been applied to qualify_single and
3001 search_parents which are the other dnslookup options that can cause
3004 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3007 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3009 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3010 service" if it encountered an MX record with an empty target hostname.
3011 The message is now "an MX or SRV record indicated no SMTP service".
3013 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3014 if the DNS of the sending site is misconfigured. This is quite a
3015 common situation. This change restores the behaviour of treating a
3016 helo verification defer as a failure.
3018 PH/51 If self=fail was set on a router, the bounce message did not include the
3019 actual error message.
3025 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3027 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3028 command itself ended in error, the underlying message about the transport
3029 filter timeout was being overwritten with the pipe command error. Now the
3030 underlying error message should be appended to the second error message.
3032 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3034 PH/02 Reduce the amount of output that "make" produces by default. Full output
3035 can still be requested.
3037 PH/03 The warning log line about a condition test deferring for a "warn" verb
3038 was being output only once per connection, rather than after each
3039 occurrence (because it was using the same function as for successful
3040 "warn" verbs). This seems wrong, so I have changed it.
3042 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3043 it should not have, which might have caused a crash in the right
3044 circumstances, but probably never did.
3046 PH/04 Installed a modified version of Tony Finch's patch to make submission
3047 mode fix the return path as well as the Sender: header line, and to
3048 add a /name= option so that you can make the user's friendly name appear
3051 TF/03 Added the control = fakedefer ACL modifier.
3053 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3054 Mark Lowes for thorough testing.
3056 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3058 TK/03 Merged latest SRS patch from Miles Wilton.
3060 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3061 with the definition in sysexits.h (which is #included earlier).
3062 Fortunately, Exim does not actually use EX_OK. The code used to try to
3063 preserve the sysexits.h value, by assumimg that macro definitions were
3064 scanned for macro replacements. I have been disabused of this notion,
3065 so now the code just undefines EX_OK before #including unistd.h.
3067 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3068 in the smtp transport. When a block could not be written in a single
3069 write() function, the timeout was being re-applied to each part-write.
3070 This seems wrong - if the receiver was accepting one byte at a time it
3071 would take for ever. The timeout is now adjusted when this happens. It
3072 doesn't have to be particularly precise.
3074 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3075 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3077 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3078 without a domain if the check to <postmaster@domain> fails.
3080 SC/01 Eximstats: added -xls and the ability to specify output files
3081 (patch written by Frank Heydlauf).
3083 SC/02 Eximstats: use FileHandles for outputing results.
3085 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3087 SC/04 Eximstats: fixed display of large numbers with -nvr option
3089 SC/05 Eximstats: fixed merging of reports with empty tables.
3091 SC/06 Eximstats: added the -include_original_destination flag
3093 SC/07 Eximstats: removed tabs and trailing whitespace.
3095 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3097 TK/06 MBOX spool code: Add real "From " MBOX separator line
3098 so the .eml file is really in mbox format (even though
3099 most programs do not really care). Patch from Alex Miller.
3101 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3102 The latter is generated from $received_to and is only set if the
3103 message has one envelope recipient. SA can use these headers,
3104 obviously out-of-the-box. Patch from Alex Miller.
3106 PH/08 The ${def test on a variable was returning false if the variable's
3107 value was "0", contrary to what the specification has always said!
3108 The result should be true unless the variable is empty.
3110 PH/09 The syntax error of a character other than { following "${if
3111 def:variable_name" (after optional whitespace) was not being diagnosed.
3112 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3113 accidental colon was present, for example, could give incorrect results.
3115 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3116 result is used (not including appendfile, where other changes are about
3119 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3120 This involved changing a lot of size variables from int to off_t. It
3121 should work with maildirs and everything.
3123 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3124 spamd dying while we are connected to it.
3126 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3129 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3130 function for BDB 4.3.
3132 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3134 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3135 the sources for PCRE have been re-arranged and more files are now
3138 PH/15 The code I had for printing potentially long long variables in PH/11
3139 above was not the best (it lost precision). The length of off_t variables
3140 is now inspected at build time, and an appropriate printing format (%ld
3141 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3142 to be "long long int" or "long int". This is needed for the internal
3143 formatting function string_vformat().
3145 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3146 the configuration file to be ":syslog", then the script "guesses" where
3147 the logs files are, rather than using the compiled in default. In our
3148 case the guess is not the same as the compiled default, so the script
3149 suddenly stopped working when I started to use syslog. The patch checks
3150 to see if log_file_path is "". If so, it attempts to read it from exim
3151 with no configuration file to get the compiled in version, before it
3152 falls back to the previous guessing code."
3154 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3155 implementing BATV in an Exim configuration. See NewStuff for the gory
3158 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3159 Makefile that are specific to HP-UX.
3161 PH/18 If the "use_postmaster" option was set for a recipient callout together
3162 with the "random" option, the postmaster address was used as the MAIL
3163 FROM address for the random test, but not for the subsequent recipient
3164 test. It is now used for both.
3166 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3167 patch removes a few documentation additions to RFC 3028, because the
3168 latest draft now contains them. It adds the new en;ascii-case comparator
3169 and a new error check for 8bit text in MIME parts. Comparator and
3170 require names are now matched exactly. I enabled the subaddress
3171 extension, but it is not well tested yet (read: it works for me)."
3173 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3174 rework some of the code of TK/09 above to avoid the hardwired use of
3175 "%lld" and "long long". Replaced the call to snprintf() with a call to
3178 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3179 records point to non-existent hosts", "retry timeout exceeded", and
3180 "retry time not reached for any host after a long failure period".
3182 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3183 experimental DomainKeys support:
3185 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3186 (2) On an error such as an illegally used "control", the wrong name for
3187 the control was given.
3189 These problems did NOT occur unless DomainKeys support was compiled.
3191 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3193 PH/24 Added ${if match_ip condition.
3195 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3196 databases so that it will be absolutely obvious if a crash occurs in the
3197 DB library. This is a regular occurrence (often caused by mis-matched
3200 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3201 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3202 picked up on a user's system that detects such things. There doesn't seem
3203 to be a gcc warning option for this - only an attribute that has to be
3204 put on the function's prototype. It seems that in Fedora Core 4 they have
3205 set this on a number of new functions. No doubt there will be more in due
3208 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3209 specify a transport. However, if an address that was verified by such a
3210 router was the subject of a callout, Exim crashed because it tried to
3211 read the rcpt_include_affixes from the non-existent transport. Now it
3212 just assumes that the setting of that option is false. This bug was
3213 introduced by 4.51/PH/31.
3215 PH/28 Changed -d+all to exclude +memory, because that information is very
3216 rarely of interest, but it makes the output a lot bigger. People tend to
3217 do -d+all out of habit.
3219 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3220 code in os-type was giving problems when libc.so lives in lib64, like on
3223 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3224 aren't the modern standard, and it seems that some systems' include files
3225 don't always have them. Exim was already checking for some of the newer
3226 ones like T_AAAA, and defining it itself. I've added checks for all the
3227 record types that Exim uses.
3229 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3230 not automatically generating a new one, as it is supposed to. This
3231 prevented TLS from working. If the file did exist, but contained invalid
3232 data, a new version was generated, as expected. It was only the case of a
3233 non-existent file that was broken.
3235 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3236 with a change in libdomainkeys > 0.64.
3238 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3239 from DNS. If the selector record carries the flag, it now has
3240 precedence over the domain-wide flag.
3242 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3244 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3245 the use of a port name that isn't defined in /etc/services) occurred, the
3246 message was deferred as in a normal delivery, and thus remained on the
3247 spool, instead of being failed because of the mua_wrapper setting. This
3248 is now fixed, and I tidied up some of the mua_wrapper messages at the
3251 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3252 the messages in a hash of arrays rather than using individual hashes.
3253 This is a bit cleaner and results in dramatic memory savings, albeit
3254 at a slight CPU cost.
3256 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3257 as requested by Marc Sherman.
3259 SC/10 Eximstats: added histograms for user specified patterns as requested
3262 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3264 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3265 fopen() in the content-scanning modules that did not already have it.
3271 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3272 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3274 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3276 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3278 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3279 to test Sieve filters that use "vacation".
3281 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3282 that changes the way the GnuTLS parameters are stored in the cache file.
3283 The new format can be generated externally. For backward compatibility,
3284 if the data in the cache doesn't make sense, Exim assumes it has read an
3285 old-format file, and it generates new data and writes a new file. This
3286 means that you can't go back to an older release without removing the
3289 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3290 work if there are any delivery delays because "one_time" forces the
3291 parent to be marked "delivered", so its unseen clone is never tried
3292 again. For this reason, Exim now forbids the simultaneous setting of
3295 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3296 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3297 entry if you want to know the details. The fix, however, neglected to
3298 consider the case when local delivery batching is involved. The test for
3299 "previously delivered" was not happening when checking to see if an
3300 address could be batched with a previous (undelivered) one; under
3301 certain circumstances this could lead to multiple deliveries to the same
3304 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3305 in its include files, and this causes problems building Exim.
3307 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3308 header_syntax) but Exim was just ignoring anything given after a slash.
3309 In particular, this caused confusion with an attempt to use "verify =
3310 reverse_host_lookup/defer_ok". An error is now given when options are
3311 supplied for verify items that do not have them. (Maybe reverse_host_
3312 lookup should have a defer_ok option, but that's a different point.)
3314 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3315 defined by RFC 821) to 2048, because there were problems with some AUTH
3316 commands, and RFC 1869 says the size should be increased for extended
3317 SMTP commands that take arguments.
3319 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3322 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3323 "unknown" error; now it says that the functionality isn't in the binary.
3325 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3326 an address' error message when a string expansion fails (syntax or
3327 whatever). Otherwise the password may appear in the log. Following change
3328 PH/42 below, there is no longer a chance of it appearing in a bounce
3331 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3333 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3334 of its addresses was ever tried. (Bugzilla bug #2.)
3336 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3337 the result incorrectly in the debug output. (It correctly added a newline
3338 to what was transported.)
3340 TF/01 Added $received_time.
3342 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3343 commented out examples of how to interface to a virus scanner and to
3344 SpamAssassin. Also added commented examples of av_scanner and
3345 spamd_address settings.
3347 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3348 and controls are allowed in which ACLs. There were a couple of minor
3349 errors. Some of the entries in the conditions table (which is a table of
3350 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3351 negation of where the condition IS allowed.
3353 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3355 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3356 header file does not have a version number, so I've had to invent a new
3357 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3358 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3359 radiusclient), but it was contributed by a Radius user.
3361 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3362 files or directories, for OpenSSL.
3364 PH/19 When an Exim process that is running as root has to create an Exim log
3365 file, it does so in a subprocess that runs as exim:exim so as to get the
3366 ownership right at creation (otherwise, other Exim processes might see
3367 the file with the wrong ownership). There was no test for failure of this
3368 fork() call, which would lead to the process getting stuck as it waited
3369 for a non-existent subprocess. Forks do occasionally fail when resources
3370 run out. I reviewed all the other calls to fork(); they all seem to check
3373 PH/20 When checking for unexpected SMTP input at connect time (before writing
3374 the banner), Exim was not dealing correctly with a non-positive return
3375 from the read() function. If the client had disconnected by this time,
3376 the result was a log entry for a synchronization error with an empty
3377 string after "input=" when read() returned zero. If read() returned -1
3378 (an event I could not check), uninitialized data bytes were printed.
3379 There were reports of junk text (parts of files, etc) appearing after
3382 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3384 PH/22 Added support for macro redefinition, and (re)definition in between
3385 driver and ACL definitions.
3387 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3388 forgetting to use the resulting value; it was using the unexpanded value.
3390 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3391 hadn't been configured. The fix is from Juergen Kreileder, who
3392 understands it better than I do:
3394 "Here's what I see happening with three configured cyrus_sasl
3395 authenticators configured (plain, login, cram-md5):
3397 On startup auth_cyrus_sasl_init() gets called for each of these.
3398 This means three calls to sasl_listmech() without a specified mech_list.
3399 => SASL tests which mechs of all available mechs actually work
3400 => three warnings about OTP not working
3401 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3403 With the patch, sasl_listmech() also gets called three times. But now
3404 SASL's mech_list option is set to the server_mech specified in the the
3405 authenticator. Or in other words, the answer from sasl_listmech()
3406 gets limited to just the mech you're testing for (which is different
3408 => the return list contains just 'plain' or 'login', 'cram-md5' or
3409 nothing depending on the value of ob->server_mech.
3411 I've just tested the patch: Authentication still works fine,
3412 unavailable mechs specified in the exim configuration are still
3413 caught, and the auth.log warnings about OTP are gone."
3415 PH/25 When debugging is enabled, the contents of the command line are added
3416 to the debugging output, even when log_selector=+arguments is not
3419 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3420 answer is "GNU", and only if the return is "GNU/something" is the answer
3423 PH/27 $acl_verify_message is now set immediately after the failure of a
3424 verification in an ACL, and so is available in subsequent modifiers. In
3425 particular, the message can be preserved by coding like this:
3427 warn !verify = sender
3428 set acl_m0 = $acl_verify_message
3430 Previously, $acl_verify_message was set only while expanding "message"
3431 and "log_message" when a very denied access.
3433 PH/28 Modified OS/os.c-Linux with
3435 -#ifndef OS_LOAD_AVERAGE
3436 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3438 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3439 nomenclature these days.)
3441 PH/29 Installed patch from the Sieve maintainer that adds the options
3442 sieve_useraddress and sieve_subaddress to the redirect router.
3444 PH/30 In these circumstances:
3445 . Two addresses routed to the same list of hosts;
3446 . First host does not offer TLS;
3447 . First host accepts first address;
3448 . First host gives temporary error to second address;
3449 . Second host offers TLS and a TLS session is established;
3450 . Second host accepts second address.
3451 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3452 and peerdn, if requested) that were in fact used only for the second
3455 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3456 attention to any local part prefix or suffix that was matched by the
3457 router that accepted the address. It now behaves in the same way as it
3458 does for delivery: the affixes are removed from the local part unless
3459 rcpt_include_affixes is set on the transport.
3461 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3462 timeout during the DATA phase of an incoming message.
3464 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3465 applied a patch sanctioned by the Sieve maintainer.
3467 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3468 the uid or gid is negative. A case of a negative gid caused this to be
3469 noticed. The fix allows for either to be negative.
3471 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3472 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3475 PH/36 The change PH/12 above was broken. Fixed it.
3477 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3478 the grounds that routing the same address twice would always produce the
3479 same answer. This might have been true once, but it is certainly no
3480 longer true now. Routing a child address may depend on the previous
3481 routing that produced that child. Some complicated redirection strategies
3482 went wrong when messages had multiple recipients, and made Exim's
3483 behaviour dependent on the order in which the addresses were given.
3485 I have moved the duplicate checking until after the routing is complete.
3486 Exim scans the addresses that are assigned to local and remote
3487 transports, and removes any duplicates. This means that more work will be
3488 done, as duplicates will always all be routed, but duplicates are
3489 presumably rare, so I don't expect this is of any significance.
3491 For deliveries to pipes, files, and autoreplies, the duplicate checking
3492 still happens during the routing process, since they are not going to be
3495 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3496 It corrects a timeout issue with spamd. This is Ian's comment: "The
3497 background is that sometimes spamd either never reads data from a
3498 connection it has accepted, or it never writes response data. The exiscan
3499 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3500 blindly assumes that writes won't block so it may never time out."
3502 PH/39 Allow G after quota size as well as K and M.
3504 PH/40 The value set for $authenticated_id in an authenticator may not contain
3505 binary zeroes or newlines because the value is written to log lines and
3506 to spool files. There was no check on this. Now the value is run through
3507 the string_printing() function so that such characters are converted to
3508 printable escape sequences.
3510 PH/41 $message_linecount is a new variable that contains the total number of
3511 lines in the message. Compare $body_linecount, which is the count for the
3514 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3515 bounce and delay warning messages, except in certain special cases, which
3518 (a) An SMTP error message from a remote host;
3519 (b) A message specified in a :fail: redirection;
3520 (c) A message specified in a "fail" command in a system filter;
3521 (d) A message specified in a FAIL return from the queryprogram router;
3522 (e) A message specified by the cannot_route_message router option.
3524 In these cases only, Exim does include the error details in bounce and
3525 warning messages. There are also a few cases where bland messages such
3526 as "unrouteable address" or "local delivery error" are given.
3528 PH/43 $value is now also set for the "else" part of a ${run expansion.
3530 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3531 being worked on, but at least Exim now implements the latest version to
3534 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3535 process to complete was treated as a delivery failure, a timeout while
3536 writing the message to the pipe was logged, but erroneously treated as a
3537 successful delivery. Such timeouts include transport filter timeouts. For
3538 consistency with the overall process timeout, these timeouts are now
3539 treated as errors, giving rise to delivery failures by default. However,
3540 there is now a new Boolean option for the pipe transport called
3541 timeout_defer, which, if set TRUE, converts the failures into defers for
3542 both kinds of timeout. A transport filter timeout is now identified in
3545 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3546 systems where "make" and "gmake" are different, calling "gmake" at top
3547 level broke things. I've arranged for the value of $(MAKE) to be passed
3548 from the Makefile to this script so that it can call the same version of
3552 A note about Exim versions 4.44 and 4.50
3553 ----------------------------------------
3555 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3556 changes of various kinds. As a consequence, a big documentation update was
3557 needed. This delayed the release for rather longer than seemed good, especially
3558 in the light of a couple of (minor) security issues. Therefore, the changes
3559 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3560 release. So 4.44 and 4.50 are in effect two different branches that both start
3563 I have left the 4.50 change log unchanged; it contains all the changes since
3564 4.43. The change log for 4.44 is below; many of its items are identical to
3565 those for 4.50. This seems to be the most sensible way to preserve the
3566 historical information.
3572 1. Minor wording change to the doc/README.SIEVE file.
3574 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3575 computation of the current number of files was incorrect.
3577 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3578 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3581 4. Give more explanation in the error message when the command for a transport
3582 filter fails to execute.
3584 5. There are several places where Exim runs a non-Exim command in a
3585 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3586 was being done only for the command run by the queryprogram router. It is
3587 now done for all such subprocesses. The other cases are: ${run, transport
3588 filters, and the commands run by the lmtp and pipe transports.
3590 6. Added CONFIGURE_GROUP build-time option.
3592 7. Some older OS have a limit of 256 on the maximum number of file
3593 descriptors. Exim was using setrlimit() to set 1000 as a large value
3594 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3595 systems. I've change it so that if it can't get 1000, it tries for 256.
3597 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3598 was an oversight, and furthermore, ever since the addition of extra
3599 controls (e.g. 4.43/32), the checks on when to allow different forms of
3600 "control" were broken. There should now be diagnostics for all cases when a
3601 control that does not make sense is encountered.
3603 9. Added the /retain_sender option to "control=submission".
3605 10. $recipients is now available in the predata ACL (oversight).
3607 11. Tidy the search cache before the fork to do a delivery from a message
3608 received from the command line. Otherwise the child will trigger a lookup
3609 failure and thereby defer the delivery if it tries to use (for example) a
3610 cached ldap connection that the parent has called unbind on.
3612 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3613 of $address_data from the recipient verification was clobbered by the
3614 sender verification.
3616 13. The value of address_data from a sender verification is now available in
3617 $sender_address_data in subsequent conditions in the ACL statement.
3619 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3621 15. Added a new option "connect=<time>" to callout options, to set a different
3624 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3625 was its contents. (It was OK if the option was not defined at all.)
3627 17. A "Completed" log line is now written for messages that are removed from
3628 the spool by the -Mrm option.
3630 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3631 information about exactly what failed.
3633 19. Added -dd to debug only the daemon process.
3635 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3636 handles timeouts, both on the server side and network timeouts. Renamed the
3637 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3639 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3640 It is now set to "smtps".
3642 22. $host_address is now set to the target address during the checking of
3643 ignore_target_hosts.
3645 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3646 being passed; this would have caused $sender_host_name to have been used if
3647 matching the list had actually called for a host name (not very likely,
3648 since this list is usually IP addresses). A host name is now passed as
3651 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3652 code that specifies a non-restarting handler (typically sigaction() in
3653 modern systems) in an attempt to fix a rare and obscure crash bug.
3655 25. Narrowed the window for a race in the daemon that could cause it to ignore
3656 SIGCHLD signals. This is not a major problem, because they are used only to
3657 wake it up if nothing else does.
3659 26. A malformed maildirsize file could cause Exim to calculate negative values
3660 for the mailbox size or file count. Odd effects could occur as a result.
3661 The maildirsize information is now recalculated if the size or filecount
3664 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3665 support for a long time. Removed HAVE_SYS_VFS_H.
3667 28. Installed the latest version of exipick from John Jetmore.
3669 29. In an address list, if the pattern was not a regular expression, an empty
3670 subject address (from a bounce message) matched only if the pattern was an
3671 empty string. Non-empty patterns were not even tested. This was the wrong
3672 because it is perfectly reasonable to use an empty address as part of a
3673 database query. An empty address is now tested by patterns that are
3674 lookups. However, all the other forms of pattern expect the subject to
3675 contain a local part and a domain, and therefore, for them, an empty
3676 address still always fails if the pattern is not itself empty.
3678 30. Exim went into a mad DNS loop when attempting to do a callout where the
3679 host was specified on an smtp transport, and looking it up yielded more
3680 than one IP address.
3682 31. Re-factored the code for checking spool and log partition space into a
3683 function that finds that data and another that does the check. The former
3684 is then used to implement four new variables: $spool_space, $log_space,
3685 $spool_inodes, and $log_inodes.
3687 32. The RFC2047 encoding function was originally intended for short strings
3688 such as real names; it was not keeping to the 75-character limit for
3689 encoded words that the RFC imposes. It now respects the limit, and
3690 generates multiple encoded words if necessary. To be on the safe side, I
3691 have increased the buffer size for the ${rfc2047: expansion operator from
3694 33. It is now permitted to omit both strings after an "if" condition; if the
3695 condition is true, the result is "true". As before, when the second string
3696 is omitted, a false condition yields an empty string. This makes it less
3697 cumbersome to write custom ACL and router conditions.
3699 34. Failure to deliver a bounce message always caused it to be frozen, even if
3700 there was an errors_to setting on the router. The errors_to setting is now
3703 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3704 canonical form (fully expanded) before being placed in
3705 $sender_host_address.
3707 36. The table in the code that translates DNS record types into text (T_A to
3708 "A" for instance) was missing entries for NS and CNAME. It is just possible
3709 that this could have caused confusion if both these types were looked up
3710 for the same domain, because the text type is used as part of Exim's
3711 per-process caching. But the chance of anyone hitting this buglet seems
3714 37. The dnsdb lookup has been extended in a number of ways.
3716 (1) There is a new type, "zns", which walks up the domain tree until it
3717 finds some nameserver records. It should be used with care.
3719 (2) There is a new type, "mxh", which is like "mx" except that it returns
3720 just the host names, not the priorities.
3722 (3) It is now possible to give a list of domains (or IP addresses) to be
3723 looked up. The behaviour when one of the lookups defers can be
3724 controlled by a keyword.
3726 (4) It is now possible to specify the separator character for use when
3727 multiple records are returned.
3729 38. The dnslists ACL condition has been extended: it is now possible to supply
3730 a list of IP addresses and/or domains to be looked up in a particular DNS
3733 39. Added log_selector=+queue_time_overall.
3735 40. When running the queue in the test harness, wait just a tad after forking a
3736 delivery process, to get repeatability of debugging output.
3738 41. Include certificate and key file names in error message when GnuTLS fails
3739 to set them up, because the GnuTLS error message doesn't include the name
3740 of the failing file when there is a problem reading it.
3742 42. Allow both -bf and -bF in the same test run.
3744 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3746 44. The "Exiscan patch" is now merged into the mainline Exim source.
3748 45. Sometimes the final signoff response after QUIT could fail to get
3749 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3750 < 0 before doing a fflush(). This bug looks as though it goes back to the
3751 introduction of TLS in release 3.20, but "sometimes" must have been rare
3752 because the tests only now provoked it.
3754 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3755 (this can affect the format of dates).
3757 47. exim_tidydb, when checking for the continued existence of a message for
3758 which it has found a message-specific retry record, was not finding
3759 messages that were in split spool directories. Consequently, it was
3760 deleting retry records that should have stayed in existence.
3762 48. Steve fixed some bugs in eximstats.
3764 49. The SPA authentication driver was not abandoning authentication and moving
3765 on to the next authenticator when an expansion was forced to fail,
3766 contradicting the general specification for all authenticators. Instead it
3767 was generating a temporary error. It now behaves as specified.
3769 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3770 (the order specifies the preference for clients). The order is now AES256,
3771 AES128, 3DES, ARCFOUR128.
3773 51. Small patch to Sieve code - explicitly set From: when generating an
3776 52. Exim crashed if a remote delivery caused a very long error message to be
3777 recorded - for instance if somebody sent an entire SpamAssassin report back
3778 as a large number of 550 error lines. This bug was coincidentally fixed by
3779 increasing the size of one of Exim's internal buffers (big_buffer) that
3780 happened as part of the Exiscan merge. However, to be on the safe side, I
3781 have made the code more robust (and fixed the comments that describe what
3784 53. Now that there can be additional text after "Completed" in log lines (if
3785 the queue_time_overall log selector is set), a one-byte patch to exigrep
3786 was needed to allow it to recognize "Completed" as not the last thing in
3789 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3790 patch that reportedly fixes this has been added. I am not expert enough to
3791 create a test for it. This is what the patch creator wrote:
3793 "I found a little strange behaviour of ldap code when working with
3794 Windows 2003 AD Domain, where users was placed in more than one
3795 Organization Units. When I tried to give exim partial DN, the exit code
3796 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3797 But simultaneously result of request was absolutely normal ldap result,
3798 so I produce this patch..."
3800 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3801 REFERENCE, so I have modified the code to exclude the patch when that macro
3804 55. Some experimental protocols are using DNS PTR records for new purposes. The
3805 keys for these records are domain names, not reversed IP addresses. The
3806 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3807 leaves it alone. Component reversal etc. now happens only for IP addresses.
3810 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3812 57. Double the size of the debug message buffer (to 2048) so that more of very
3813 long debug lines gets shown.
3815 58. The exicyclog utility now does better if the number of log files to keep
3816 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3818 59. Two changes related to the smtp_active_hostname option:
3820 (1) $smtp_active_hostname is now available as a variable.
3821 (2) The default for smtp_banner uses $smtp_active_hostname instead
3822 of $primary_hostname.
3824 60. The host_aton() function is supposed to be passed a string that is known
3825 to be a valid IP address. However, in the case of IPv6 addresses, it was
3826 not checking this. This is a hostage to fortune. Exim now panics and dies
3827 if the condition is not met. A case was found where this could be provoked
3828 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3829 components; fortuitously, this particular loophole had already been fixed
3830 by change 4.50/55 above.
3832 If there are any other similar loopholes, the new check in host_aton()
3833 itself should stop them being exploited. The report I received stated that
3834 data on the command line could provoke the exploit when Exim was running as
3835 exim, but did not say which command line option was involved. All I could
3836 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3837 running as the user.
3840 61. There was a buffer overflow vulnerability in the SPA authentication code
3841 (which came originally from the Samba project). I have added a test to the
3842 spa_base64_to_bits() function which I hope fixes it.
3845 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3846 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3848 63. The daemon start-up calls getloadavg() while still root for those OS that
3849 need the first call to be done as root, but it missed one case: when
3850 deliver_queue_load_max is set with deliver_drop_privilege. This is
3851 necessary for the benefit of the queue runner, because there is no re-exec
3852 when deliver_drop_privilege is set.
3854 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3855 This has been fixed.
3857 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3858 was in use, was not putting the data itself into the right store pool;
3859 consequently, it could be overwritten for a subsequent message in the same
3860 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3863 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3865 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3866 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3867 the function were treating the return as a boolean value, which happened to
3868 work because 0=false and not-0=true, but is not correct code.
3870 68. The host_aton() function was not handling scoped IPv6 addresses (those
3871 with, for example, "%eth0" on the end) correctly.
3873 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3874 negated items (that is, ~something) in unsigned ints. Some compilers
3875 apparently mutter when there is no cast.
3877 70. If an address verification called from an ACL failed, and did not produce a
3878 user-specific message (i.e. there was only a "system" message), nothing was
3879 put in $acl_verify_message. In this situation, it now puts the system
3882 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3883 session; change 4.31/43 added the unwanted input to the log line - except
3884 that it did not do this in the start of session case. It now does.
3886 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3887 This is wrong and can cause the other end to generate a synchronization
3888 error if it is another Exim or anything else that does the synchronization
3889 check. A QUIT command is no longer sent after a timeout.
3891 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3892 during host lookups.
3894 74. The defer_ok option of callout verification was not working if it was used
3895 when verifying addresses in header lines, that is, for this case:
3897 verify = header_sender/callout=defer_ok
3899 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3900 those file descriptors could be used for SMTP connections. If anything
3901 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3902 could be sent to the SMTP client, causing chaos. The daemon now opens
3903 stdin, stdout, and stderr to /dev/null when it puts itself into the
3906 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3907 log by default. The user can override this with suitable Perl magic.
3909 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3910 the log message when discard triggers, was not working for the DATA ACL or
3911 for the non-SMTP ACL.
3913 78. Error message wording change in sieve.c.
3915 79. If smtp_accept_max_per_host was set, the number of connections could be
3916 restricted to fewer than expected, because the daemon was trying to set up
3917 a new connection before checking whether the processes handling previous
3918 connections had finished. The check for completed processes is now done
3919 earlier. On busy systems, this bug wouldn't be noticed because something
3920 else would have woken the daemon, and it would have reaped the completed
3923 80. If a message was submitted locally by a user whose login name contained one
3924 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3925 It caused a spool format error. I have fixed the spool reading code. A
3926 related problem was that the "from" clause in the Received: line became
3927 illegal because of the space(s). It is now covered by ${quote_local_part.
3929 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3932 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3934 83. Patch from Sieve maintainer to fix unterminated string problem in
3935 "vacation" handling.
3937 84. Some minor changes to the Linux configuration files to help with other
3938 OS variants using glibc.
3940 85. One more patch for Sieve to update vacation handling to latest spec.
3943 ----------------------------------------------------
3944 See the note above about the 4.44 and 4.50 releases.
3945 ----------------------------------------------------
3951 1. Change 4.43/35 introduced a bug that caused file counts to be
3952 incorrectly computed when quota_filecount was set in an appendfile
3955 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3956 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3959 3. Give more explanation in the error message when the command for a transport
3960 filter fails to execute.
3962 4. There are several places where Exim runs a non-Exim command in a
3963 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3964 was being done only for the command run by the queryprogram router. It is
3965 now done for all such subprocesses. The other cases are: ${run, transport
3966 filters, and the commands run by the lmtp and pipe transports.
3968 5. Some older OS have a limit of 256 on the maximum number of file
3969 descriptors. Exim was using setrlimit() to set 1000 as a large value
3970 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3971 systems. I've change it so that if it can't get 1000, it tries for 256.
3973 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3974 was an oversight, and furthermore, ever since the addition of extra
3975 controls (e.g. 4.43/32), the checks on when to allow different forms of
3976 "control" were broken. There should now be diagnostics for all cases when a
3977 control that does not make sense is encountered.
3979 7. $recipients is now available in the predata ACL (oversight).
3981 8. Tidy the search cache before the fork to do a delivery from a message
3982 received from the command line. Otherwise the child will trigger a lookup
3983 failure and thereby defer the delivery if it tries to use (for example) a
3984 cached ldap connection that the parent has called unbind on.
3986 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3987 of $address_data from the recipient verification was clobbered by the
3988 sender verification.
3990 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3991 was its contents. (It was OK if the option was not defined at all.)
3993 11. A "Completed" log line is now written for messages that are removed from
3994 the spool by the -Mrm option.
3996 12. $host_address is now set to the target address during the checking of
3997 ignore_target_hosts.
3999 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4000 being passed; this would have caused $sender_host_name to have been used if
4001 matching the list had actually called for a host name (not very likely,
4002 since this list is usually IP addresses). A host name is now passed as
4005 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4006 code that specifies a non-restarting handler (typically sigaction() in
4007 modern systems) in an attempt to fix a rare and obscure crash bug.
4009 15. Narrowed the window for a race in the daemon that could cause it to ignore
4010 SIGCHLD signals. This is not a major problem, because they are used only to
4011 wake it up if nothing else does.
4013 16. A malformed maildirsize file could cause Exim to calculate negative values
4014 for the mailbox size or file count. Odd effects could occur as a result.
4015 The maildirsize information is now recalculated if the size or filecount
4018 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4019 support for a long time. Removed HAVE_SYS_VFS_H.
4021 18. Updated exipick to current release from John Jetmore.
4023 19. Allow an empty sender to be matched against a lookup in an address list.
4024 Previously the only cases considered were a regular expression, or an
4027 20. Exim went into a mad DNS lookup loop when doing a callout where the
4028 host was specified on the transport, if the DNS lookup yielded more than
4031 21. The RFC2047 encoding function was originally intended for short strings
4032 such as real names; it was not keeping to the 75-character limit for
4033 encoded words that the RFC imposes. It now respects the limit, and
4034 generates multiple encoded words if necessary. To be on the safe side, I
4035 have increased the buffer size for the ${rfc2047: expansion operator from
4038 22. Failure to deliver a bounce message always caused it to be frozen, even if
4039 there was an errors_to setting on the router. The errors_to setting is now
4042 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4043 canonical form (fully expanded) before being placed in
4044 $sender_host_address.
4046 24. Updated eximstats to version 1.33
4048 25. Include certificate and key file names in error message when GnuTLS fails
4049 to set them up, because the GnuTLS error message doesn't include the name
4050 of the failing file when there is a problem reading it.
4052 26. Expand error message when OpenSSL has problems setting up cert/key files.
4055 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4056 (this can affect the format of dates).
4058 28. exim_tidydb, when checking for the continued existence of a message for
4059 which it has found a message-specific retry record, was not finding
4060 messages that were in split spool directories. Consequently, it was
4061 deleting retry records that should have stayed in existence.
4063 29. eximstats updated to version 1.35
4064 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4065 1.35 - bugfix such that pie charts by volume are generated correctly
4067 30. The SPA authentication driver was not abandoning authentication and moving
4068 on to the next authenticator when an expansion was forced to fail,
4069 contradicting the general specification for all authenticators. Instead it
4070 was generating a temporary error. It now behaves as specified.
4072 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4073 (the order specifies the preference for clients). The order is now AES256,
4074 AES128, 3DES, ARCFOUR128.
4076 31. Small patch to Sieve code - explicitly set From: when generating an
4079 32. Exim crashed if a remote delivery caused a very long error message to be
4080 recorded - for instance if somebody sent an entire SpamAssassin report back
4081 as a large number of 550 error lines. This bug was coincidentally fixed by
4082 increasing the size of one of Exim's internal buffers (big_buffer) that
4083 happened as part of the Exiscan merge. However, to be on the safe side, I
4084 have made the code more robust (and fixed the comments that describe what
4087 33. Some experimental protocols are using DNS PTR records for new purposes. The
4088 keys for these records are domain names, not reversed IP addresses. The
4089 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4090 leaves it alone. Component reversal etc. now happens only for IP addresses.
4093 34. The host_aton() function is supposed to be passed a string that is known
4094 to be a valid IP address. However, in the case of IPv6 addresses, it was
4095 not checking this. This is a hostage to fortune. Exim now panics and dies
4096 if the condition is not met. A case was found where this could be provoked
4097 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4098 components; fortuitously, this particular loophole had already been fixed
4099 by change 4.50/55 or 4.44/33 above.
4101 If there are any other similar loopholes, the new check in host_aton()
4102 itself should stop them being exploited. The report I received stated that
4103 data on the command line could provoke the exploit when Exim was running as
4104 exim, but did not say which command line option was involved. All I could
4105 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4106 running as the user.
4109 35. There was a buffer overflow vulnerability in the SPA authentication code
4110 (which came originally from the Samba project). I have added a test to the
4111 spa_base64_to_bits() function which I hope fixes it.
4114 36. The daemon start-up calls getloadavg() while still root for those OS that
4115 need the first call to be done as root, but it missed one case: when
4116 deliver_queue_load_max is set with deliver_drop_privilege. This is
4117 necessary for the benefit of the queue runner, because there is no re-exec
4118 when deliver_drop_privilege is set.
4120 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4121 was in use, was not putting the data itself into the right store pool;
4122 consequently, it could be overwritten for a subsequent message in the same
4123 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4126 38. Sometimes the final signoff response after QUIT could fail to get
4127 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4128 < 0 before doing a fflush(). This bug looks as though it goes back to the
4129 introduction of TLS in release 3.20, but "sometimes" must have been rare
4130 because the tests only now provoked it.
4136 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4137 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4138 Now it returns an integer. A number of calls were still expecting a T/F
4139 return. Fortuitously, in all cases, the tests worked in OK situations,
4140 which is the norm. However, things would have gone wrong on any write
4141 failures on the smtp file descriptor. This function is used when sending
4142 messages over SMTP and also when doing verify callouts.
4144 2. When Exim is called to do synchronous delivery of a locally submitted
4145 message (the -odf or -odi options), it no longer closes stderr before doing
4148 3. Implemented the mua_wrapper option.
4150 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4152 5. Implemented the functions header_remove(), header_testname(),
4153 header_add_at_position(), and receive_remove_recipient(), and exported them
4156 6. If an ACL "warn" statement specified the addition of headers, Exim already
4157 inserted X-ACL-Warn: at the start if there was no header name. However, it
4158 was not making this test for the second and subsequent header lines if
4159 there were newlines in the string. This meant that an invalid header could
4160 be inserted if Exim was badly configured.
4162 7. Allow an ACL "warn" statement to add header lines at the start or after all
4163 the Received: headers, as well as at the end.
4165 8. Added the rcpt_4xx retry error code.
4167 9. Added postmaster_mailfrom=xxx to callout verification option.
4169 10. Added mailfrom=xxxx to the callout verification option, for verify=
4172 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4173 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4175 12. Inserted some casts to stop certain compilers warning when using pointer
4176 differences as field lengths or precisions in printf-type calls (mostly
4177 affecting debugging statements).
4179 13. Added optional readline() support for -be (dynamically loaded).
4181 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4182 same clock tick as a message's arrival, so that its received time was the
4183 same as the "first fail" time on the retry record, and that message
4184 remained on the queue past the ultimate address timeout, every queue runner
4185 would try a delivery (because it was past the ultimate address timeout) but
4186 after another failure, the ultimate address timeout, which should have then
4187 bounced the address, did not kick in. This was a "< instead of <=" error;
4188 in most cases the first failure would have been in the next clock tick
4189 after the received time, and all would be well.
4191 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4192 being recognized when the domain list was tested by the match_domain
4193 condition in an expansion string.
4195 16. Added the ${str2b64: operator.
4197 17. Exim was always calling setrlimit() to set a large limit for the number of
4198 processes, without checking whether the existing limit was already
4199 adequate. (It did check for the limit on file descriptors.) Furthermore,
4200 errors from getrlimit() and setrlimit() were being ignored. Now they are
4201 logged to the main and panic logs, but Exim does carry on, to try to do its
4202 job under whatever limits there are.
4204 18. Imported PCRE 5.0.
4206 19. Trivial typo in log message " temporarily refused connection" (the leading
4209 20. If the log selector return_path_on_delivery was set and an address was
4210 redirected to /dev/null, the delivery process crashed because it assumed
4211 that a return path would always be set for a "successful" delivery. In this
4212 case, the whole delivery is bypassed as an optimization, and therefore no
4215 21. Internal re-arrangement: the function for sending a challenge and reading
4216 a response while authentication was assuming a zero-terminated challenge
4217 string. It's now changed to take a pointer and a length, to allow for
4218 binary data in such strings.
4220 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4222 23. Exim was not respecting finduser_retries when seeking the login of the
4223 uid under which it was called; it was always trying 10 times. (The default
4224 setting of finduser_retries is zero.) Also, it was sleeping after the final
4225 failure, which is pointless.
4227 24. Implemented tls_on_connect_ports.
4229 25. Implemented acl_smtp_predata.
4231 26. If the domain in control=submission is set empty, Exim assumes that the
4232 authenticated id is a complete email address when it generates From: or
4233 Sender: header lines.
4235 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4236 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4237 chown and chgrp in /bin and hostname in /usr/bin.
4239 28. Exim was keeping the "process log" file open after each use, just as it
4240 does for the main log. This opens the possibility of it remaining open for
4241 long periods when the USR1 signal hits a daemon. Occasional processlog
4242 errors were reported, that could have been caused by this. Anyway, it seems
4243 much more sensible not to leave this file open at all, so that is what now
4246 29. The long-running daemon process does not normally write to the log once it
4247 has entered its main loop, and it closes the log before doing so. This is
4248 so that log files can straightforwardly be renamed and moved. However,
4249 there are a couple of unusual error situations where the daemon does write
4250 log entries, and I had neglected to close the log afterwards.
4252 30. The text of an SMTP error response that was received during a remote
4253 delivery was being truncated at 512 bytes. This is too short for some of
4254 the long messages that one sometimes sees. I've increased the limit to
4257 31. It is now possible to make retry rules that apply only when a message has a
4258 specific sender, in particular, an empty sender.
4260 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4261 it possible to be selective about when SMTP synchronization is enforced.
4263 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4265 32. Implemented hosts_connection_nolog.
4267 33. Added an ACL for QUIT.
4269 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4272 35. Added mailbox_size and mailbox_filecount to appendfile.
4274 36. Added control = no_multiline_responses to ACLs.
4276 37. There was a bug in the logic of the code that waits for the clock to tick
4277 in the case where the clock went backwards by a substantial amount such
4278 that the microsecond fraction of "now" was more than the microsecond
4279 fraction of "then" (but the whole seconds number was less).
4281 38. Added support for the libradius Radius client library this is found on
4282 FreeBSD (previously only the radiusclient library was supported).
4288 1. When certain lookups returned multiple values in the form name=value, the
4289 quoting of the values was not always being done properly. Specifically:
4290 (a) If the value started with a double quote, but contained no whitespace,
4292 (b) If the value contained whitespace other than a space character (i.e.
4293 tabs or newlines or carriage returns) it was not quoted.
4294 This fix has been applied to the mysql and pgsql lookups by writing a
4295 separate quoting function and calling it from the lookup code. The fix
4296 should probably also be applied to nisplus, ibase and oracle lookups, but
4297 since I cannot test any of those, I have not disturbed their existing code.
4299 2. A hit in the callout cache for a specific address caused a log line with no
4300 reason for rejecting RCPT. Now it says "Previous (cached) callout
4301 verification failure".
4303 3. There was an off-by-one bug in the queryprogram router. An over-long
4304 return line was truncated at 256 instead of 255 characters, thereby
4305 overflowing its buffer with the terminating zero. As well as fixing this, I
4306 have increased the buffer size to 1024 (and made a note to document this).
4308 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4309 when Exim is waiting for an SMTP response from a remote server, Exim
4310 restarts its select() call on the socket, thereby resetting its timeout.
4311 This is not a problem when such interrupts are rare. Somebody set up a cron
4312 job to run exiwhat every 2 minutes, which is less than the normal select()
4313 timeout (5 or 10 minutes). This meant that the select() timeout never
4314 kicked in because it was always reset. I have fixed this by comparing the
4315 time when an interrupt arrives with the time at the start of the first call
4316 to select(). If more time than the timeout has elapsed, the interrupt is
4317 treated as a timeout.
4319 5. Some internal re-factoring in preparation for the addition of Sieve
4320 extensions (by MH). In particular, the "personal" test is moved to a
4321 separate function, and given an option for scanning Cc: and Bcc: (which is
4322 not set for Exim filters).
4324 6. When Exim created an email address using the login of the caller as the
4325 local part (e.g. when creating a From: or Sender: header line), it was not
4326 quoting the local part when it contained special characters such as @.
4328 7. Installed new OpenBSD configuration files.
4330 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4331 try to make them clearer.
4333 9. Callout options, other than the timeout value, were being ignored when
4334 verifying sender addresses in header lines. For example, when using
4336 verify = header_sender/callout=no_cache
4338 the cache was (incorrectly) being used.
4340 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4341 only the Cygwin environment.
4343 11. When return_path_on_delivery was set as a log selector, if different remote
4344 addresses in the same message used different return paths and parallel
4345 remote delivery occurred, the wrong values would sometimes be logged.
4346 (Whenever a remote delivery process finished, the return path value from
4347 the most recently started remote delivery process was logged.)
4349 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4350 lines when AUTH and/or TLS are in use. This is the "received protocol"
4351 field. Exim used to use "asmtp" for authenticated SMTP, without any
4352 indication (in the protocol name) for TLS use. Now it follows the RFC and
4353 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4354 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4355 names appear in log lines as well as in Received: header lines.
4357 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4358 extensions, and comparison tests, and to fix some bugs.
4360 14. Changes to the "personal" filter test:
4362 (1) The test was buggy in that it was just doing the equivalent of
4363 "contains" tests on header lines. For example, if a user's address was
4364 anne@some.where, the "personal" test would incorrectly be true for
4366 To: susanne@some.where
4368 This test is now done by extracting each address from the header in turn,
4369 and checking the entire address. Other tests that are part of "personal"
4370 are now done using regular expressions (for example, to check local parts
4371 of addresses in From: header lines).
4373 (2) The list of non-personal local parts in From: addresses has been
4374 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4375 taken from the Sieve specification recommendations.
4377 (3) If the message contains any header line starting with "List-" it is
4378 treated as non-personal.
4380 (4) The test for "circular" in the Subject: header line has been removed
4381 because it now seems ill-conceived.
4383 15. Minor typos in src/EDITME comments corrected.
4385 16. Installed latest exipick from John Jetmore.
4387 17. If headers_add on a router specified a text string that was too long for
4388 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4389 of string_sprintf() is now avoided.
4391 18. $message_body_size was not set (it was always zero) when running the DATA
4392 ACL and the local_scan() function.
4394 19. For the "mail" command in an Exim filter, no default was being set for
4395 the once_repeat time, causing a random time value to be used if "once" was
4396 specified. (If the value happened to be <= 0, no repeat happened.) The
4397 default is now 0s, meaning "never repeat". The "vacation" command was OK
4398 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4399 (I found it when inspecting the code).
4401 20. There is now an overall timeout for performing a callout verification. It
4402 defaults to 4 times the callout timeout, which applies to individual SMTP
4403 commands during the callout. The overall timeout applies when there is more
4404 than one host that can be tried. The timeout is checked before trying the
4405 next host. This prevents very long delays if there are a large number of
4406 hosts and all are timing out (e.g. when the network connections are timing
4407 out). The value of the overall timeout can be changed by specifying an
4408 additional sub-option for "callout", called "maxwait". For example:
4410 verify = sender/callout=5s,maxwait=20s
4412 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4413 to the end before writing, but this should make it even safer).
4415 22. Exim was forgetting that it had advertised PIPELINING for the second and
4416 subsequent messages on an SMTP connection. It was also not resetting its
4417 memory on STARTTLS and an internal HELO.
4419 23. When Exim logs an SMTP synchronization error within a session, it now
4420 records whether PIPELINING has been advertised or not.
4422 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4423 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4424 rather than long int.
4426 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4428 26. Added the never_mail option to autoreply.
4434 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4435 crash if the getsockname() call failed; this can happen if a connection is
4436 closed very soon after it is established. The problem was simply in the
4437 order in which certain operations were done, causing Exim to try to write
4438 to the SMTP stream before it had set up the file descriptor. The bug has
4439 been fixed by making things happen in the correct order.
4445 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4446 before the connection was closed, thus losing the rejection response.
4448 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4449 some early Solaris releases, but causes trouble in current releases where
4450 socklen_t is defined.
4452 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4455 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4458 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4459 that releasing the top of it at the end releases what was used for sub-
4460 expansions (unless the block got too big). However, discard this block if
4461 the first thing is a variable or header, so that we can use its block when
4462 it is dynamic (useful for very large $message_headers, for example).
4464 6. Lookups now cache *every* query, not just the most recent. A new, separate
4465 store pool is used for this. It can be recovered when all lookup caches are
4466 flushed. Lookups now release memory at the end of their result strings.
4467 This has involved some general refactoring of the lookup sources.
4469 7. Some code has been added to the store_xxx() functions to reduce the amount
4470 of flapping under certain conditions.
4472 8. log_incoming_interface used to affect only the <= reception log lines. Now
4473 it causes the local interface and port to be added to several more SMTP log
4474 lines, for example "SMTP connection from", and rejection lines.
4476 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4478 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4480 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4481 could be overwritten at the end of the current message (or the start of a
4482 new message if it was set in a HELO ACL). The value is now preserved for
4483 the duration of the SMTP connection.
4485 12. If a transport had a headers_rewrite setting, and a matching header line
4486 contained an unqualified address, that address was qualified, even if it
4487 did not match any rewriting rules. The underlying bug was that the values
4488 of the flags that permit the existence of unqualified sender and recipient
4489 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4490 non-local messages, and by -bnq for local messages) were not being
4491 preserved with the message after it was received.
4493 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4494 "next input=" as part of the text comprising the host identity instead of
4495 the correct text. The code was using the same buffer for two different
4496 strings. However, depending on which order the printing function evaluated
4497 its arguments, the bug did not always show up. Under Linux, for example, my
4498 test suite worked just fine.
4500 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4501 doesn't work with some older versions of Perl. It has been changed to "my",
4502 which in any case is probably the better facility to use.
4504 15. A really picky compiler found some instances of statements for creating
4505 error messages that either had too many or two few arguments for the format
4508 16. The size of the buffer for calls to the DNS resolver has been increased
4509 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4510 for addresses that have a lot of PTR records. This alleviates a problem; it
4511 does not fully solve it.
4513 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4514 buffer now truncates the list and logs the incident, which is the same
4515 action as happens when Exim is looking up a host name and its aliases.
4516 Previously in this situation something unpredictable would happen;
4517 sometimes it was "internal error: store_reset failed".
4519 18. If a server dropped the connection unexpectedly when an Exim client was
4520 using GnuTLS and trying to read a response, the client delivery process
4521 crashed while trying to generate an error log message.
4523 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4524 string, for example:
4526 warn message = H1: something\nH2: something
4528 the text was added as a single header line from Exim's point of view
4529 though it ended up OK in the delivered message. However, searching for the
4530 second and subsequent header lines using $h_h2: did not work. This has been
4531 fixed. Similarly, if a system filter added multiple headers in this way,
4532 the routers could not see them.
4534 20. Expanded the error message when iplsearch is called with an invalid key to
4535 suggest using net-iplsearch in a host list.
4537 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4538 ACLs are no longer actually imposed (and a message to that effect is
4541 22. If a "gecos" field in a passwd entry contained escaped characters, in
4542 particular, if it contained a \" sequence, Exim got it wrong when building
4543 a From: or a Sender: header from that name. A second bug also caused
4544 incorrect handling when an unquoted " was present following a character
4545 that needed quoting.
4547 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4548 was not being matched caselessly.
4550 24. Arranged for all hyphens in the exim.8 source to be escaped with
4553 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4554 properly. Recipient callout cache records were still being keyed to include
4555 the sender, even when use_sender was set false. This led to far more
4556 callouts that were necessary. The sender is no longer included in the key
4557 when use_sender is false.
4559 26. Added "control = submission" modifier to ACLs.
4561 27. Added the ${base62d: operator to decode base 62 numbers.
4563 28. dnsdb lookups can now access SRV records.
4565 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4566 the configuration file.
4568 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4569 (-v) mode. This makes the output for a verbose queue run more intelligible.
4571 31. Added a use_postmaster feature to recipient callouts.
4573 32. Added the $body_zerocount variable, containing the number of binary zero
4574 bytes in the message body.
4576 33. The time of last modification of the "new" subdirectory is now used as the
4577 "mailbox time last read" when there is a quota error for a maildir
4580 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4582 35. Added +ignore_unknown as a special item in host lists.
4584 36. Code for decoding IPv6 addresses in host lists is now included, even if
4585 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4586 address was recognized as an IP address, but was then not correctly decoded
4587 into binary, causing unexpected and incorrect effects when compared with
4594 1. Very minor rewording of debugging text in manualroute to say "list of
4595 hosts" instead of "hostlist".
4597 2. If verify=header_syntax was set, and a header line with an unqualified
4598 address (no domain) and a large number of spaces between the end of the
4599 name and the colon was received, the reception process suffered a buffer
4600 overflow, and (when I tested it) crashed. This was caused by some obsolete
4601 code that should have been removed. The fix is to remove it!
4603 3. When running in the test harness, delay a bit after writing a bounce
4604 message to get a bit more predictability in the log output.
4606 4. Added a call to search_tidyup() just before forking a reception process. In
4607 theory, someone could use a lookup in the expansion of smtp_accept_max_
4608 per_host which, without the tidyup, could leave open a database connection.
4610 5. Added the variables $recipient_data and $sender_data which get set from a
4611 lookup success in an ACL "recipients" or "senders" condition, or a router
4612 "senders" option, similar to $domain_data and $local_part_data.
4614 6. Moved the writing of debug_print from before to after the "senders" test
4617 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4618 problems for message scanning, either using a data ACL, or using
4619 local_scan() because the Received: header was not generated till after they
4620 were called (in order to set the time as the time of reception completion).
4621 I have revised the way this works. The header is now generated after the
4622 body is received, but before the ACL or local_scan() are called. After they
4623 are run, the timestamp in the header is updated.
4629 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4630 before starting a queue runner without re-exec. This happened only when
4631 deliver_drop_privilege was set or when the Exim user was set to root. The
4632 effect of the bug was that timeouts during subsequent deliveries caused
4633 crashes instead of being properly handled. The handler is now left at its
4634 default (and expected) setting.
4636 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4637 message, again when deliver_drop_privilege is set or Exim is run as root.
4638 The bug described in (1) was not present in this case, but the tidying up
4639 of the other signals was missing. I have made the two cases consistent.
4641 3. The ignore_target_hosts setting on a manualroute router was being ignored
4642 for hosts that were looked up using the /MX notation.
4644 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4647 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4648 operated on the sender address. After changing the $sender_address to <>
4649 for the sender address verify, Exim was re-instated it as the original
4650 (before rewriting) address, but remembering that it had rewritten it, so it
4651 wasn't rewriting it again. This bug also had the effect of breaking the
4652 sender address verification caching when the sender address was rewritten.
4654 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4655 This has been changed so that if the ip literal address matches
4656 ignore_target_hosts, the router declines.
4658 7. Added expansion conditions match_domain, match_address, and match_local_
4659 part (NOT match_host).
4661 8. The placeholder for the Received: header didn't have a length field set.
4663 9. Added code to Exim itself and to exim_lock to test for a specific race
4664 condition that could lead to file corruption when using MBX delivery. The
4665 issue is with the lockfile that is created in /tmp. If this file is removed
4666 after a process has opened it but before that process has acquired a lock,
4667 there is the potential for a second process to recreate the file and also
4668 acquire a lock. This could lead to two Exim processes writing to the file
4669 at the same time. The added code performs the same test as UW imapd; it
4670 checks after acquiring the lock that its file descriptor still refers to
4671 the same named file.
4673 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4674 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4677 11. Added the smtp_active_hostname option. If used, this will typically be made
4678 to depend on the incoming interface address. Because $interface_address is
4679 not set up until the daemon has forked a reception process, error responses
4680 that can happen earlier (such as "too many connections") no longer contain
4683 12. If an expansion in a condition on a "warn" statement fails because a lookup
4684 defers, the "warn" statement is abandoned, and the next ACL statement is
4685 processed. Previously this caused the whole ACL to be aborted.
4687 13. Added the iplsearch lookup type.
4689 14. Added ident_timeout as a log selector.
4691 15. Added tls_certificate_verified as a log selector.
4693 16. Added a global option tls_require_ciphers (compare the smtp transport
4694 option of the same name). This controls incoming TLS connections.
4696 17. I finally figured out how to make tls_require_ciphers do a similar thing
4697 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4698 before starting the TLS session.
4700 18. Tabs are now shown as \t in -bP output.
4702 19. If the log selector return_path_on_delivery was set, Exim crashed when
4703 bouncing a message because it had too many Received: header lines.
4705 20. If two routers both had headers_remove settings, and the first one included
4706 a superfluous trailing colon, the final name in the first list and the
4707 first name in the second list were incorrectly joined into one item (with a
4708 colon in the middle).
4714 1. Added -C and -D options to the exinext utility, mainly to make it easier
4715 to include in the automated testing, but these could be helpful when
4716 multiple configurations are in use.
4718 2. The exinext utility was not formatting the output nicely when there was
4719 an alternate port involved in the retry record key, nor when there was a
4720 message id as well (for retries that were specific to a specific message
4721 and a specific host). It was also confused by IPv6 addresses, because of
4722 the additional colons they contain. I have fixed the IPv4 problem, and
4723 patched it up to do a reasonable job for IPv6.
4725 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4726 delivery, the log line now contains "pipelined" if PIPELINING was used.
4728 4. An SMTP transport process used to panic and die if the bind() call to set
4729 an explicit outgoing interface failed. This has been changed; it is now
4730 treated in the same way as a connect() failure.
4732 5. A reference to $sender_host_name in the part of a conditional expansion
4733 that was being skipped was still causing a DNS lookup. This no longer
4736 6. The def: expansion condition was not recognizing references to header lines
4737 that used bh_ and bheader_.
4739 7. Added the _cache feature to named lists.
4741 8. The code for checking quota_filecount in the appendfile transport was
4742 allowing one more file than it should have been.
4744 9. For compatibility with Sendmail, the command line option
4752 and sets the incoming protocol and host name (for trusted callers). The
4753 host name and its colon can be omitted when only the protocol is to be set.
4754 Note the Exim already has two private options, -pd and -ps, that refer to
4755 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4756 "s", but I don't think that's a major issue.
4758 10. A number of refactoring changes to the code, none of which should affect
4761 (a) The number of logging options was getting close to filling up the
4762 32-bit word that was used as a bit map. I have split them into two classes:
4763 those that are passed in the argument to log_write(), and those that are
4764 only ever tested independently outside of that function. These are now in
4765 separate 32-bit words, so there is plenty of room for expansion again.
4766 There is no change in the user interface or the logging behaviour.
4768 (b) When building, for example, log lines, the code previously used a
4769 macro that called string_cat() twice, in order to add two strings. This is
4770 not really sufficiently general. Furthermore, there was one instance where
4771 it was actually wrong because one of the argument was used twice, and in
4772 one call a function was used. (As it happened, calling the function twice
4773 did not affect the overall behaviour.) The macro has been replaced by a
4774 function that can join an arbitrary number of extra strings onto a growing
4777 (c) The code for expansion conditions now uses a table and a binary chop
4778 instead of a serial search (which was left over from when there were very
4779 few conditions). Also, it now recognizes conditions like "pam" even when
4780 the relevant support is not compiled in: a suitably worded error message is
4781 given if an attempt is made to use such a condition.
4783 11. Added ${time_interval:xxxxx}.
4785 12. A bug was causing one of the ddress fields not to be passed back correctly
4786 from remote delivery subprocesses. The field in question was not being
4787 subsequently used, so this caused to problems in practice.
4789 13. Added new log selectors queue_time and deliver_time.
4791 14. Might have fixed a bug in maildirsizefile handling that threw up
4792 "unexpected character" debug warnings, and recalculated the data
4793 unnecessarily. In any case, I expanded the warning message to give more
4796 15. Added the message "Restricted characters in address" to the statements in
4797 the default ACL that block characters like @ and % in local parts.
4799 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4800 Three changes have been made:
4802 (a) There was a serious bug; a negative response to MAIL caused the whole
4803 recipient domain to be cached as invalid, thereby blocking all messages
4804 to all local parts at the same domain, from all senders. This bug has
4805 been fixed. The domain is no longer cached after a negative response to
4806 MAIL if the sender used is not empty.
4808 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4811 (c) A new callout option, "use_sender" has been added for people who want
4812 the modified behaviour.
4818 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4821 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4822 indeed breaks things for older releases.
4824 3. Added additional logging to the case where there is a problem reading data
4825 from a filter that is running in a subprocess using a pipe, in order to
4826 try to track down a specific problem.
4828 4. Testing facility fudge: when running in the test harness and attempting
4829 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4830 getting "No route to host". Convert this to a timeout.
4832 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4835 6. Some OS don't have socklen_t but use size_t instead. This affects the
4836 fifth argument of getsockopt() amongst other things. This is now
4837 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4838 can be set for individual OS. I have set it for SunOS5, OSF1, and
4839 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4840 some earlier ones do not.
4842 7. Change 4.30/15 was not doing the test caselessly.
4844 8. The standard form for an IPv6 address literal was being rejected by address
4845 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4846 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4847 this, as well as the form without the "IPv6" on the front (but only when
4848 address literals are enabled, of course).
4850 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4852 10. Exim crashed if a message with an empty sender address specified by -f
4853 encountered a router with an errors_to setting. This could be provoked only
4854 by a command such as
4858 where an empty string was supplied; "<>" did not hit this bug.
4860 11. Installed PCRE release 4.5.
4862 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4863 remained set. It is now erased.
4865 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4866 times from message ids (which are base 36 rather than the normal 62).
4868 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4869 were being counted as actual protocol errors, and logged if the log
4870 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4871 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4872 and DATA following a set of rejected RCPTs do not count as protocol errors.
4873 In other words, Exim assumes they were pipelined, though this may not
4874 actually be the case. Of course, in all cases the client gets an
4875 appropriate error code.
4877 15. If a lookup fails in an ACL condition, a message about the failure may
4878 be available; it is used if testing the ACL cannot continue, because most
4879 such messages specify what the cause of the deferral is. However, some
4880 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4881 that caused an old message to be retained and used if a later statement
4882 caused a defer, replacing the real cause of the deferral.
4884 16. If an IP address had so many PTR records that the DNS lookup buffer
4885 was not large enough to hold them, Exim could crash while trying to process
4886 the truncated data. It now detects and logs this case.
4888 17. Further to 4.21/58, another change has been made: if (and only if) the
4889 first line of a message (the first header line) ends with CRLF, a bare LF
4890 in a subsequent header line has a space inserted after it, so as not to
4891 terminate the header.
4893 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4894 unnecessarily, used atoi() instead of strtol(), and didn't check the
4895 termination when getting file sizes from file names by regex.
4897 19. Completely re-implemented the support for maildirsize files, in the light
4898 of a number of problems with the previous contributed implementation
4899 (4.30/29). In particular:
4901 . If the quota is zero, the maildirsize file is maintained, but no quota is
4904 . If the maildir directory does not exist, it is created before any attempt
4905 to write a maildirsize file.
4907 . The quota value in the file is just a cache; if the quota is changed in
4908 the transport, the new value overrides.
4910 . A regular expression is available for excluding directories from the
4913 20. The autoreply transport checks the characters in options that define the
4914 message's headers; it allows continued headers, but it was checking with
4915 isspace() after an embedded newline instead of explicitly looking for a
4918 21. If all the "regular" hosts to which an address was routed had passed their
4919 expiry times, and had not reached their retry times, the address was
4920 bounced, even if fallback hosts were defined. Now Exim should go on to try
4923 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4924 equivalent code in the SMTP transport. Some hosts send humungous responses
4925 to HELO/EHLO, more than 1024 it seems.
4927 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4928 gives compiler warnings in some environments. I've now done it "properly",
4931 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4932 (because of problems with the built-in one) was declared to return uschar *
4933 instead of char *, causing compiler failure.
4935 25. Fixed a file descriptor leak when processing alias/forward files.
4937 26. Fixed a minor format string issue in dbfn.c.
4939 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4941 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4942 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4945 29. When a sender address is verified, it is cached, to save repeating the test
4946 when there is more than one recipient in a message. However, when the
4947 verification involves a callout, it is possible for different callout
4948 options to be set for different recipients. It is too complicated to keep
4949 track of this in the cache, so now Exim always runs a verification when a
4950 callout is required, relying on the callout cache for the optimization.
4951 The overhead is duplication of the address routing, but this should not be
4954 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4955 to be verified with callout=postmaster, and the main callout worked but the
4956 postmaster check failed, the verification correctly failed. However, if a
4957 subsequent RCPT command asked for sender verification *without* the
4958 postmaster check, incorrect caching caused this verification also to fail,
4961 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4962 it was not caching the DNS options (qualify_single, search_parents) that
4963 were used when the lookup failed. A subsequent lookup with different
4964 options therefore always gave the same answer, though there were cases
4965 where it should not have. (Example: a "domains = !$mx_any" option on a
4966 dnslookup router: the "domains" option is always processed without any
4967 widening, but the router might have qualify_single set.) Now Exim uses the
4968 cached value only when the same options are set.
4970 32. Added John Jetmore's "exipick" utility to the distribution.
4972 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4973 than a timeout (e.g. a certificate is required, and is not provided), an
4974 Exim server now closes the connection immediately. Previously it waited for
4975 the client to close - but if the client is SSL, it seems that they each
4976 wait for each other, leading to a delay before one of them times out.
4978 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4979 maintained 0.8.x compatibility because I don't think many are using it, and
4980 it is clearly obsolete.
4982 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4985 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4986 client certificate was expired. A simple patch fixes this, though I don't
4987 understand the full logic of why the verify callback is called multiple
4990 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4991 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4992 which causes problems with some clients (such as the Certicom SSL Plus
4993 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4994 disables the coutermeasure allowing Eudora to connect."
4996 38. Exim was not checking that a write() to a log file succeeded. This could
4997 lead to Bad Things if a log got too big, in particular if it hit a file
4998 size limit. Exim now panics and dies if it cannot write to a log file, just
4999 as it does if it cannot open a log file.
5001 39. Modified OS/Makefile-Linux so that it now contains
5003 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5005 The two -D definitions ensure that Exim is compiled with large file
5006 support, which makes it possible to handle log files that are bigger than
5009 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5010 instance) a domain was checked against a named list that involved a lookup,
5011 causing $domain_data to be set, then another domain was checked against the
5012 same list, then the first domain was re-checked, the value of $domain_data
5013 after the final check could be wrong. In particular, if the second check
5014 failed, it could be set empty. This bug probably also applied to
5017 41. The strip_trailing_dot option was not being applied to the address given
5018 with the -f command-line option.
5020 42. The code for reading a message's header from the spool was incrementing
5021 $received_count, but never initializing it. This meant that the value was
5022 incorrect (doubled) while delivering a message in the same process in which
5023 it was received. In the most common configuration of Exim, this never
5024 happens - a fresh exec is done - but it can happen when
5025 deliver_drop_privilege is set.
5027 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5028 it now includes up to 150 characters of the unexpected data in the log
5031 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5032 and building data strings. The size of both of these buffers was 10 000
5033 bytes - far larger than anybody would *ever* want, thought I. Needless to
5034 say, somebody hit the limit. I have increased the maximum line length to
5035 20 000 and the maximum data length of concatenated lines to 100 000. I have
5036 also fixed two bugs, because there was no checking on these buffers. Tsk,
5037 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5038 buffer is too small.
5040 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5041 lsearch lookups. Now it does.
5043 46. When parsing a route_list item in a manualroute router, a fixed-length
5044 buffer was used for the list of hosts. I made this 1024 bytes long,
5045 thinking that nobody would ever have a list of hosts that long. Wrong.
5046 Somebody had a whole pile of complicated expansion conditions, and the
5047 string was silently truncated, leading to an expansion error. It turns out
5048 that it is easier to change to an unlimited length (owing to other changes
5049 that have happened since this code was originally written) than to build
5050 structure for giving a limitation error. The length of the item that
5051 expands into the list of hosts is now unlimited.
5053 47. The lsearch lookup could not handle data where the length of text line was
5054 more than 4095 characters. Such lines were truncated, leading to shortened
5055 data being returned. It should now handle lines of any length.
5057 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5058 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5061 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5062 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5063 no longer happens in comment lines. A list of replacements is now placed
5064 at the head of all of the source files, except those whose only change is
5065 to replace PERL_COMMAND in the very first #! line.
5067 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5068 messages on the queue, with a bottom-up merge sort, using code contributed
5069 by Michael Haardt. This should make operations like -bp somewhat faster on
5070 large queues. It won't affect queue runners, except when queue_run_in_order
5073 51. Installed eximstats 1.31 in the distribution.
5075 52. Added support for SRV lookups to the dnslookup router.
5077 53. If an ACL referred to $message_body or $message_body_end, the value was not
5078 reset for any messages that followed in the same SMTP session.
5080 54. The store-handling optimization for building very long strings was not
5081 differentiating between the different store pools. I don't think this
5082 actually made any difference in practice, but I've tidied it.
5084 55. While running the routers to verify a sender address, $sender_address
5085 was still set to the sender address. This is wrong, because when routing to
5086 send a bounce to the sender, it would be empty. Therefore, I have changed
5087 it so that, while verifying a sender address, $sender_address is set to <>.
5088 (There is no change to what happens when verifying a recipient address.)
5090 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5091 target A or AAAA records (if not already returned) without resetting the
5092 qualify_single or search_parents options of the DNS resolver. These are
5093 inappropriate in this case because the targets of MX and SRV records must
5094 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5095 target that, when qualified, matched something in the local domain. These
5096 two options are now turned off when doing these lookups.
5098 57. It seems that at least some releases of Reiserfs (which does not have the
5099 concept of a fixed number of inodes) returns zero and not -1 for the
5100 number of available inodes. This interacted badly with check_spool_inodes,
5101 which assumed that -1 was the "no such thing" setting. What I have done is
5102 to check that the total number of inodes is greater than zero before doing
5103 the test of how many are available.
5105 58. When a "warn" ACL statement has a log_message modifier, the message is
5106 remembered, and not repeated. This is to avoid a lot of repetition when a
5107 message has many recipients that cause the same warning to be written.
5108 Howewer, Exim was preserving the list of already written lines for an
5109 entire SMTP session, which doesn't seem right. The memory is now reset if a
5110 new message is started.
5112 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5113 debugging output unless log_rewrite was also set.
5115 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5116 of (int)(handle) when we know that handle contains (void *)(-1).
5118 61. The Exim daemon panic-logs an error return when it closes the incoming
5119 connection. However "connection reset by peer" seems to be common, and
5120 isn't really an error worthy of noting specially, so that particular error
5123 62. When Exim is trying to find all the local interfaces, it used to panic and
5124 die if the ioctl to get the interface flags failed. However, it seems that
5125 on at least one OS (Solaris 9) it is possible to have an interface that is
5126 included in the list of interfaces, but for which you get a failure error
5127 for this call. This happens when the interface is not "plumbed" into a
5128 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5129 failure of the "get flags" call assumes that the interface is down.
5131 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5132 makes life easier for people who are doing arithmetic on fields extracted
5133 from dates, where you often get leading zeros that should not be
5134 interpreted as octal.
5136 64. Added qualify_domain to the redirect router, to override the global
5139 65. If a pathologically long header line contained very many addresses (the
5140 report of this problem mentioned 10 000) and each of them was rewritten,
5141 Exim could use up a very large amount of memory. (It kept on making new
5142 copies of the header line as it rewrote, and never released the old ones.)
5143 At the expense of a bit more processing, the header rewriting function has
5144 been changed so that it no longer eats memory in this way.
5146 66. The generation of the Received: header has been moved from the time that a
5147 message starts to be received, to the time that it finishes. The timestamp
5148 in the Received: header should now be very close to that of the <= log
5149 line. There are two side-effects of this change:
5151 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5152 logged header lines no longer include the local Received: line, because
5153 it has not yet been created. The same applies to a copy of the message
5154 that is returned to a non-SMTP sender when a message is rejected.
5156 (b) When a filter file is tested using -bf, no additional Received: header
5157 is added to the test message. After some thought, I decided that this
5160 This change does not affect the value of $received_for. It is still set
5161 after address rewriting, but before local_scan() is called.
5163 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5165 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5166 gave an unhelpful panic error message, and a defer error. I have managed to
5167 change this behaviour so that it now rejects any supplied certificate,
5168 which seems right, as the list of acceptable certificates is empty.
5170 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5171 gave an unhelpful defer error. I have not managed to make this reject any
5172 supplied certificates, but the error message it gives is "no certificate
5173 supplied", which is not helpful.
5175 70. exigrep's output now also includes lines that are not associated with any
5176 message, but which match the given pattern. Implemented by a patch from
5177 Martin Sluka, which also tidied up the Perl a bit.
5179 71. Recipient callout verification, like sender verification, was using <> in
5180 the MAIL FROM command. This isn't really the right thing, since the actual
5181 sender may affect whether the remote host accepts the recipient or not. I
5182 have changed it to use the actual sender in the callout; this means that
5183 the cache record is now keyed on a recipient/sender pair, not just the
5184 recipient address. There doesn't seem to be a real danger of callout loops,
5185 since a callout by the remote host to check the sender would use <>.
5186 [SEE ABOVE: changed after hitting problems.]
5188 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5189 temporary errors. However, in the case of such a code being given after
5190 the end of a data transmission (i.e. after ".") Exim was failing to write
5191 a retry record for the message. (Yes, there was some broken host that was
5192 actually sending 8xx at this point.)
5194 73. An unknown lookup type in a host list could cause Exim to panic-die when
5195 the list was checked. (An example that provoked this was putting <; in the
5196 middle of a list instead of at the start.) If this happened during a DATA
5197 ACL check, a -D file could be left lying around. This kind of configuration
5198 error no longer causes Exim to die; instead it causes a defer errror. The
5199 incident is still logged to the main and panic logs.
5201 74. Buglet left over from Exim 3 conversion. The message "too many messages
5202 in one connection" was written to the rejectlog but not the mainlog, except
5203 when address rewriting (yes!) was being logged.
5205 75. Added write_rejectlog option.
5207 76. When a system filter was run not as root (that is, when system_filter_user
5208 was set), the values of the $n variables were not being returned to the
5209 main process; thus, they were not subsequently available in the $sn
5212 77. Added +return_path_on_delivery log selector.
5214 78. A connection timeout was being treated differently from recipients deferred
5215 when testing hosts_max_try with a message that was older than the host's
5216 retry timeout. (The host should not be counted, thus allowing all hosts to
5217 be tried at least once before bouncing.) This may have been the cause of an
5218 occasionally reported bug whereby a message would remain on the queue
5219 longer than the retry timeout, but would be bounced if a delivery was
5220 forced. I say "may" because I never totally pinned down the problem;
5221 setting up timeout/retry tests is difficult. See also the next item.
5223 79. The ultimate address timeout was not being applied to errors that involved
5224 a combination of host plus message (for example, a timeout on a MAIL
5225 command). When an address resolved to a number of possible hosts, and they
5226 were not all tried for each delivery (e.g. because of hosts_max_try), a
5227 message could remain on the queue longer than the retry timeout.
5229 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5232 81. Fixed an obscure SMTP outgoing bug which required at least the following
5233 conditions: (a) there was another message waiting for the same server;
5234 (b) the server returned 5xx to all RCPT commands in the first message so
5235 that the message was not completed; (c) the server dropped the connection
5236 or gave a negative response to the RSET that Exim sends to abort the
5237 transaction. The observed case was a dropped connection after DATA that had
5238 been sent in pipelining mode. That is, the server had advertised PIPELINING
5239 but was not implementing it correctly. The effect of the bug was incorrect
5240 behaviour, such as trying another host, and this could lead to a crash.
5246 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5247 and daemon.c were passed as pointers to ints; they should have been
5248 pointers to socklen_t variables (which are typically unsigned ints).
5250 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5253 3. Fixed a really odd bug that affected only the testing scheme; patching a
5254 certain fixed string in the binary changed the value of another string that
5255 happened to be identical to the end of the original first string.
5257 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5258 name", it returns that address as the IP address. On some operating
5259 systems (e.g. Solaris), it also passes back the IP address string as the
5260 "host name". However, on others (e.g. Linux), it passes back an empty
5261 string. Exim wasn't checking for this, and was changing the host name to an
5262 empty string, assuming it had been canonicized.
5264 5. Although rare, it is permitted to have more than one PTR record for a given
5265 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5266 all the names associated with an address, because they do in Solaris.
5267 However, it seems that they do not in Linux for data that comes from the
5268 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5269 I found this out when I moved to a new Linux workstation and tried to run
5270 the Exim test suite.
5272 To get round this problem I have changed the code so that it now does its
5273 own call to the DNS to look up PTR records when searching for a host name.
5274 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5275 addresses that are only in /etc/hosts are still found.
5277 This behaviour is, however, controlled by an option called host_lookup_
5278 order, which defaults to "bydns:byaddr". If people want to use the other
5279 order, or indeed, just use one or the other means of lookup, they can
5280 specify it in this variable.
5282 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5283 some operating systems, this comes back from gethostbyaddr() as an empty
5284 string, and this is what Exim used to test for. However, it seems that in
5285 other systems, "." is yielded. Exim now tests for this case too.
5287 7. The values of check_spool_space and check_log_space are now held internally
5288 as a number of kilobytes instead of an absolute number of bytes. If a
5289 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5290 kilobyte. This means that much larger values can be stored.
5292 8. Exim monitor: an attempt to get the action menu when not actually pointing
5293 at a message produces an empty menu entitled "No message selected". This
5294 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5295 no entries in it ("Shell widget menu has zero width and/or height"). So I
5296 have added a single, blank menu entry in this case.
5298 9. Added ${quote_local_part.
5300 10. MIME decoding is now applied to the contents of Subject: header lines when
5303 11. Now that a reference to $sender_host_address automatically causes a reverse
5304 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5305 host lookup before query-style lookups in lists that might use this
5306 variable. This has therefore been abolished, and the "net-" prefix is no
5307 longer necessary for query-style lookups.
5309 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5310 have been a typo for LFLAGS, so it has been changed.
5312 13. The install script calls Exim with "-C /dev/null" in order to find the
5313 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5314 to be output. Howeve, since Exim outputs its version number before the
5315 error, it didn't break the script. It just looked ugly. I fixed this by
5316 always allowing "-C /dev/null" if the caller is root.
5318 14. Ignore overlarge ACL variable number when reading spool file - insurance
5319 against a later release with more variables having written the file.
5321 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5322 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5323 this, as well as the form without the "IPv6" on the front.
5325 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5326 OS/Makefile-Darwin file.
5328 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5329 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5331 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5332 variable. However, some OS use ints for inodes. I've added cast to long int
5333 to get rid of the compiler warning.
5335 19. I had forgotten to lock out "/../" in configuration file names when
5336 ALT_CONFIG_PREFIX was set.
5338 20. Routers used for verification do not need to specify transports. However,
5339 if such a router generated a host list, and callout was configured, Exim
5340 crashed, because it could not find a port number from the (non-existent)
5341 transport. It now assumes port 25 in this circumstance.
5343 21. Added the -t option to exigrep.
5345 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5346 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5347 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5348 them is set, LOOKUP_LSEARCH is forced.
5350 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5352 24. Added sender and host information to the "rejected by local_scan()" log
5353 line; previously there was no indication of these.
5355 25. Added .include_if_exists.
5357 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5358 Linux. It turns out that not all file systems support this. Apparently some
5359 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5360 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5361 not supported on the file descriptor, is now ignored when Exim is trying to
5362 sync a directory. This applies only to Linux.
5364 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5366 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5367 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5370 29. Added support for maildirsize files from supplied patch (modified a bit).
5372 30. The use of :fail: followed by an empty string could lead Exim to respond to
5373 sender verification failures with (e.g.):
5375 550 Verification failed for <xxx>
5376 550 Sender verify failed
5378 where the first response line was missing the '-' that indicates it is not
5379 the final line of the response.
5381 31. The loop for finding the name of the user that called Exim had a hardwired
5382 limit of 10; it now uses the value of finduser_retries, which is used for
5383 all other user lookups.
5385 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5388 33. Exim was neglecting to zero errno before one call of strtol() when
5389 expanding a string and expecting an integer value. On some systems this
5390 resulted in spurious "integer overflow" errors. Also, it was casting the
5391 result into an int without checking.
5393 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5394 did not work. The code looks as if it has *never* worked, though it appears
5395 to have been documented since at least releast 1.62. I have made it work.
5397 35. The "timeout_DNS" error in retry rules, also documented since at least
5398 1.62, also never worked. As it isn't clear exactly what this means, and
5399 clearly it isn't a major issue, I have abolished the feature by treating it
5400 as "timeout", and writing a warning to the main and panic logs.
5402 36. The display of retry rules for -brt wasn't always showing the error code
5405 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5406 timeout_connect_A, timeout_connect_MX.
5408 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5409 to the empty sender.
5411 39. The daemon was not analysing the content of -oX till after it had closed
5412 stderr and disconnected from the controlling terminal. This meant that any
5413 syntax errors were only noted on the panic log, and the return code from
5414 the command was 0. By re-arranging the code a little, I've made the
5415 decoding happen first, so such errors now appear on stderr, and the return
5416 code is 1. However, the actual setting up of the sockets still happens in
5417 the disconnected process, so errors there are still only recorded on the
5420 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5421 connections (as happens on some IP stacks) was logged at start up time as
5422 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5423 it from "IPv6 and IPv4", which means that two separate sockets are being
5426 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5427 says whether AF_INET or AF_INET6 was passed as an argument.
5429 42. Exiwhat output was messed up when time zones were included in log
5432 43. Exiwhat now gives more information about the daemon's listening ports,
5433 and whether -tls-on-connect was used.
5435 44. The "port" option of the smtp transport is now expanded.
5437 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5438 silently ignored. Now an error message is written to the main and panic
5441 46. There's a new ACL modifier called "logwrite" which writes to a log file
5442 as soon as it is encountered.
5444 47. Added $local_user_uid and $local_user_gid at routing time.
5446 48. Exim crashed when trying to verify a sender address that was being
5449 49. Exim was recognizing only a space character after ".include". It now also
5450 recognizes a tab character.
5452 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5453 extracting the relevant information from the specification. The man page no
5454 longer contains scrambled data for the -d option, and I've added a section
5455 at the front about calling Exim under different names.
5457 51. Added "extra_headers" argument to the "mail" command in filter files.
5459 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5462 53. Installed eximstats 1.29.
5464 54. Added transport_filter_timeout as a generic transport option.
5466 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5467 Cc: header lines. This was required by RFC 822, but it not required by RFC
5470 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5471 incoming messages that did not have them. Now it does so only if the
5472 message originates locally, that is, if there is no associated remote host
5473 address. When Resent- header lines are present, this applies to the Resent-
5474 lines rather than the non-Resent- lines.
5476 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5477 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5479 58. Messages for configuration errors now include the name of the main
5480 configuration file - useful now that there may be more than one file in a
5481 list (.included file names were always shown).
5483 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5484 for those rare installations that do not start the daemon as root or run it
5485 setuid root. I've cut out the call to initgroups() if the daemon is not
5488 60. The Exim user and group can now be bound into the binary as text strings
5489 that are looked up at the start of Exim's processing.
5491 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5493 62. Added $mailstore_basename variable.
5495 63. Installed patch to sieve.c from Michael Haardt.
5497 64. When Exim failed to open the panic log after failing to open the main log,
5498 the original message it was trying to log was written to stderr and debug
5499 output, but if they were not available (the usual case in production), it
5500 was lost. Now it is written to syslog before the two lines that record the
5501 failures to open the logs.
5503 65. Users' Exim filters run in subprocesses under the user's uid. It is
5504 possible for a "deliver" command or an alias in a "personal" command to
5505 provoke an address rewrite. If logging of address rewriting is configured,
5506 this fails because the process is not running as root or exim. There may be
5507 a better way of dealing with this, but for the moment (because 4.30 needs
5508 to be released), I have disabled address rewrite logging when running a
5509 filter in a non-root, non-exim process.
5515 1. The buildconfig auxiliary program wasn't quoting the value set for
5516 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5517 not defined. This bug was masked in 4.22 by the effect that was fixed in
5520 2. Some messages that were rejected after a message id was allocated were
5521 shown as "incomplete" by exigrep. It no longer does this for messages that
5522 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5524 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5525 have allow_domain_literals set, the ID did not get logged in the <= line.
5526 Domain literals are now always recognized in Message-ID: header lines.
5528 4. The first argument for a ${extract expansion item is the key name or field
5529 number. Leading and trailing spaces in this item were not being ignored,
5530 causing some misleading effects.
5532 5. When deliver_drop_privilege was set, single queue runner processes started
5533 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5534 same command in the process it spins off) were not dropping privilege.
5536 6. When the daemon running as "exim" started a queue runner, it always
5537 re-executed Exim in the spun-off process. This is a waste of effort when
5538 deliver_drop_privilege is set. The new process now just calls the
5539 queue-runner function directly.
5545 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5548 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5549 ignored. Though the use of -oP was forcing the writing of a pid file, it
5550 was always written to the default place.
5552 3. If the message "no IP address found for host xxxx" is generated during
5553 incoming verification, it is now followed by identification of the incoming
5554 connection (so you can more easily find what provoked it).
5556 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5558 5. Added some features to "harden" Exim a bit more against certain attacks:
5560 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5561 be put in Local/Makefile. This is like the never_users runtime option,
5562 but it cannot be overridden. The default setting is "root".
5564 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5565 prefix string with which any file named in a -C command line option
5568 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5569 is retained for -C and -D only if the caller of Exim is root. Without
5570 it, the exim user may also use -C and -D and retain privilege.
5572 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5573 command line option is disabled.
5575 6. Macro names set by the -D option must start with an upper case letter, just
5576 like macro names defined in the configuration file.
5578 7. Added "dereference=" facility to LDAP.
5580 8. Two instances of the typo "uknown" in the source files are fixed.
5582 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5583 the Configure-Makefile script screwed up while processing it.
5585 10. Incorporated PCRE 4.4.
5587 11. The SMTP synchronization check was not operating right at the start of an
5588 SMTP session. For example, it could not catch a HELO sent before the client
5589 waited for the greeting. There is now a check for outstanding input at the
5590 point when the greeting is written. Because of the duplex, asynchronous
5591 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5592 way, but not yet received, when the check is performed.
5594 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5595 on TCP/IP sockets, because this apparently causes some broken clients to
5598 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5599 unchanged) from the Cygwin maintainer.
5601 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5602 was supported instead of testing for mbx. Effectively a typo.
5604 15. The spa authenticator server code was not checking that the input it
5605 received was valid base64.
5607 16. The debug output line for the "set" modifier in ACLs was not showing the
5608 name of the variable that was being set.
5610 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5612 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5613 lookup on its own. Something else was needed to trigger the lookup. For
5614 example, a match in host_lookup or the need for a host name in a host list.
5615 Now, if $sender_host_name is referenced and the host name has not yet been
5616 looked up, a lookup is performed. If the lookup fails, the variable remains
5617 empty, and $host_lookup_failed is set to "1".
5619 19. Added "eqi" as a case-independent comparison operator.
5621 20. The saslauthd authentication condition could segfault if neither service
5622 nor realm was specified.
5624 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5625 error message that was logged was misleading, and incoming SMTP
5626 connections were dropped. The message is now more accurate, and temporary
5627 errors are given to SMTP connections.
5629 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5630 (except RSET) with a 421 error, until QUIT is received. However, it was
5631 failing to send a response to QUIT.
5633 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5634 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5635 ACL is now run after the helo_try_verify_hosts code.
5637 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5638 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5639 case-independent, so other case variants are also recognized). Apparently
5640 some systems use these upper case variants.
5642 25. If more than two messages were waiting for the same host, and a transport
5643 filter was specified for the transport, Exim sent two messages over the
5644 same TCP/IP connection, and then failed with "socket operation on non-
5645 socket" when it tried to send the third.
5647 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5649 27. The extern definition of crypt16() in expand.c was not being excluded when
5650 the OS had its own crypt16() function.
5652 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5653 as a preferred synonym for return_size_limit, both as an option and as an
5656 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5658 30. Changed the default configuration ACL to relax the local part checking rule
5659 for addresses that are not in any local domains. For these addresses,
5660 slashes and pipe symbols are allowed within local parts, but the sequence
5661 /../ is explicitly forbidden.
5663 31. SPA server authentication was not clearing the challenge buffer before
5666 32. log_message in a "warn" ACL statement was writing to the reject log as
5667 well as to the main log, which contradicts the documentation and doesn't
5668 seem right (because no rejection is happening). So I have stopped it.
5670 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5671 However, I am unable to do any testing of this.
5673 34. Fixed an infelicity in the appendfile transport. When checking directories
5674 for a mailbox, to see if any needed to be created, it was accidentally
5675 using path names with one or more superfluous leading slashes; tracing
5676 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5678 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5679 contents are added to the log line that is written for every discarded
5680 recipient. (Previously a log_message setting was ignored.)
5682 36. The ${quote: operator now quotes the string if it is empty.
5684 37. The install script runs exim in order to find its version number. If for
5685 some reason other than non-existence or emptiness, which it checks, it
5686 could not run './exim', it was installing it with an empty version number,
5687 i.e. as "exim-". This error state is now caught, and the installation is
5690 38. An argument was missing from the function that creates an error message
5691 when Exim fails to connect to the socket for saslauthd authentication.
5692 This could cause Exim to crash, or give a corrupted message.
5694 39. Added isip, isip4, and isip6 to ${if conditions.
5696 40. The ACL variables $acl_xx are now saved with the message, and can be
5697 accessed later in routers, transports, and filters.
5699 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5700 strings in the file are not string-expanded.
5702 42. If a MAIL command specified a SIZE value that was too large to fit into an
5703 int variable, the check against message_size_limit failed. Such values are
5704 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5705 day this will have to be increased, but I don't think I want to be around
5706 when emails are that large.
5713 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5714 iconv() is not standard in FreeBSD.
5716 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5717 IPv6 enabled. The observed symptom was a segmentation fault on return from
5718 the function os_common_find_running_interfaces() in src/os.c.
5720 3. In the check_special_case() function in daemon.c I had used "errno" as an
5721 argument name, which causes warnings on some systems. This was basically a
5722 typo, since it was named "eno" in the comments!
5724 4. The code that waits for the clock to tick (at a resolution of some fraction
5725 of a second) so as to ensure message-id uniqueness was always waiting for
5726 at least one whole tick, when it could have waited for less. [This is
5727 almost certainly not relevant at current processor speeds, where it is
5728 unlikely to ever wait at all. But we try to future-proof.]
5730 5. The function that sleeps for a time interval that includes fractions of a
5731 second contained a race. It did not block SIGALRM between setting the
5732 timer, and suspending (a couple of lines later). If the interval was short
5733 and the sigsuspend() was delayed until after it had expired, the suspension
5734 never ended. On busy systems this could lead to processes getting stuck for
5737 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5738 process, before it forks any delivery processes. The open lookup caching
5739 mechanism meant that the open file or database connection was passed into
5740 the delivery process. The problem was that delivery processes always tidy
5741 up cached lookup data. This could cause a problem for the next delivery
5742 process started by the queue runner, because the external queue runner
5743 process does not know about the closure. So the next delivery process
5744 still has data in the lookup cache. In the case of a file lookup, there was
5745 no problem because closing a file descriptor in a subprocess doesn't affect
5746 the parent. However, if the lookup was caching a connection to a database,
5747 the connection was closed, and the second delivery process was likely to
5748 see errors such as "PGSQL: query failed: server closed the connection
5749 unexpectedly". The problem has been fixed by closing all cached lookups
5750 in a queue runner before running a delivery process.
5752 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5753 seem to have the "const" qualifier which it has on other OS. I've
5756 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5757 *of the same type* (client or server) with the same public name that an
5758 error should be diagnosed.
5760 9. When Exim looked up a host name for an IP address, but failed to find the
5761 original IP address when looking up the host name (a safety check), it
5762 output the message "<ip address> does not match any IP for NULL", which was
5763 confusing, to say the least. The bug was that the host name should have
5764 appeared instead of "NULL".
5766 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5767 user that is built into the binary, and the -C or -D options is used, root
5768 privilege is dropped before the configuration file is read. In addition,
5769 logging is switched to stderr instead of the normal log files. If the
5770 configuration then re-defines the Exim user, the unprivileged environment
5771 is probably not what is expected, so Exim logs a panic warning message (but
5774 However, if deliver_drop_privilege is set, the unprivileged state may well
5775 be exactly what is intended, so the warning has been cut out in that case,
5776 and Exim is allowed to try to write to its normal log files.
5782 1. smtp_return_error_details was not giving details for temporary sender
5783 or receiver verification errors.
5785 2. Diagnose a configuration error if two authenticators have the same public
5788 3. Exim used not to create the message log file for a message until the first
5789 delivery attempt. This could be confusing when incoming messages were held
5790 for policy or load reasons. The message log file is now created at the time
5791 the message is received, and an initial "Received" line is written to it.
5793 4. The automatically generated man page for command line options had a minor
5794 bug that caused no ill effects; however, a more serious problem was that
5795 the procedure for building the man page automatically didn't always
5796 operate. Consequently, release 4.20 contains an out-of-date version. This
5797 shouldn't happen again.
5799 5. When building Exim with embedded Perl support, the script that builds the
5800 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5801 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5803 6. The freeze_tell option was not being used for messages that were frozen on
5804 arrival, either by an ACL or by local_scan().
5806 7. Added the smtp_incomplete_transaction log selector.
5808 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5809 was accepting AUTH without a new EHLO.
5811 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5812 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5815 10. Logging of TCP/IP connections (when configured) now happens in the main
5816 daemon process instead of the child process, so that the TCP/IP connection
5817 count is more accurate (but it can never be perfect).
5819 11. The use of "drop" in a nested ACL was not being handled correctly in the
5820 outer ACL. Now, if condition failure induced by the nested "drop" causes
5821 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5822 or "require"), the connection is dropped.
5824 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5825 that yield "discard" can now be used with an "accept" or a "discard" verb,
5826 but an error is generated for any others (because I can't see a useful way
5827 to define what should happen).
5829 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5830 are now processed in the same way as lines in the Exim configuration file.
5831 In particular, continuation lines are supported.
5833 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5835 15. Added -ti meaning -t -i.
5837 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5838 domains, and warn by logging if others are found.
5840 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5841 returned from GIFCONF when Exim is trying to find the list of interfaces on
5842 a host. The code in os.c has been modified to copy each ifreq to an aligned
5843 structure in all cases.
5845 Also, in some cases, the returned ifreq's were being copied to a 'struct
5846 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5847 means the last couple of bytes of an IPv6 address could be chopped if the
5848 ifreq contained only a normal sockaddr (14 bytes storage).
5850 18. Named domain lists were not supported in the hosts_treat_as_local option.
5851 An entry such as +xxxx was not recognized, and was treated as a literal
5854 19. Ensure that header lines added by a DATA ACL are included in the reject log
5855 if the ACL subsequently rejects the message.
5857 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5858 MD5 (which is deprecated).
5860 21. When testing a filter file using -bf, Exim was writing a message when it
5861 took the sender from a "From " line in the message, but it was not doing so
5862 when it took $return_path from a Return-Path: header line. It now does.
5864 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5865 with a valid header line field name (a series of printing characters
5866 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5868 23. Changed "disc" in the source to "disk" to conform to the documentation and
5869 the book and for uniformity.
5871 24. Ignore Sendmail's -Ooption=value command line item.
5873 25. When execve() failed while trying to run a command in a pipe transport,
5874 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5875 could be confused with a return value of 69 from the command itself. This
5876 has been changed to 127, the value the shell returns if it is asked to run
5877 a non-existent command. The wording for the related log line suggests a
5878 non-existent command as the problem.
5880 26. If received_header_text expands to an empty string, do not add a Received:
5881 header line to the message. (Well, it adds a token one on the spool, but
5882 marks it "old" so that it doesn't get used or transmitted.)
5884 27. Installed eximstats 1.28 (addition of -nt option).
5886 28. There was no check for failure on the call to getsockname() in the daemon
5887 code. This can fail if there is a shortage of resources on the system, with
5888 ENOMEM, for example. A temporary error is now given on failure.
5890 29. Contrary to the C standard, it seems that in some environments, the
5891 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5892 program. Exim now does this explicitly; it affects the formatting of
5893 timestamps using strftime().
5895 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5896 complaints. I've now initialized all the variables, to avoid this.
5898 32. Header lines added by a system filter were not being "seen" during
5899 transport-time rewrites.
5901 33. The info_callback() function passed to OpenSSL is set up with type void
5902 (*)(SSL *, int, int), as described somewhere. However, when calling the
5903 function (actually a macro) that sets it up, the type void(*)() is
5904 expected. I've put in a cast to prevent warnings from picky compilers.
5906 34. If a DNS black list lookup found a CNAME record, but there were no A
5907 records associated with the domain it pointed at, Exim crashed.
5909 35. If a DNS black list lookup returned more than one A record, Exim ignored
5910 all but the first. It now scans all returned addresses if a particular IP
5911 value is being sought. In this situation, the contents of the
5912 $dnslist_value variable are a list of all the addresses, separated by a
5915 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5916 to accept a host name and all its aliases if the forward lookup for any of
5917 them yielded the IP address of the incoming connection. Now it accepts only
5918 those names whose forward lookup yields the correct IP address. Any other
5919 names are discarded. This closes a loophole whereby a rogue DNS
5920 administrator could create reverse DNS records to break through a
5921 wildcarded host restriction in an ACL.
5923 37. If a user filter or a system filter that ran in a subprocess used any of
5924 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5925 the wrong values were passed to the pipe command ($thisaddress had the
5926 value of $0, $0 had the value of $1, etc). This bug was introduced by
5927 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5929 38. Improved the line breaking for long SMTP error messages from ACLs.
5930 Previously, if there was no break point between 40 and 75 characters, Exim
5931 left the rest of the message alone. Two changes have been made: (a) I've
5932 reduced the minimum length to 35 characters; (b) if it can't find a break
5933 point between 35 and 75 characters, it looks ahead and uses the first one
5934 that it finds. This may give the occasional overlong line, but at least the
5935 remaining text gets split now.
5937 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5938 file descriptors might be low, and that setting 1000 would always raise it.
5939 It turns out that in some environments, the limit is already over 1000 and
5940 that lowering it causes trouble. So now Exim takes care not to decrease it.
5942 40. When delivering a message, the value of $return_path is set to $sender_
5943 address at the start of routing (routers may change the value). By an
5944 oversight, this default was not being set up when an address was tested by
5945 -bt or -bv, which affected the outcome if any router or filter referred to
5948 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5949 log only when "message" or "log_message" are set. However, if one of the
5950 conditions was an address verification, or a call to a nested ACL, the
5951 messages generated by the underlying test were being passed through. This
5952 no longer happens. The underlying message is available in $acl_verify_
5953 message for both "message" and "log_message" expansions, so it can be
5954 passed through if needed.
5956 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5957 new expansion $bh_ to give the encoded byte string without charset
5958 translation. Translation happens only if iconv() is available; HAVE_ICONV
5959 indicates this at build time. HEADERS_CHARSET gives the charset to
5960 translate to; headers_charset can change it in the configuration, and
5961 "headers charset" can change it in an individual filter file.
5963 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5964 that creates RFC 2047 encoded "words" labels them as that charset instead
5965 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5966 expansion operator; (ii) when Exim creates a From: line for a local
5967 message; (iii) when a header line is rewritten to include a "phrase" part.
5969 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5970 buggy, causing it to skip the first lines of messages whose message ID
5971 ended in 'D'. This would not have bitten before Exim release 4.14, because
5972 message IDs were unlikely to end in 'D' before then. The effect was to have
5973 incorrect size information for certain domains.
5975 45. #include "config.h" was missing at the start of the crypt16.c module. This
5976 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5979 46. If there was a timeout during a "random" callout check, Exim treated it as
5980 a failure of the random address, and carried on sending RSET and the real
5981 address. If the delay was just some slowness somewhere, the response to the
5982 original RCPT would be taken as a response to RSET and so on, causing
5983 mayhem of various kinds.
5985 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5986 when I implemented it. It didn't allow for the fact that some option values
5987 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5988 the right test for positive values.
5990 48. Domain names in DNS records are case-independent. Exim always looks them up
5991 in lower case. Some resolvers return domain names in exactly the case they
5992 appear in the zone file, that is, they may contain uppercase letters. Not
5993 all resolvers do this - some return always lower case. Exim was treating a
5994 change of case by a resolver as a change of domain, similar to a widening
5995 of a domain abbreviation. This triggered its re-routing code and so it was
5996 trying to route what was effectively the same domain again. This normally
5997 caused routing to fail (because the router wouldn't handle the domain
5998 twice). Now Exim checks for this case specially, and just changes the
5999 casing of the domain that it ultimately uses when it transmits the message
6002 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6005 50. If a filter generated a file delivery with a non-absolute name (possible if
6006 no home directory exists for the router), the forbid_file option was not
6009 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6010 the existing equality matching.
6012 52. Exim was using ints instead of ino_t variables in some places where it was
6013 dealing with inode numbers.
6015 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6016 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6017 called TMPDIR, and if it finds it is different, it changes its value.
6019 54. The smtp_printf() function is now made available to local_scan() so
6020 additional output lines can be written before returning. There is also an
6021 smtp_fflush() function to enable the detection of a dropped connection.
6022 The variables smtp_input and smtp_batched_input are exported to
6025 55. Changed the default runtime configuration: the message "Unknown user"
6026 has been removed from the ACL, and instead placed on the localuser router,
6027 using the cannot_route_message feature. This means that any verification
6028 failures that generate their own messages won't get overridden. Similarly,
6029 the "Unrouteable address" message that was in the ACL for unverifiable
6030 relay addresses has also been removed.
6032 56. Added hosts_avoid_esmtp to the smtp transport.
6034 57. The exicyclog script was not checking for the esoteric option
6035 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6036 will work only if exicyclog is run under the appropriate euid.
6038 58. Following a discussion on the list, the rules by which Exim recognises line
6039 endings on incoming messages have been changed. The -dropcr and drop_cr
6040 options are now no-ops, retained only for backwards compatibility. The
6041 following line terminators are recognized: LF CRLF CR. However, special
6042 processing applies to CR:
6044 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6045 nor a local message in the state where . is a terminator.
6047 (ii) If a bare CR is encountered in a header line, an extra space is added
6048 after the line terminator so as not to end the header. The reasoning
6049 behind this is that bare CRs in header lines are most likely either
6050 to be mistakes, or people trying to play silly games.
6052 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6053 was being incorrectly given as 18 bytes larger than it should have been.
6054 This is a VOB (very old bug).
6056 60. This may never have affected anything current, but just in case it has:
6057 When the local host is found other than at the start of a list of hosts,
6058 the local host, those with the same MX, and any that follow, are discarded.
6059 When the list in question was part of a longer list of hosts, the following
6060 hosts (not currently being processed) were also being discarded. This no
6061 longer happens. I'm not sure if this situation could ever has previously
6064 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6067 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6068 header. This is something that is recommended in a new Internet Draft, and
6069 is something that is documented as being done by Sendmail. There are two
6070 possible values. For messages generated by the autoreply transport, Exim
6073 Auto-Submitted: auto-replied
6075 whereas for all other generated messages (e.g. bounces) it adds
6077 Auto-Submitted: auto-generated
6079 63. The "personal" condition in filters now includes a test for the
6080 Auto-Submitted: header. If it contains the string "auto-" the message it
6081 not considered personal.
6083 64. Added rcpt_include_affixes as a generic transport option.
6085 65. Added queue_only_override (default true).
6087 66. Added the syslog_duplication option.
6089 67. If what should have been the first header line of a message consisted of
6090 a space followed by a colon, Exim was mis-interpreting it as a header line.
6091 It isn't of course - it is syntactically invalid and should therefore be
6092 treated as the start of the message body. The misbehaviour could have
6093 caused a number of strange effects, including loss of data in subsequent
6094 header lines, and spool format errors.
6096 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6097 client host had authenticated. This control can now be exercised by an ACL
6098 for more flexibility.
6100 69. By default, callouts do not happen when testing with -bh. There is now a
6101 variant, -bhc, which does actually run the callout code, including
6102 consulting and updating the callout cache.
6104 70. Added support for saslauthd authentication, courtesy of Alexander
6107 71. If statvfs() failed on the spool or log directories while checking their
6108 size for availability, Exim confusingly gave the error "space shortage".
6109 Furthermore, in debugging mode it crashed with a floating point exception.
6110 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6111 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6112 serious problem, Exim now writes to the main and panic logs when this
6113 happens, with details of the failure. It then refuses to accept the
6114 incoming message, giving the message "spool directory problem" or "log
6115 directory problem" with a 421 code for SMTP messages.
6117 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6118 0, 1, and 2 exist, because some OS complain for execs without them (see
6119 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6120 descriptors. However, the code omitted to check that the open succeeded,
6121 causing mysterious errors if for some reason the permissions on /dev/null
6122 got screwed. Now Exim writes a message to the main and panic logs, and
6123 bombs out if it can't open /dev/null.
6125 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6126 interact so that it is all more flexible. It is supposed to remain
6127 backwards compatible. Also added extra_local_interfaces.
6129 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6130 to bomb out with an assertion failure - to the client this appears as a
6131 connection drop. This problem occurs in the part of the code that was taken
6132 from the Samba project. Fortunately, the assertion is in a very simple
6133 function, so I have fixed this by reproducing the function inline in the
6134 one place where it is called, and arranging for authentication to fail
6135 instead of killing the process with assert().
6137 75. The SPA client code was not working when the server requested OEM rather
6138 than Unicode encoding.
6140 76. Added code to make require_files with a specific uid setting more usable in
6141 the case where statting the file as root fails - usually a non-root-mounted
6142 NFS file system. When this happens and the failure is EACCES, Exim now
6143 forks a subprocess and does the per-uid checking as the relevant uid.
6145 77. Added process_log_path.
6147 78. If log_file_path was not explicitly set, a setting of check_log_space or
6148 check_log_inodes was ignored.
6150 79. If a space check for the spool or log partitions fails, the incident is now
6151 logged. Of course, in the latter case the data may get lost...
6153 80. Added the %p formatting code to string_format() so that it can be used to
6154 print addresses in debug_print(). Adjusted all the address printing in the
6155 debugging in store.c to use %p rather than %d.
6157 81. There was a concern that a line of code in smtp_in.c could overflow a
6158 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6159 initially expressed, the concern was not well-founded, because trailing
6160 spaces are removed early. However, if the trailing spaces were followed by
6161 a NULL, they did not get removed, so the overflow was possible. Two fixes
6164 (a) I re-wrote the offending code in a cleaner fashion.
6165 (b) If an incoming SMTP command contains a NULL character, it is rejected
6168 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6169 runs initgroups(), so that if the Exim user is in any additional groups,
6170 they will be used during message reception.
6176 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob