1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
18 TL/01 When running the test suite, the README says that variables such as
19 no_msglog_check are global and can be placed anywhere in a specific
20 test's script, however it was observed that placement needed to be near
21 the beginning for it to behave that way. Changed the runtest perl
22 script to read through the entire script once to detect and set these
23 variables, reset to the beginning of the script, and then run through
24 the script parsing/test process like normal.
26 TL/02 The BSD's have an arc4random API. One of the functions to induce
27 adding randomness was arc4random_stir(), but it has been removed in
28 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
29 function when detected.
31 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
32 cause callback expansion.
34 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
35 syntax errors in an expansion can be treated as a string instead of
36 logging or causing an error, due to the internal use of bool_lax
37 instead of bool when processing it.
39 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
40 server certificates when making smtp deliveries.
42 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
44 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
46 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
48 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
49 Merged patch from Sebastian Wiedenroth.
51 JH/05 Fix results-pipe from transport process. Several recipients, combined
52 with certificate use, exposed issues where response data items split
53 over buffer boundaries were not parsed properly. This eventually
54 resulted in duplicates being sent. This issue only became common enough
55 to notice due to the introduction of conection certificate information,
56 the item size being so much larger. Found and fixed by Wolfgang Breyha.
58 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
59 size buffer was used, resulting in syntax errors when an expansion
62 JH/07 Add support for directories of certificates when compiled with a GnuTLS
63 version 3.3.6 or later.
65 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
66 is EXPERIMENTAL_EVENT, the main-configuration and transport options
67 both become "event_action", the variables become $event_name, $event_data
68 and $event_defer_errno. There is a new variable $verify_mode, usable in
69 routers, transports and related events. The tls:cert event is now also
70 raised for inbound connections, if the main configuration event_action
73 TL/06 In test suite, disable OCSP for old versions of openssl which contained
74 early OCSP support, but no stapling (appears to be less than 1.0.0).
76 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
77 server certificate names available under the smtp transport option
78 "tls_verify_cert_hostname" now do not permit multi-component wildcard
81 JH/10 Time-related extraction expansions from certificates now use the main
82 option "timezone" setting for output formatting, and are consistent
83 between OpenSSL and GnuTLS compilations. Bug 1541.
85 JH/11 Fix a crash in mime ACL when meeting a zero-length parameter in the
88 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
89 include certificate info, eximon was claiming there were spoolfile
92 JH/13 Buf 1521: Fix ldap lookup for single-attr request, multiple-attr return.
94 JH/14 Log delivery-related information more consistently, using the sequence
95 "H=<name> [<ip>]" wherever possible.
100 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
101 checkers that were complaining about end of non-void function with no
104 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
105 This was a regression intruduced in 4.83 by another bugfix.
107 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
109 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
110 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
116 TF/01 Correctly close the server side of TLS when forking for delivery.
118 When a message was received over SMTP with TLS, Exim failed to clear up
119 the incoming connection properly after forking off the child process to
120 deliver the message. In some situations the subsequent outgoing
121 delivery connection happened to have the same fd number as the incoming
122 connection previously had. Exim would try to use TLS and fail, logging
123 a "Bad file descriptor" error.
125 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
126 utilities have not been installed.
128 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
129 temporary space as the ACL may create new global variables.
131 TL/01 LDAP support uses per connection or global context settings, depending
132 upon the detected version of the libraries at build time.
134 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
135 to extract and use the src ip:port in logging and expansions as if it
136 were a direct connection from the outside internet. PPv2 support was
137 updated based on HAProxy spec change in May 2014.
139 JH/02 Add ${listextract {number}{list}{success}{fail}}.
141 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
142 Properly escape header and check for NULL return.
144 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
147 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
149 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
150 characters in header names, implemented as a verify condition.
151 Contributed by Michael Fischer v. Mollard.
153 TL/05 Rename SPF condition results err_perm and err_temp to standardized
154 results permerror and temperror. Previous values are deprecated but
155 still accepted. In a future release, err_perm and err_temp will be
156 completely removed, which will be a backward incompatibility if the
157 ACL tests for either of these two old results. Patch contributed by
158 user bes-internal on the mailing list.
160 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
162 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
163 selectors, in both main and reject logs.
165 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
168 JH/07 Add malware type "sock" for talking to simple daemon.
170 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
172 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
173 routers/transports under cutthrough routing.
175 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
176 numbers. Touch up "bool" conditional to keep the same definition.
178 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
180 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
182 JH/12 Expand items in router/transport headers_add or headers_remove lists
183 individually rather than the list as a whole. Bug 1452.
185 Required for reasonable handling of multiple headers_ options when
186 they may be empty; requires that headers_remove items with embedded
187 colons must have them doubled (or the list-separator changed).
189 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
190 view the policy declared in the DMARC record. Currently, $dmarc_status
191 is a combined value of both the record presence and the result of the
194 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
196 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
197 dnslookup router and the smtp transport (applying to the forward
200 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
201 of ldap servers used for a specific lookup. Patch provided by Heiko
204 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
205 New variable $lookup_dnssec_authenticated for observability.
207 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
208 Patch submitted by Lars Timman.
210 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
212 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
213 Requires trusted mode and valid format message id, aborts otherwise.
214 Patch contributed by Heiko Schlichting.
216 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
217 certextract with support for various fields. Bug 1358.
219 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
220 is requested by default, modifiable by smtp transport option
223 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
224 operate on certificate variables to give certificate fingerprints
225 Also new ${sha256:cert_variable}.
227 JH/23 The PRDR feature is moved from being Experimental into the mainline.
229 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
230 Christian Aistleitner.
232 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
234 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
235 file. Patch from Wolfgang Breyha.
237 JH/25 Expand the coverage of the delivery $host and $host_address to
238 client authenticators run in verify callout. Bug 1476.
240 JH/26 Port service names are now accepted for tls_on_connect_ports, to
241 align with daemon_smtp_ports. Bug 72.
243 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
244 support and error reporting did not work properly.
246 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
247 and is readable. Patch from Andrew Colin Kissa.
249 TL/14 Enhance documentation of ${run expansion and how it parses the
250 commandline after expansion, particularly in the case when an
251 unquoted variable expansion results in an empty value.
253 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
255 PP/02 Fix internal collision of T_APL on systems which support RFC3123
256 by renaming away from it. Addresses GH issue 15, reported by
259 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
261 TL/15 SECURITY: prevent double expansion in math comparison functions
262 (can expand unsanitized data). Not remotely exploitable.
269 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
271 PP/02 Make -n do something, by making it not do something.
272 When combined with -bP, the name of an option is not output.
274 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
277 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
278 $sender_host_name and config options to manage this, and basic check
281 PP/05 DSCP support for outbound connections and control modifier for inbound.
283 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
284 (Only plugin which currently uses this is kerberos4, which nobody should
285 be using, but we should make it available and other future plugins might
286 conceivably use it, even though it would break NAT; stuff *should* be
287 using channel bindings instead).
289 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
290 name; added for Sendmail compatibility; requires admin caller.
291 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
292 just ignore it); requires trusted caller.
293 Also parse but ignore: -Ac -Am -X<logfile>
296 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
298 TL/02 Add +smtp_confirmation as a default logging option.
300 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
301 Patch by Magnus Holmgren from 2007-02-20.
303 TL/04 Bugzilla 1281 - Spec typo.
304 Bugzilla 1283 - Spec typo.
305 Bugzilla 1290 - Spec grammar fixes.
307 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
309 TL/06 Add Experimental DMARC support using libopendmarc libraries.
311 TL/07 Fix an out of order global option causing a segfault. Reported to dev
312 mailing list by by Dmitry Isaikin.
314 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
316 JH/02 Support "G" suffix to numbers in ${if comparisons.
318 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
320 NM/01 Bugzilla 1197 - Spec typo
321 Bugzilla 1196 - Spec examples corrections
323 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
325 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
326 gnutls_enable_pkcs11, but renamed to more accurately indicate its
329 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
330 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
332 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
333 "acl {{name}{arg}...}", and optional args on acl condition
336 JH/05 Permit multiple router/transport headers_add/remove lines.
338 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
340 JH/07 Avoid using a waiting database for a single-message-only transport.
341 Performance patch from Paul Fisher. Bugzilla 1262.
343 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
346 JH/09 Add $headers_added variable, with content from use of ACL modifier
347 add_header (but not yet added to the message). Bugzilla 199.
349 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
350 Pulled from Bugzilla 817 by Wolfgang Breyha.
352 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
354 (nb: this is the same fix as in Exim 4.80.1)
356 JH/11 Add A= logging on delivery lines, and a client_set_id option on
359 JH/12 Add optional authenticated_sender logging to A= and a log_selector
362 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
364 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
365 advertise SMTP AUTH mechanism to us, instead of a generic
366 protocol violation error. Also, make Exim more robust to bad
367 data from the Dovecot auth socket.
369 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
371 When a queue runner is handling a message, Exim first routes the
372 recipient addresses, during which it prunes them based on the retry
373 hints database. After that it attempts to deliver the message to
374 any remaining recipients. It then updates the hints database using
377 So if a recipient address works intermittently, it can get repeatedly
378 deferred at routing time. The retry hints record remains fresh so the
379 address never reaches the final cutoff time.
381 This is a fairly common occurrence when a user is bumping up against
382 their storage quota. Exim had some logic in its local delivery code
383 to deal with this. However it did not apply to per-recipient defers
384 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
386 This change adds a proper retry rule check during routing so that the
387 final cutoff time is checked against the message's age. We only do
388 this check if there is an address retry record and there is not a
389 domain retry record; this implies that previous attempts to handle
390 the address had the retry_use_local_parts option turned on. We use
391 this as an approximation for the destination being like a local
392 delivery, as in LMTP.
394 I suspect this new check makes the old local delivery cutoff check
395 redundant, but I have not verified this so I left the code in place.
397 TF/02 Correct gecos expansion when From: is a prefix of the username.
399 Test 0254 submits a message to Exim with the header
403 When I ran the test suite under the user fanf2, Exim expanded
404 the header to contain my full name, whereas it should have added
405 a Resent-Sender: header. It erroneously treats any prefix of the
406 username as equal to the username.
408 This change corrects that bug.
410 GF/01 DCC debug and logging tidyup
411 Error conditions log to paniclog rather than rejectlog.
412 Debug lines prefixed by "DCC: " to remove any ambiguity.
414 TF/03 Avoid unnecessary rebuilds of lookup-related code.
416 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
417 Bug spotted by Jeremy Harris; was flawed since initial commit.
418 Would have resulted in OCSP responses post-SNI triggering an Exim
419 NULL dereference and crash.
421 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
423 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
424 Bug detection, analysis and fix by Samuel Thibault.
425 Bugzilla 1331, Debian bug #698092.
427 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
429 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
430 Server implementation by Todd Lyons, client by JH.
431 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
432 config variable "prdr_enable" controls whether the server
433 advertises the facility. If the client requests PRDR a new
434 acl_data_smtp_prdr ACL is called once for each recipient, after
435 the body content is received and before the acl_smtp_data ACL.
436 The client is controlled by bolth of: a hosts_try_prdr option
437 on the smtp transport, and the server advertisement.
438 Default client logging of deliveries and rejections involving
439 PRDR are flagged with the string "PRDR".
441 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
442 fclose(). Diagnosis by Todd Lyons.
444 PP/17 Update configure.default to handle IPv6 localhost better.
445 Patch by Alain Williams (plus minor tweaks).
448 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
449 This is now consistent with GnuTLS, and is now documented: the
450 previous undocumented portable approach to treating the option as
451 unset was to force an expansion failure. That still works, and
452 an empty string is now equivalent.
454 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
455 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
456 not performing validation itself.
458 PP/20 Added force_command boolean option to pipe transport.
459 Patch from Nick Koston, of cPanel Inc.
461 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
464 TF/04 Added udpsend ACL modifer and hexquote expansion operator
466 PP/21 Fix eximon continuous updating with timestamped log-files.
467 Broken in a format-string cleanup in 4.80, missed when I repaired the
468 other false fix of the same issue.
469 Report and fix from Heiko Schlichting.
472 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
473 Report from Prashanth Katuri.
475 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
476 It's SecureTransport, so affects any MacOS clients which use the
477 system-integrated TLS libraries, including email clients.
479 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
480 using a MIME ACL for non-SMTP local injection.
481 Report and assistance in diagnosis by Warren Baker.
483 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
485 JH/16 Fix comparisons for 64b. Bugzilla 1385.
487 TL/09 Add expansion variable $authenticated_fail_id to keep track of
488 last id that failed so it may be referenced in subsequent ACL's.
490 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
493 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
494 ldap library initialization, allowing self-signed CA's to be
495 used. Also properly sets require_cert option later in code by
496 using NULL (global ldap config) instead of ldap handle (per
497 session). Bug diagnosis and testing by alxgomz.
499 TL/12 Enhanced documentation in the ratelimit.pl script provided in
500 the src/util/ subdirectory.
502 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
503 renamed to Transport Post Delivery Action by Jeremy Harris, as
506 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
507 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
508 redis_servers = needs to be configured which will be used by the redis
509 lookup. Patch from Warren Baker, of The Packet Hub.
511 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
513 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
514 hostname or reverse DNS when processing a host list. Used suggestions
515 from multiple comments on this bug.
517 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
519 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
520 Missed a few lines, added it to make the runtest require no keyboard
523 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
524 contains upper case chars. Make router use caseful_local_part.
526 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
527 support when GnuTLS has been built with p11-kit.
533 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
535 This, or similar/improved, will also be change PP/11 of 4.82.
541 PP/01 Handle short writes when writing local log-files.
542 In practice, only affects FreeBSD (8 onwards).
543 Bugzilla 1053, with thanks to Dmitry Isaikin.
545 NM/01 Bugzilla 949 - Documentation tweak
547 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
550 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
552 PP/02 Implemented gsasl authenticator.
554 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
556 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
557 `pkg-config foo` for cflags/libs.
559 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
560 with rest of GSASL and with heimdal_gssapi.
562 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
563 `pkg-config foo` for cflags/libs for the TLS implementation.
565 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
566 properties get this fed in as external SSF. A number of robustness
567 and debugging improvements to the cyrus_sasl authenticator.
569 PP/08 cyrus_sasl server now expands the server_realm option.
571 PP/09 Bugzilla 1214 - Log authentication information in reject log.
572 Patch by Jeremy Harris.
574 PP/10 Added dbmjz lookup type.
576 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
578 PP/12 MAIL args handles TAB as well as SP, for better interop with
579 non-compliant senders.
580 Analysis and variant patch by Todd Lyons.
582 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
583 Bug report from Lars Müller <lars@samba.org> (via SUSE),
584 Patch from Dirk Mueller <dmueller@suse.com>
586 PP/13 tls_peerdn now print-escaped for spool files.
587 Observed some $tls_peerdn in wild which contained \n, which resulted
588 in spool file corruption.
590 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
591 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
592 or write after TLS renegotiation, which otherwise led to messages
595 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
596 as a tracking header (ie: a signed header comes before the signature).
597 Patch from Wolfgang Breyha.
599 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
600 comma-sep list; embedded commas doubled.
602 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
604 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
606 Report and patch from Dmitry Banschikov.
608 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
609 Removed SSL_clear() after SSL_new() which led to protocol negotiation
610 failures. We appear to now support TLS1.1+ with Exim.
612 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
613 lets Exim select keys and certificates based upon TLS SNI from client.
614 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
615 before an outbound SMTP session. New log_selector, +tls_sni.
617 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
618 NULL dereference. Report and patch from Alun Jones.
620 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
621 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
622 Not seeing resolver debug output on NetBSD, but suspect this is a
623 resolver implementation change.
625 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
626 Left warnings. Added "eximon gdb" invocation mode.
628 PP/21 Defaulting "accept_8bitmime" to true, not false.
630 PP/22 Added -bw for inetd wait mode support.
632 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
633 locate the relevant includes and libraries. Made this the default.
635 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
636 Bugzilla 1246, report and most of solution from Tomasz Kusy.
638 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
639 This may cause build issues on older platforms.
641 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
642 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
643 gnutls_require_mac & gnutls_require_protocols (no longer supported).
644 Added SNI support via GnuTLS too.
645 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
647 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
649 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
650 Applied second patch from Janne, implementing suggestion to default
651 multiple-strings-in-record handling to match SPF spec.
653 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
655 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
656 read-only, out of scope).
657 Patch from Wolfgang Breyha, report from Stuart Northfield.
659 PP/29 Fix three issues highlighted by clang analyser static analysis.
660 Only crash-plausible issue would require the Cambridge-specific
661 iplookup router and a misconfiguration.
662 Report from Marcin Mirosław.
664 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
666 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
667 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
668 As part of this, removing so much warning spew let me fix some minor
669 real issues in debug logging.
671 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
672 assignment on my part. Fixed.
674 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
675 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
676 Janne Snabb (who went above and beyond: thank you).
678 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
679 string otherwise requires a connection and a bunch more work and it's
680 relatively easy to get wrong. Should also expose TLS library linkage
683 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
684 64-bit ${eval} (JH/03).
686 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
687 GNU libc to support some of the 64-bit stuff, should not lead to
688 conflicts. Defined before os.h is pulled in, so if a given platform
689 needs to override this, it can.
691 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
692 protection layer was required, which is not implemented.
693 Bugzilla 1254, patch from Wolfgang Breyha.
695 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
696 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
697 tls_dhparam take prime identifiers. Also unbreak combination of
698 OpenSSL+DH_params+TLSSNI.
700 PP/39 Disable SSLv2 by default in OpenSSL support.
706 PP/01 Solaris build fix for Oracle's LDAP libraries.
707 Bugzilla 1109, patch from Stephen Usher.
709 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
711 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
714 TF/02 Fix a couple more cases where we did not log the error message
715 when unlink() failed. See also change 4.74-TF/03.
717 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
718 lock up or crash if it happened to be inside a call to libc when it
719 got a SIGUSR1 from exiwhat.
721 The SIGUSR1 handler appends the current process status to the process
722 log which is later printed by exiwhat. It used to use the general
723 purpose logging code to do this, but several functions it calls are
724 not safe for signals.
726 The new output code in the SIGUSR1 handler is specific to the process
727 log, and simple enough that it's easy to inspect for signal safety.
728 Removing some special cases also simplifies the general logging code.
729 Removing the spurious timestamps from the process log simplifies
732 TF/04 Improved ratelimit ACL condition.
734 The /noupdate option has been deprecated in favour of /readonly which
735 has clearer semantics. The /leaky, /strict, and /readonly update modes
736 are mutually exclusive. The update mode is no longer included in the
737 database key; it just determines when the database is updated. (This
738 means that when you upgrde Exim will forget old rate measurements.)
740 Exim now checks that the per_* options are used with an update mode that
741 makes sense for the current ACL. For example, when Exim is processing a
742 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
743 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
744 must specify per_mail/readonly. If you omit the update mode it defaults to
745 /leaky where that makes sense (as before) or /readonly where required.
747 The /noupdate option is now undocumented but still supported for
748 backwards compatibility. It is equivalent to /readonly except that in
749 ACLs where /readonly is required you may specify /leaky/noupdate or
750 /strict/noupdate which are treated the same as /readonly.
752 A useful new feature is the /count= option. This is a generalization
753 of the per_byte option, so that you can measure the throughput of other
754 aggregate values. For example, the per_byte option is now equivalent
755 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
757 The per_rcpt option has been generalized using the /count= mechanism
758 (though it's more complicated than the per_byte equivalence). When it is
759 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
760 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
761 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
762 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
763 Note that using per_rcpt with a non-readonly update mode in more than
764 one ACL will cause the recipients to be double-counted. (The per_mail
765 and per_byte options don't have this problem.)
767 The handling of very low rates has changed slightly. If the computed rate
768 is less than the event's count (usually one) then this event is the first
769 after a long gap. In this case the rate is set to the same as this event's
770 count, so that the first message of a spam run is counted properly.
772 The major new feature is a mechanism for counting the rate of unique
773 events. The new per_addr option counts the number of different
774 recipients that someone has sent messages to in the last time period. It
775 behaves like per_rcpt if all the recipient addresses are different, but
776 duplicate recipient addresses do not increase the measured rate. Like
777 the /count= option this is a general mechanism, so the per_addr option
778 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
779 example, measure the rate that a client uses different sender addresses
780 with the options per_mail/unique=$sender_address. There are further
781 details in the main documentation.
783 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
785 TF/06 Removed a few PCRE remnants.
787 TF/07 Automatically extract Exim's version number from tags in the git
788 repository when doing development or release builds.
790 PP/02 Raise smtp_cmd_buffer_size to 16kB.
791 Bugzilla 879. Patch from Paul Fisher.
793 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
794 Heavily based on revision 40f9a89a from Simon Arlott's tree.
797 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
799 PP/05 Variable $av_failed, true if the AV scanner deferred.
800 Bugzilla 1078. Patch from John Horne.
802 PP/06 Stop make process more reliably on build failure.
803 Bugzilla 1087. Patch from Heiko Schlittermann.
805 PP/07 Make maildir_use_size_file an _expandable_ boolean.
806 Bugzilla 1089. Patch from Heiko Schlittermann.
808 PP/08 Handle ${run} returning more data than OS pipe buffer size.
809 Bugzilla 1131. Patch from Holger Weiß.
811 PP/09 Handle IPv6 addresses with SPF.
812 Bugzilla 860. Patch from Wolfgang Breyha.
814 PP/10 GnuTLS: support TLS 1.2 & 1.1.
816 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
819 PP/11 match_* no longer expand right-hand-side by default.
820 New compile-time build option, EXPAND_LISTMATCH_RHS.
821 New expansion conditions, "inlist", "inlisti".
823 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
825 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
827 PP/14 fix log_write() format string regression from TF/03.
828 Bugzilla 1152. Patch from Dmitry Isaikin.
834 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
836 PP/02 Harmonised TLS library version reporting; only show if debugging.
837 Layout now matches that introduced for other libraries in 4.74 PP/03.
839 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
841 PP/04 New "dns_use_edns0" global option.
843 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
846 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
847 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
849 TK/01 Updated PolarSSL code to 0.14.2.
850 Bugzilla 1097. Patch from Andreas Metzler.
852 PP/07 Catch divide-by-zero in ${eval:...}.
855 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
858 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
859 format-string attack -- SECURITY: remote arbitrary code execution.
861 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
862 time unintentionally subject to list matching rules, letting the header
863 cause arbitrary Exim lookups (of items which can occur in lists, *not*
864 arbitrary string expansion). This allowed for information disclosure.
866 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
867 INT_MIN/-1 -- value coerced to INT_MAX.
873 NM/01 Workround for PCRE version dependency in version reporting
876 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
877 This fixes portability to compilers other than gcc, notably
878 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
880 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
881 makefiles for portability to HP-UX and POSIX correctness.
883 PP/01 Permit LOOKUP_foo enabling on the make command-line.
884 Also via indented variable definition in the Makefile.
885 (Debugging by Oliver Heesakkers).
887 PP/02 Restore caching of spamd results with expanded spamd_address.
888 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
890 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
891 Improves build reliability. Fix from: Frank Elsner
893 NM/02 Fix wide character breakage in the rfc2047 coding
894 Fixes bug 1064. Patch from Andrey N. Oktyabrski
896 NM/03 Allow underscore in dnslist lookups
897 Fixes bug 1026. Patch from Graeme Fowler
899 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
900 Code patches from Adam Ciarcinski of NetBSD.
902 NM/04 Fixed exiqgrep to cope with mailq missing size issue
905 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
906 is logged, to avoid truncation. Patch from John Horne.
908 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
909 Patch from Jakob Hirsch.
911 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
912 SQL string expansion failure details.
913 Patch from Andrey Oktyabrski.
915 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
916 Patch from Simon Arlott.
918 PP/09 New lookups functionality failed to compile on old gcc which rejects
919 extern declarations in function scope.
920 Patch from Oliver Fleischmann
922 PP/10 Use sig_atomic_t for flags set from signal handlers.
923 Check getgroups() return and improve debugging.
924 Fixed developed for diagnosis in bug 927 (which turned out to be
927 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
928 Patch from Mark Zealey.
930 PP/12 Bugzilla 1056: Improved spamd server selection.
931 Patch from Mark Zealey.
933 PP/13 Bugzilla 1086: Deal with maildir quota file races.
934 Based on patch from Heiko Schlittermann.
936 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
937 Patch from Uwe Doering, sign-off by Michael Haardt.
939 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
940 variable declaration deep within a block. Bug and patch from
943 PP/15 lookups-Makefile IRIX compatibilty coercion.
945 PP/16 Make DISABLE_DKIM build knob functional.
947 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
948 Patch by Simon Arlott
950 TF/03 Fix valgrind.h portability to C89 compilers that do not support
951 variable argument macros. Our copy now differs from upstream.
957 TF/01 Failure to get a lock on a hints database can have serious
958 consequences so log it to the panic log.
960 TF/02 Log LMTP confirmation messages in the same way as SMTP,
961 controlled using the smtp_confirmation log selector.
963 TF/03 Include the error message when we fail to unlink a spool file.
965 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
966 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
967 for maintaining out-of-tree patches for some time.
969 PP/01 Bugzilla 139: Documentation and portability issues.
970 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
971 Handle per-OS dynamic-module compilation flags.
973 PP/02 Let /dev/null have normal permissions.
974 The 4.73 fixes were a little too stringent and complained about the
975 permissions on /dev/null. Exempt it from some checks.
976 Reported by Andreas M. Kirchwitz.
978 PP/03 Report version information for many libraries, including
979 Exim version information for dynamically loaded libraries. Created
980 version.h, now support a version extension string for distributors
981 who patch heavily. Dynamic module ABI change.
983 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
984 privilege escalation vulnerability whereby the Exim run-time user
985 can cause root to append content of the attacker's choosing to
988 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
991 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
992 If dropping privileges for untrusted macros, we disabled normal logging
993 on the basis that it would fail; for the Exim run-time user, this is not
994 the case, and it resulted in successful deliveries going unlogged.
995 Fixed. Reported by Andreas Metzler.
1001 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1002 only prepend for the Resent-* case. Fixes regression introduced in
1003 Exim 4.70 by NM/22 for Bugzilla 607.
1005 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1006 increasing numbers of administrators be bitten by this.
1008 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1010 PP/03 Bugzilla 994: added openssl_options main configuration option.
1012 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1014 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1016 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1018 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1019 without a peer certificate, leading to a segfault because of an
1020 assumption that peers always have certificates. Be a little more
1021 paranoid. Problem reported by Martin Tscholak.
1023 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1024 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1025 NB: ClamAV planning to remove STREAM in "middle of 2010".
1026 CL also introduces -bmalware, various -d+acl logging additions and
1027 more caution in buffer sizes.
1029 PP/09 Implemented reverse_ip expansion operator.
1031 PP/10 Bugzilla 937: provide a "debug" ACL control.
1033 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1035 PP/12 Bugzilla 973: Implement --version.
1037 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1039 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1041 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1043 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1044 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1045 ignore trailing whitespace.
1047 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1049 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1052 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1053 Notification from Dr Andrew Aitchison.
1055 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1056 ExtendedDetectionInfo response format.
1057 Notification from John Horne.
1059 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1062 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1063 XSL and documented dependency on system catalogs, with examples of how
1066 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1069 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1070 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1071 configuration file which is writeable by the Exim user or group.
1073 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1074 of configuration files to cover files specified with the -C option if
1075 they are going to be used with root privileges, not just the default
1078 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1079 option (effectively making it always true).
1081 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1082 files to be used while preserving root privileges.
1084 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1085 that rogue child processes cannot use them.
1087 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1088 run-time user, instead of root.
1090 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1091 Exim run-time user without dropping privileges.
1093 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1094 result string, instead of calling string_vformat() twice with the same
1097 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1098 for other users. Others should always drop root privileges if they use
1099 -C on the command line, even for a whitelisted configure file.
1101 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1103 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1109 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1110 $data_path, and $header_path variables; fixed documentation bugs and
1113 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1114 exipick to access non-standard spools, including the "frozen" queue
1117 NM/01 Bugzilla 965: Support mysql stored procedures.
1118 Patch from Alain Williams
1120 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1122 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1123 Patch from Andreas Metzler
1125 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1126 Patch from Kirill Miazine
1128 NM/05 Bugzilla 671: Added umask to procmail example.
1130 JJ/03 installed exipick 20100323.0, fixing doc bug
1132 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1133 directory. Notification and patch from Dan Rosenberg.
1135 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1137 TK/02 Improve log output when DKIM signing operation fails.
1139 MH/01 Treat the transport option dkim_domain as a colon separated
1140 list, not as a single string, and sign the message with each element,
1141 omitting multiple occurences of the same signer.
1143 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1144 Bugzilla 985, 986. Patch by Simon Arlott
1146 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1147 Patch by Simon Arlott
1149 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1150 MBX locking. Notification from Dan Rosenberg.
1156 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1158 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1160 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1162 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1164 NM/04 Bugzilla 915: Flag broken perl installation during build.
1170 TK/01 Added patch by Johannes Berg that expands the main option
1171 "spamd_address" if it starts with a dollar sign.
1173 TK/02 Write list of recipients to X-Envelope-Sender header when building
1174 the mbox-format spool file for content scanning (suggested by Jakob
1177 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1178 (http://www.dcc-servers.net/) support via dccifd. Activated by
1179 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1181 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1182 by Mark Daniel Reidel <mr@df.eu>.
1184 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1185 When building exim an external PCRE library is now needed -
1186 PCRE is a system library on the majority of modern systems.
1187 See entry on PCRE_LIBS in EDITME file.
1189 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1190 conversation. Added nologin parameter to request.
1191 Patch contributed by Kirill Miazine.
1193 TF/01 Do not log submission mode rewrites if they do not change the address.
1195 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1197 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1198 log files in place. Contributed by Roberto Lima.
1200 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1202 TF/03 Bugzilla 615: When checking the local_parts router precondition
1203 after a local_part_suffix or local_part_prefix option, Exim now
1204 does not use the address's named list lookup cache, since this
1205 contains cached lookups for the whole local part.
1207 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1208 Robert Millan. Documentation is in experimental-spec.txt.
1210 TF/04 Bugzilla 668: Fix parallel build (make -j).
1212 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1214 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1215 Patch provided by Jan Srzednicki.
1217 TF/05 Leading white space used to be stripped from $spam_report which
1218 wrecked the formatting. Now it is preserved.
1220 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1221 that they are available at delivery time.
1223 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1225 TF/08 TLS error reporting now respects the incoming_interface and
1226 incoming_port log selectors.
1228 TF/09 Produce a more useful error message if an SMTP transport's hosts
1229 setting expands to an empty string.
1231 NM/06 Bugzilla 744: EXPN did not work under TLS.
1232 Patch provided by Phil Pennock.
1234 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1235 Patch provided by Richard Godbee.
1237 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1238 acl_smtp_notquit, added index entry.
1240 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1241 Patch provided by Eugene Bujak.
1243 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1244 accept(). Patch provided by Maxim Dounin.
1246 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1247 Patch provided by Phil Pennock.
1249 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1251 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1252 Patch provided by Brad "anomie" Jorsch.
1254 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1255 Patch provided by Dean Brooks.
1257 TK/05 Add native DKIM support (does not depend on external libraries).
1259 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1260 Patch provided by Graeme Fowler.
1262 NM/16 Bugzilla 851: Documentation example syntax fix.
1264 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1266 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1269 NM/19 Bugzilla 745: TLS version reporting.
1270 Patch provided by Phil Pennock.
1272 NM/20 Bugzilla 167: bool: condition support.
1273 Patch provided by Phil Pennock.
1275 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1276 clients. Patch provided by Phil Pennock.
1278 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1279 Patch provided by Brad "anomie" Jorsch.
1281 NM/23 Bugzilla 687: Fix misparses in eximstats.
1282 Patch provided by Heiko Schlittermann.
1284 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1285 Patch provided by Heiko Schlittermann.
1287 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1288 plus update to original patch.
1290 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1292 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1293 Patch provided by David Brownlee.
1295 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1297 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1299 NM/30 Bugzilla 888: TLS documentation bugfixes.
1301 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1303 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1304 Unlike the original bugzilla I have changed all shell scripts in src tree.
1306 NM/33 Bugzilla 898: Transport filter timeout fix.
1307 Patch by Todd Rinaldo.
1309 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1310 Patch by Serge Demonchaux.
1312 NM/35 Bugzilla 39: Base64 decode bug fixes.
1313 Patch by Jakob Hirsch.
1315 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1317 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1319 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1321 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1327 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1328 ALT-N's libdkim that I have put here:
1329 http://duncanthrax.net/exim-experimental/
1331 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1332 were called 'true' and it seems that C99 defines that as a reserved
1333 keyword to be used with 'bool' variable types. That means you could
1334 not include C99-style headers which use bools without triggering
1335 build errors in sieve.c.
1337 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1338 as mailq or other aliases. Changed the --help handling significantly
1339 to do whats expected. exim_usage() emits usage/help information.
1341 SC/01 Added the -bylocaldomain option to eximstats.
1343 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1345 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1347 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1353 PH/01 Another patch from the Sieve maintainer.
1355 PH/02 When an IPv6 address is converted to a string for single-key lookup
1356 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1357 dots are used instead of colons so that keys in lsearch files need not
1358 contain colons. This was done some time before quoting was made available
1359 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1360 (notated using the quote facility) so as to distinguish them from IPv4
1361 keys. This meant that lookups for IP addresses in host lists did not work
1362 for iplsearch lookups.
1364 This has been fixed by arranging for IPv6 addresses to be expressed with
1365 colons if the lookup type is iplsearch. This is not incompatible, because
1366 previously such lookups could never work.
1368 The situation is now rather anomolous, since one *can* have colons in
1369 ordinary lsearch keys. However, making the change in all cases is
1370 incompatible and would probably break a number of configurations.
1372 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1375 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1376 conversion specification without a maximum field width, thereby enabling
1377 a rogue spamd server to cause a buffer overflow. While nobody in their
1378 right mind would setup Exim to query an untrusted spamd server, an
1379 attacker that gains access to a server running spamd could potentially
1380 exploit this vulnerability to run arbitrary code as the Exim user.
1382 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1383 $primary_hostname instead of what libspf2 thinks the hosts name is.
1385 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1386 a directory entry by the name of the lookup key. Previously, if a
1387 symlink pointed to a non-existing file or a file in a directory that
1388 Exim lacked permissions to read, a lookup for a key matching that
1389 symlink would fail. Now it is enough that a matching directory entry
1390 exists, symlink or not. (Bugzilla 503.)
1392 PH/03 The body_linecount and body_zerocount variables are now exported in the
1395 PH/04 Added the $dnslist_matched variable.
1397 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1398 This means they are set thereafter only if the connection becomes
1401 PH/06 Added the client_condition to authenticators so that some can be skipped
1402 by clients under certain conditions.
1404 PH/07 The error message for a badly-placed control=no_multiline_responses left
1405 "_responses" off the end of the name.
1407 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1409 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1410 (without spaces) instead of just copying the configuration text.
1412 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1414 PH/11 Added $max_received_linelength.
1416 PH/12 Added +ignore_defer and +include_defer to host lists.
1418 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1419 way in which PCRE > 7.0 is built.
1421 PH/14 Implemented queue_only_load_latch.
1423 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1424 MAIL command. The effect was to mangle the value on 64-bit systems.
1426 PH/16 Another patch from the Sieve maintainer.
1428 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1430 PH/18 If a system quota error occurred while trying to create the file for
1431 a maildir delivery, the message "Mailbox is full" was not appended to the
1432 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1433 only to a quota excession during the actual writing of the file.
1435 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1436 characters?) which causes problems in log lines. The DN values are now
1437 passed through string_printing() before being added to log lines.
1439 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1440 and InterBase are left for another time.)
1442 PH/21 Added message_body_newlines option.
1444 PH/22 Guard against possible overflow in moan_check_errorcopy().
1446 PH/23 POSIX allows open() to be a macro; guard against that.
1448 PH/24 If the recipient of an error message contained an @ in the local part
1449 (suitably quoted, of course), incorrect values were put in $domain and
1450 $local_part during the evaluation of errors_copy.
1456 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1457 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1460 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1461 issue a MAIL command.
1463 PH/02 In an ACL statement such as
1465 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1467 if a client was not listed at all, or was listed with a value other than
1468 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1469 the condition was not true (as it should be), so access was not denied.
1470 The bug was that the ! inversion was incorrectly passed on to the second
1471 item. This has been fixed.
1473 PH/03 Added additional dnslists conditions == and =& which are different from
1474 = and & when the dns lookup returns more than one IP address.
1476 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1477 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1479 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1480 FSYNC, which compiles an option called disable_fsync that allows for
1481 bypassing fsync(). The documentation is heavily laced with warnings.
1483 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1485 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1486 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1487 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1488 including adding "make clean"; (3) Added -fPIC when compiling the test
1489 dynamically loaded module, to get rid of a warning.
1491 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1492 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1493 = 0s. The bug is otherwise harmless.
1495 PH/07 There was a bug in the dovecot authenticator such that the value of
1496 $auth1 could be overwritten, and so not correctly preserved, after a
1497 successful authentication. This usually meant that the value preserved by
1498 the server_setid option was incorrect.
1500 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1502 PH/09 Installed PCRE release 7.0.
1504 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1505 run for batched SMTP input. It is now run at the start of every message
1506 in the batch. While fixing this I discovered that the process information
1507 (output by running exiwhat) was not always getting set for -bs and -bS
1508 input. This is fixed, and it now also says "batched" for BSMTP.
1510 PH/11 Added control=no_pipelining.
1512 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1513 patch, slightly modified), and move the expansion of helo_data till after
1514 the connection is made in the smtp transport (so it can use these
1517 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1519 PH/14 Added log_selector = +pid.
1521 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1523 PH/16 Add ${if forany and ${if forall.
1525 PH/17 Added dsn_from option to vary the From: line in DSNs.
1527 PH/18 Flush SMTP output before performing a callout, unless control =
1528 no_callout_flush is set.
1530 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1531 was true (the default) a successful delivery failed to delete the retry
1532 item, thus causing premature timeout of the address. The bug is now
1535 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1537 PH/21 Long custom messages for fakedefer and fakereject are now split up
1538 into multiline reponses in the same way that messages for "deny" and
1539 other ACL rejections are.
1541 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1542 with slight modification.
1544 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1545 draft, changing the syntax and factoring some duplicate code".
1547 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1548 for deliveries of the second and subsequent messages over the same SMTP
1551 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1552 ${reduce, with only minor "tidies".
1554 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1556 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1557 expansion side effects.
1559 PH/27 When a message times out after an over-quota error from an Exim-imposed
1560 quota, the bounce message says "mailbox is full". This message was not
1561 being given when it was a system quota that was exceeded. It now should
1564 MH/03 Made $recipients available in local_scan(). local_scan() already has
1565 better access to the recipient list through recipients_list[], but
1566 $recipients can be useful in postmaster-provided expansion strings.
1568 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1569 in the case of a MAIL command with additional options following the
1570 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1571 were accidentally chopped off.
1573 PH/29 SMTP synchronization checks are implemented when a command is read -
1574 there is a check that no more input is waiting when there shouldn't be
1575 any. However, for some commands, a delay in an ACL can mean that it is
1576 some time before the response is written. In this time, more input might
1577 arrive, invalidly. So now there are extra checks after an ACL has run for
1578 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1579 pipelining has not been advertised.
1581 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1583 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1584 returned in the SMTP response when smtp_return_error_details was false.
1585 This has been fixed.
1587 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1588 instead of the C I/O that was originally supplied, because problems were
1589 reported on Solaris.
1591 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1592 Exim which did not show up earlier: it was assuming that a call to
1593 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1594 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1595 was a macro that became an assignment, so it seemed to work. This has
1596 changed to a proper function call with a void return, hence the compile
1597 error. Exim's code has been fixed.
1599 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1602 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1604 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1606 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1607 because it turns out to be too big or there is a timeout) there is no
1608 "Completed" line in the log. When some messages of this type were
1609 selected by exigrep, they were listed as "not completed". Others were
1610 picked up by some special patterns. I have improved the selection
1611 criteria to be more general.
1613 PH/38 The host_find_failed option in the manualroute router can now be set
1614 to "ignore", to completely ignore a host whose IP address cannot be
1615 found. If all hosts are ignored, the behaviour is controlled by the new
1616 host_all_ignored option.
1618 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1619 homing or because of multiple MX records with /mx) generated more than
1620 one IP address, and the following item turned out to be the local host,
1621 all the secondary addresses of the first item were incorrectly removed
1622 from the list, along with the local host and any following hosts (which
1623 is what is supposed to happen).
1625 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1626 whoever called Exim into the -H file. In the case of the daemon it was
1627 behaving confusingly. When first started, it used values for whoever
1628 started the daemon, but after a SIGHUP it used the Exim user (because it
1629 calls itself on a restart). I have changed the code so that it now always
1632 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1633 message are rejected with the same error (e.g. no authentication or bad
1634 sender address), and a DATA command is nevertheless sent (as can happen
1635 with PIPELINING or a stupid MUA), the error message that was given to the
1636 RCPT commands is included in the rejection of the DATA command. This is
1637 intended to be helpful for MUAs that show only the final error to their
1640 PH/42 Another patch from the Sieve maintainer.
1642 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1643 Eximstats - Fixed some broken HTML links and added missing column headers
1645 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1646 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1648 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1650 PH/43 Yet another patch from the Sieve maintainer.
1652 PH/44 I found a way to check for a TCP/IP connection going away before sending
1653 the response to the final '.' that terminates a message, but only in the
1654 case where the client has not sent further data following the '.'
1655 (unfortunately, this is allowed). However, in many cases there won't be
1656 any further data because there won't be any more messages to send. A call
1657 to select() can be used: if it shows that the input is "ready", there is
1658 either input waiting, or the socket has been closed. An attempt to read
1659 the next input character can distinguish the two cases. Previously, Exim
1660 would have sent an OK response which the client would never have see.
1661 This could lead to message repetition. This fix should cure that, at
1662 least in a lot of common cases.
1664 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1665 advertised in response to EHLO.
1671 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1672 fixed by 4.65/MH/01 (is this a record?) are fixed:
1674 (i) An empty string was always treated as zero by the numeric comparison
1675 operators. This behaviour has been restored.
1677 (ii) It is documented that the numeric comparison operators always treat
1678 their arguments as decimal numbers. This was broken in that numbers
1679 starting with 0 were being interpreted as octal.
1681 While fixing these problems I realized that there was another issue that
1682 hadn't been noticed. Values of message_size_limit (both the global option
1683 and the transport option) were treated as octal if they started with 0.
1684 The documentation was vague. These values are now always treated as
1685 decimal, and I will make that clear in the documentation.
1691 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1692 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1695 MH/01 Don't check that the operands of numeric comparison operators are
1696 integers when their expansion is in "skipping" mode (fixes bug
1697 introduced by 4.64-PH/07).
1699 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1700 child addresses, Exim now panics and dies. Previously, because the count
1701 is held in a short int, deliveries were likely to be lost. As such a
1702 large number of recipients for a single message is ridiculous
1703 (performance will be very, very poor), I have chosen to impose a limit
1704 rather than extend the field.
1710 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1711 leftover -K file (the existence of which was triggered by #402).
1712 While we were at it, introduced process PID as part of the -K
1713 filename. This should rule out race conditions when creating
1716 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1717 processing considerably. Previous code took too long for large mails,
1718 triggering a timeout which in turn triggers #401.
1720 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1721 in the DK code in transports.c. sendfile() is not really portable,
1722 hence the _LINUX specificness.
1724 TF/01 In the add_headers option to the mail command in an Exim filter,
1725 there was a bug that Exim would claim a syntax error in any
1726 header after the first one which had an odd number of characters
1729 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1730 callout verification, Exim cached a "reject" for the entire domain. This
1731 is correct for most verifications, but it is not correct for a recipient
1732 verification with use_sender or use_postmaster set, because in that case
1733 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1734 case of MAIL FROM:<> rejection from other early rejections (e.g.
1735 rejection of HELO). When verifying a recipient using a non-null MAIL
1736 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1737 Whatever the result of the callout, the value of the domain cache is
1738 left unchanged (for any other kind of callout, getting as far as trying
1739 RCPT means that the domain itself is ok).
1741 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1744 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1745 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1746 session, the daemon ignores ECONNECT errors and logs others; it now
1747 ignores EPIPE as well.
1749 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1750 (quoted-printable decoding).
1752 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1753 later the small subsequent patch to fix an introduced bug.
1755 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1757 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1759 PH/08 An error is now given if message_size_limit is specified negative.
1761 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1762 to be given (somewhat) arbitrary names.
1764 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1767 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1768 miscellaneous code fixes
1770 PH/10 Added the log_reject_target ACL modifier to specify where to log
1773 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1774 hostname. This is wrong, because it relates to the incoming message (and
1775 probably the interface on which it is arriving) and not to the outgoing
1776 callout (which could be using a different interface). This has been
1777 changed to use the value of the helo_data option from the smtp transport
1778 instead - this is what is used when a message is actually being sent. If
1779 there is no remote transport (possible with a router that sets up host
1780 addresses), $smtp_active_hostname is used.
1782 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1783 tweaks were necessary in order to get it to work (see also 21 below):
1784 (a) The code assumed that strncpy() returns a negative number on buffer
1785 overflow, which isn't the case. Replaced with Exim's string_format()
1787 (b) There were several signed/unsigned issues. I just did the minimum
1788 hacking in of casts. There is scope for a larger refactoring.
1789 (c) The code used strcasecmp() which is not a standard C function.
1790 Replaced with Exim's strcmpic() function.
1791 (d) The code set only $1; it now sets $auth1 as well.
1792 (e) A simple test gave the error "authentication client didn't specify
1793 service in request". It would seem that Dovecot has changed its
1794 interface. Fortunately there's a specification; I followed it and
1795 changed what the client sends and it appears to be working now.
1797 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1800 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1801 address A is aliased to B and C, where B exists and C does not. Without
1802 -v the output is "A verified" because verification stops after a
1803 successful redirection if more than one address is generated. However,
1804 with -v the child addresses are also verified. Exim was outputting "A
1805 failed to verify" and then showing the successful verification for C,
1806 with its parentage. It now outputs "B failed to verify", showing B's
1807 parentage before showing the successful verification of C.
1809 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1810 look up a TXT record in a specific list after matching in a combined
1813 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1814 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1815 they consult the DNS. I had assumed they would set it the way they
1816 wanted; and indeed my experiments on Linux seem to show that in some
1817 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1818 To be on the safe side, however, I have now made the interface to
1819 host_find_byname() similar to host_find_bydns(), with an argument
1820 containing the DNS resolver options. The host_find_byname() function now
1821 sets these options at its start, just as host_find_bydns() does. The smtp
1822 transport options dns_qualify_single and dns_search_parents are passed to
1823 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1824 of host_find_byname() use the default settings of RES_DEFNAMES
1825 (qualify_single) but not RES_DNSRCH (search_parents).
1827 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1828 spool_read_header() do less string testing, by means of a preliminary
1829 switch on the second character of optional "-foo" lines. (This is
1830 overdue, caused by the large number of possibilities that now exist.
1831 Originally there were few.) While I was there, I also converted the
1832 str(n)cmp tests so they don't re-test the leading "-" and the first
1833 character, in the hope this might squeeze out yet more improvement.
1835 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1836 flag allowing group syntax was set by the header_syntax check but not
1837 turned off, possible causing trouble later; (2) The flag was not being
1838 set at all for the header_verify test, causing "group"-style headers to
1839 be rejected. I have now set it in this case, and also caused header_
1840 verify to ignore an empty address taken from a group. While doing this, I
1841 came across some other cases where the code for allowing group syntax
1842 while scanning a header line wasn't quite right (mostly, not resetting
1843 the flag correctly in the right place). These bugs could have caused
1844 trouble for malformed header lines. I hope it is now all correct.
1846 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1847 with the "reply" argument non-NULL. The code, however (which originally
1848 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1849 but it didn't always do it. This confused somebody who was copying the
1850 code for some other use. I have removed all the tests.
1852 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1853 feature that was used to support insecure browsers during the U.S. crypto
1854 embargo. It requires special client support, and Exim is probably the
1855 only MTA that supported it -- and would never use it because real RSA is
1856 always available. This code has been removed, because it had the bad
1857 effect of slowing Exim down by computing (never used) parameters for the
1858 RSA_EXPORT functionality.
1860 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1861 authenticator to fail if there's a tab character in the incoming data
1862 (there should never be unless someone is messing about, as it's supposed
1863 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1864 if the connection is using TLS or if the remote IP is the same as the
1865 local IP, and the "valid-client-cert option" if a client certificate has
1868 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1869 authenticators. This can be used for authorization after authentication
1870 succeeds. (In the case of plaintext, it servers for both authentication
1873 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1874 if any retry times were supplied.
1876 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1877 connection, where there is no client IP address to check. In this
1878 situation, the verify now always succeeds.
1880 PH/25 Applied John Jetmore's -Mset patch.
1882 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1884 PH/27 In a string expansion for a processed (not raw) header when multiple
1885 headers of the same name were present, leading whitespace was being
1886 removed from all of them, but trailing whitespace was being removed only
1887 from the last one. Now trailing whitespace is removed from each header
1888 before concatenation. Completely empty headers in a concatenation (as
1889 before) are ignored.
1891 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1892 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1894 PH/29 [Removed. This was a change that I later backed out, and forgot to
1895 correct the ChangeLog entry (that I had efficiently created) before
1896 committing the later change.]
1898 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1899 address errors (4xx response to RCPT) over the same connection as other
1900 messages routed to the same hosts. Such deliveries are always "forced",
1901 so retry times are not inspected. This resulted in far too many retries
1902 for the affected addresses. The effect occurred only when there were more
1903 hosts than the hosts_max_try setting in the smtp transport when it had
1904 the 4xx errors. Those hosts that it had tried were not added to the list
1905 of hosts for which the message was waiting, so if all were tried, there
1906 was no problem. Two fixes have been applied:
1908 (i) If there are any address or message errors in an SMTP delivery, none
1909 of the hosts (tried or untried) are now added to the list of hosts
1910 for which the message is waiting, so the message should not be a
1911 candidate for sending over the same connection that was used for a
1912 successful delivery of some other message. This seems entirely
1913 reasonable: after all the message is NOT "waiting for some host".
1914 This is so "obvious" that I'm not sure why it wasn't done
1915 previously. Hope I haven't missed anything, but it can't do any
1916 harm, as the worst effect is to miss an optimization.
1918 (ii) If, despite (i), such a delivery is accidentally attempted, the
1919 routing retry time is respected, so at least it doesn't keep
1920 hammering the server.
1922 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1923 in ${readsocket because some servers need this prod.
1925 PH/32 Added some extra debug output when updating a wait-xxx database.
1927 PH/33 The hint "could be header name not terminated by colon", which has been
1928 given for certain expansion errors for a long time, was not being given
1929 for the ${if def:h_colon_omitted{... case.
1931 PH/34 The spec says: "With one important exception, whenever a domain list is
1932 being scanned, $domain contains the subject domain." There was at least
1933 one case where this was not true.
1935 PH/35 The error "getsockname() failed: connection reset by peer" was being
1936 written to the panic log as well as the main log, but it isn't really
1937 panic-worthy as it just means the connection died rather early on. I have
1938 removed the panic log writing for the ECONNRESET error when getsockname()
1941 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1942 runs only) independently of the message's sender address. This meant
1943 that, if the 4xx error was in fact related to the sender, a different
1944 message to the same recipient with a different sender could confuse
1945 things. In particualar, this can happen when sending to a greylisting
1946 server, but other circumstances could also provoke similar problems.
1947 I have changed the default so that the retry time for these errors is now
1948 based a combination of the sender and recipient addresses. This change
1949 can be overridden by setting address_retry_include_sender=false in the
1952 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1953 remote server are returned as part of bounce messages. This was not
1954 happening for LMTP over a pipe (the lmtp transport), but now it is the
1955 same for both kinds of LMTP.
1957 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1958 in header lines that were in fact CNAMEs. This is no longer the case.
1960 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1961 and queue runs started by the daemon processed all messages. This has
1962 been fixed so that -R and -S can now usefully be given with -q<time>.
1964 PH/40 Import PCRE release 6.7 (fixes some bugs).
1966 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1968 PH/42 Give an error if -q is specified more than once.
1970 PH/43 Renamed the variables $interface_address and $interface_port as
1971 $received_ip_address and $received_port, to make it clear that these
1972 values apply to message reception, and not to the outgoing interface when
1973 a message is delivered. (The old names remain recognized, of course.)
1975 PH/44 There was no timeout on the connect() call when using a Unix domain
1976 socket in the ${readsocket expansion. There now is.
1978 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1979 be meaningful with "accept".
1981 SC/01 Eximstats V1.43
1982 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1984 SC/02 Eximstats V1.44
1985 Use a glob alias rather than an array ref in the generated
1986 parser. This improves both readability and performance.
1988 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1989 Collect SpamAssassin and rejection statistics.
1990 Don't display local sender or destination tables unless
1991 there is data to show.
1992 Added average volumes into the top table text output.
1994 SC/04 Eximstats V1.46
1995 Collect data on the number of addresses (recipients)
1996 as well as the number of messages.
1998 SC/05 Eximstats V1.47
1999 Added 'Message too big' to the list of mail rejection
2000 reasons (thanks to Marco Gaiarin).
2002 SC/06 Eximstats V1.48
2003 Mainlog lines which have GMT offsets and are too short to
2004 have a flag are now skipped.
2006 SC/07 Eximstats V1.49 (Alain Williams)
2007 Added the -emptyok flag.
2009 SC/08 Eximstats V1.50
2010 Fixes for obtaining the IP address from reject messages.
2012 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2013 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2014 whitesspace changes from 4.64-PH/27
2016 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2019 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2020 are found, allow negative numbers in numeric criteria)
2022 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2024 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2025 to match changes made in 4.64-PH/43
2027 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2029 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2030 "domain" is actually an IP address, and force a failure. This locks out
2031 those revolvers/nameservers that support "A-for-A" lookups, in
2032 contravention of the specifications.
2034 PH/48 When a host name was looked up from an IP address, and the subsequent
2035 forward lookup of the name timed out, the host name was left in
2036 $sender_host_name, contrary to the specification.
2038 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2039 restricted to single-key lookups, Exim was not diagnosing an error if
2040 * or *@ was used with a query-style lookup.
2042 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2044 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2045 long ago, but noone interested enough thought of it. Let's just say that
2046 the "1.1" means that there are some new functions that weren't there at
2047 some point in the past.
2049 PH/51 Error processing for expansion failure of helo_data from an smtp
2050 transport during callout processing was broken.
2052 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2053 tested/used via the -bh/-bhc/-bs options.
2055 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2056 bug, fixed in subsequent PCRE releases).
2058 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2059 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2061 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2067 SC/01 Use a glob alias rather than an array ref in eximstats generated
2068 parser. This improves both readability and performance.
2070 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2071 Don't display local sender or destination tables in eximstats unless
2072 there is data to show.
2073 Added average volumes into the eximstats top table text output.
2075 SC/03 Collect data on the number of addresses (recipients) as well
2076 as the number of messages in eximstats.
2078 TF/01 Correct an error in the documentation for the redirect router. Exim
2079 does (usually) call initgroups() when daemonizing.
2081 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2082 with consistent privilege compared to when running as a daemon.
2084 TF/03 Note in the spec that $authenticated_id is not set for local
2085 submissions from trusted users.
2087 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2088 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2090 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2091 by adding some example configuration directives to the default
2092 configuration file. A little bit of work is required to uncomment the
2093 directives and define how usernames and passwords are checked, but
2094 there is now a framework to start from.
2096 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2097 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2098 without this. I don't know how relevant this is to other LDAP libraries.
2100 PH/02 Add the verb name to the "unknown ACL verb" error.
2102 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2104 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2106 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2107 directory not expanded when it should be if an expanded home directory
2108 was set for the address (which is overridden by the transport).
2110 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2113 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2114 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2115 because it is too late at that time, and has no effect.
2117 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2118 security issue with \' (bugzilla #107). I could not use the
2119 PQescapeStringConn() function, because it needs a PGconn value as one of
2122 PH/08 When testing addresses using -bt, indicate those final addresses that
2123 are duplicates that would not cause an additional delivery. At least one
2124 person was confused, thinking that -bt output corresponded to deliveries.
2125 (Suppressing duplicates isn't a good idea as you lose the information
2126 about possibly different redirections that led to the duplicates.)
2128 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2129 systems where poll() doesn't work, in particular OS X.
2131 PH/10 Added more information to debugging output for retry time not reached.
2133 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2134 operations in malware.c.
2136 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2139 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2140 syslog_duplication set false, log lines that would normally be written
2141 both the the main log and to the reject log were not written to syslog at
2144 PH/14 In the default configuration, change the use of "message" in ACL warn
2145 statements to "add_header".
2147 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2148 not followed by a command (e.g. "seen endif").
2150 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2151 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2154 PH/17 Added extra conditions to the default value of delay_warning_condition
2158 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2159 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2160 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2163 The Auto-Submitted: and various List- headers are standardised, whereas I
2164 don't think Precedence: ever was.
2166 PH/18 Refactored debugging code in route_finduser() to show more information,
2167 in particular, the error code if getpwnam() issues one.
2169 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2170 This is apparently needed in addition to the PH/07 change above to avoid
2171 any possible encoding problems.
2173 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2174 but not after initializing Perl.
2176 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2177 output them only if debugging. By default they are written stderr,
2178 apparently, which is not desirable.
2180 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2183 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2186 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2188 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2189 authenticated or an ident call has been made. Suppress the default
2190 values for $authenticated_id and $authenticated_sender (but permit -oMai
2191 and -oMas) when testing with -bh.
2193 PH/24 Re-jigged the order of the tests in the default configuration so that the
2194 tests for valid domains and recipients precede the DNS black list and CSA
2195 tests, on the grounds that those ones are more expensive.
2197 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2198 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2199 This bug exists in every version of Exim that I still have, right back to
2202 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2203 However, an attempt to turn on case-sensitivity in a regex key by
2204 including (?-i) didn't work because the subject string was already
2205 lowercased, and the effects were non-intuitive. It turns out that a
2206 one-line patch can be used to allow (?-i) to work as expected.
2212 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2213 other effects) broke the use of negated acl sub-conditions.
2215 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2218 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2219 "Deny" causes Exim to reject the incoming connection with a 554 error.
2220 Unfortunately, if there is a major crisis, such as a disk failure,
2221 tcp-wrappers gives "deny", whereas what one would like would be some
2222 kind of temporary error. A kludge has been added to help with this.
2223 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2224 554 error is used if errno is still zero or contains ENOENT (which occurs
2225 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2228 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2230 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2231 errors. Otherwise a message that provokes a temporary error (when other
2232 messages do not) can cause a whole host to time out.
2234 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2235 addresses were routed directly to files or pipes from a redirect router.
2236 File deliveries just didn't batch; pipe deliveries might have suffered
2239 PH/06 A failure to get a lock for a hints database would erroneously always say
2240 "Failed to get write lock", even when it was really a read lock.
2242 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2243 of 0600. This has been changed to use the value of the lockfile_mode
2244 option (which defaults to 0600).
2246 PH/08 Applied small patch from the Sieve maintainer.
2248 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2249 folder from quota calculations, a direct delivery into this folder messed
2250 up the contents of the maildirsize file. This was because the regex was
2251 used only to exclude .Trash (or whatever) when the size of the mailbox
2252 was calculated. There was no check that a delivery was happening into an
2253 excluded directory. This bug has been fixed by ignoring all quota
2254 processing for deliveries into excluded directories.
2256 PH/10 Added the maildirfolder_create_regex option to appendfile.
2262 PH/01 The code for finding all the local interface addresses on a FreeBSD
2263 system running IPv6 was broken. This may well have applied to all BSD
2264 systems, as well as to others that have similar system calls. The broken
2265 code found IPv4 interfaces correctly, but gave incorrect values for the
2266 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2267 that it would not match correctly against @[] and not recognize the IPv6
2270 PH/02 The ipliteral router was not recognizing addresses of the form user@
2271 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2273 PH/03 Added disable_ipv6.
2275 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2276 decoded form, because it is most often used to construct To: headers
2277 lines in autoreplies, and the decoded form may well be syntactically
2278 invalid. However, $reply_address has leading white space removed, and all
2279 newlines turned into spaces so that the autoreply transport does not
2282 PH/05 If group was specified without a user on a router, and no group or user
2283 was specified on a transport, the group from the router was ignored.
2285 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2286 for visible compile-time settings that can be used to change these
2287 numbers, for those that want even more. Backwards compatibility with old
2288 spool files has been maintained. However, going back to a previous Exim
2289 release will lost any variables that are in spool files.
2291 PH/07 Two small changes when running in the test harness: increase delay when
2292 passing a TCP/IP connection to a new process, in case the original
2293 process has to generate a bounce, and remove special handling of
2294 127.0.0.2 (sic), which is no longer necessary.
2296 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2297 be the same on different OS.
2299 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2302 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2303 whether --show-vars was specified or not
2305 JJ/02 exipick: Added support for new ACL variable spool format introduced
2308 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2309 syntactically invalid From: or Reply-to: line, and a filter used this to
2310 generate an autoreply, and therefore failed to obtain an address for the
2311 autoreply, Exim could try to deliver to a non-existent relative file
2312 name, causing unrelated and misleading errors. What now happens is that
2313 it logs this as a hard delivery error, but does not attempt to create a
2316 PH/11 The exinext utility has a -C option for testing purposes, but although
2317 the given file was scanned by exinext itself; it wasn't being passed on
2318 when Exim was called.
2320 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2321 an end-of-file indication when reading a command response.
2323 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2324 compiled. In many other places in Exim, IPv6 addresses are always
2325 recognized, so I have changed this. It also means that IPv4 domain
2326 literals of the form [IPV4:n.n.n.n] are now always recognized.
2328 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2329 used if the router is not running as root, for example, when verifying at
2330 ACL time, or when using -bh. The debugging output from this situation was
2331 non-existent - all you got was a failure to exec. I have made two
2334 (a) Failures to set uid/gid, the current directory, or a process leader
2335 in a subprocess such as that created by queryprogram now generate
2336 suitable debugging ouput when -d is set.
2338 (b) The queryprogram router detects when it is not running as root,
2339 outputs suitable debugging information if -d is set, and then runs
2340 the subprocess without attempting to change uid/gid.
2342 PH/15 Minor change to Makefile for building test_host (undocumented testing
2345 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2346 additional section of a DNS packet that returns MX or SRV records.
2347 Instead, it always explicitly searches for A/AAAA records. This avoids
2348 major problems that occur when a DNS server includes only records of one
2349 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2350 fixed another bug: if SRV records were looked up and the corresponding
2351 address records were *not* found in the additional section, the port
2352 values from the SRV records were lost.
2354 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2355 using the correct key (the original address) when searching the retry
2356 rules in order to find which one to use for generating the retry hint.
2358 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2359 adding the default one. Similarly, if it contains a Reply-To: header, the
2360 errors_reply_to option, if set, is not used.
2362 PH/19 When calculating a retry time, Exim used to measure the "time since
2363 failure" by looking at the "first failed" field in the retry record. Now
2364 it does not use this if it is later than than the arrival time of the
2365 message. Instead it uses the arrival time. This makes for better
2366 behaviour in cases where some deliveries succeed, thus re-setting the
2367 "first failed" field. An example is a quota failure for a huge message
2368 when small messages continue to be delivered. Without this change, the
2369 "time since failure" will always be short, possible causing more frequent
2370 delivery attempts for the huge message than are intended.
2371 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2373 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2374 $1, $2, $3) because the numerical variables can be reset during some
2375 expansion items (e.g. "match"), thereby losing the authentication data.
2377 PH/21 Make -bV show the size of off_t variables so that the test suite can
2378 decide whether to run tests for quotas > 2G.
2380 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2381 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2382 filecount value is greater than 2G or if a quota value is greater than 2G
2383 on a system where the size of off_t is not greater than 4, a panic error
2386 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2387 never match. The debug and -bh output now contains an explicit error
2388 message indicating a malformed IPv4 address or mask.
2390 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2391 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2392 PH/23 above applies.
2394 PH/25 Do not write to syslog when running in the test harness. The only
2395 occasion when this arises is a failure to open the main or panic logs
2396 (for which there is an explicit test).
2398 PH/26 Added the /no_tell option to "control=freeze".
2400 PH/27 If a host name lookup failed very early in a connection, for example, if
2401 the IP address matched host_lookup and the reverse lookup yielded a name
2402 that did not have a forward lookup, an error message of the form "no IP
2403 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2404 could be logged. Now it outputs the IP address instead of "NULL".
2406 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2407 allows the sender and the authenticated sender to be set when
2408 submitting a message from within Exim. Since child_open_exim() is
2409 documented for local_scan(), the new function should be too.
2411 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2412 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2413 results in an empty string is now treated as unset.
2415 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2417 PH/31 Added sender_verify_fail logging option.
2419 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2420 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2421 tidied the source and removed it altogether.
2423 PH/33 When a queue run was abandoned because the load average was too high, a
2424 log line was always written; now it is written only if the queue_run log
2425 selector is set. In addition, the log line for abandonment now contains
2426 information about the queue run such as the pid. This is always present
2427 in "start" and "stop" lines but was omitted from the "abandon" line.
2429 PH/34 Omit spaces between a header name and the colon in the error message that
2430 is given when verify = headers_syntax fails (if there are lots of them,
2431 the message gets confusing).
2433 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2434 names, as there are now some PTR records that contain slashes. This check
2435 is only to protect against broken name servers that fall over on strange
2436 characters, so the fact that it applies to all lookups doesn't matter.
2438 PH/36 Now that the new test suite is complete, we can remove some of the
2439 special code in Exim that was needed for the old test suite. For example,
2440 sorting DNS records because real resolvers return them in an arbitrary
2441 order. The new test suite's fake resolver always returns records in the
2444 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2445 bounces) except when queue_only is set, to avoid logging races between
2446 the different processes.
2448 PH/38 Panic-die if .include specifies a non-absolute path.
2450 PH/39 A tweak to the "H" retry rule from its user.
2452 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2453 a label. They prevented compilation on older perls.
2455 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2456 a warning to be raised on newish perls.
2458 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2459 on queue. Changes to match documented behaviour of showing count of
2460 messages matching specified criteria.
2462 PH/40 Changed the default ident timeout from 30s to 5s.
2464 PH/41 Added support for the use of login_cap features, on those BSD systems
2465 that have them, for controlling the resources used by pipe deliveries.
2467 PH/42 The content-scanning code uses fopen() to create files in which to put
2468 message data. Previously it was not paying any attention to the mode of
2469 the files. Exim runs with umask(0) because the rest of the code creates
2470 files with open(), and sets the required mode explicitly. Thus, these
2471 files were ending up world-writeable. This was not a big issue, because,
2472 being within the spool directory, they were not world-accessible. I have
2473 created a function called modefopen, which takes an additional mode
2474 argument. It sets umask(777), creates the file, chmods it to the required
2475 mode, then resets the umask. All the relevant calls to fopen() in the
2476 content scanning code have been changed to use this function.
2478 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2479 to 24 hours. This avoids potential overflow problems when processing G
2480 and H retry rules. I suspect nobody ever tinkers with this value.
2482 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2484 PH/45 When the plaintext authenticator is running as a client, the server's
2485 challenges are checked to ensure they are valid base64 strings. By
2486 default, the authentication attempt is cancelled if an invalid string is
2487 received. Setting client_ignore_invalid_base64 true ignores these errors.
2488 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2489 they are received. Thus, the responses can be made to depend on the
2490 challenges. If an invalid string is ignored, an empty string is placed in
2493 PH/46 Messages that are created by the autoreply transport now contains a
2494 References: header, in accordance with RFCs 2822 and 3834.
2496 PH/47 Added authenticated_sender_force to the smtp transport.
2498 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2500 PH/49 Installed latest patch from the Sieve maintainer.
2502 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2503 was also set, the appendfile transport was unnecessarily scanning a
2504 directory of message files (e.g. for maildir delivery) to find the count
2505 of files (along with the size), even though it did not need this
2506 information. It now does the scan only if it needs to find either the
2507 size of the count of files.
2509 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2511 PH/52 Two bugs concerned with error handling when the smtp transport is
2514 (i) Exim was not creating retry information for temporary errors given
2515 for individual recipients after the DATA command when the smtp transport
2516 was used in LMTP mode. This meant that they could be retried too
2517 frequently, and not timed out correctly.
2519 (ii) Exim was setting the flag that allows error details to be returned
2520 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2521 recipients that were returned after the DATA command.
2523 PH/53 This is related to PH/52, but is more general: for any failing address,
2524 when detailed error information was permitted to be returned to the
2525 sender, but the error was temporary, then after the final timeout, only
2526 "retry timeout exceeded" was returned. Now it returns the full error as
2527 well as "retry timeout exceeded".
2529 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2530 do this, and (what is worse) MTAs that accept it.
2532 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2533 will now be deprecated.
2535 PH/56 New os.c-cygwin from the Cygwin maintainer.
2537 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2538 formats (previously only available in exim formats via -bpr, -bpru,
2539 and -bpra. Now also available in native and exiqgrep formats)
2541 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2542 with very large, slow to parse queues
2544 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2546 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2548 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2549 responses to authentication challenges, though it was showing the
2550 challenges; (ii) I've removed the CR characters from the debug output for
2553 PH/58 Allow for the insertion of a newline as well as a space when a string
2554 is turned into more than one encoded-word during RFC 2047 encoding. The
2555 Sieve code now uses this.
2557 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2558 data_4xx, lost_connection, tls_required.
2560 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2561 message was being sent as an SMTP response.
2563 PH/61 Add -l and -k options to exicyclog.
2565 PH/62 When verifying, if an address was redirected to one new address, so that
2566 verification continued, and the new address failed or deferred after
2567 having set something in $address_data, the value of $address_data was not
2568 passed back to the ACL. This was different to the case when no
2569 redirection occurred. The value is now passed back in both cases.
2571 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2572 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2573 use login_cap.h, so on its own it isn't the distinguishing feature. The
2574 new name refers directly to the setclassresources() function.
2576 PH/65 Added configuration files for NetBSD3.
2578 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2580 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2581 is preferred over IPv4.
2583 PH/68 The bounce_return_message and bounce_return_body options were not being
2584 honoured for bounces generated during the reception of non-SMTP messages.
2585 In particular, this applied to messages rejected by the ACL. This bug has
2586 been fixed. However, if bounce_return_message is true and bounce_return_
2587 body is false, the headers that are returned for a non-SMTP message
2588 include only those that have been read before the error was detected.
2589 (In the case of an ACL rejection, they have all been read.)
2591 PH/69 The HTML version of the specification is now built in a directory called
2592 spec_html instead of spec.html, because the latter looks like a path with
2593 a MIME-type, and this confuses some software.
2595 PH/70 Catch two compiler warnings in sieve.c.
2597 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2598 function verify_get_ident() calls ip_connect() to connect a socket, but
2599 if the "connect()" function timed out, ip_connect() used to close the
2600 socket. However, verify_get_ident() also closes the socket later, and in
2601 between Exim writes to the log, which may get opened at this point. When
2602 the socket was closed in ip_connect(), the log could get the same file
2603 descriptor number as the socket. This naturally causes chaos. The fix is
2604 not to close the socket in ip_connect(); the socket should be closed by
2605 the function that creates it. There was only one place in the code where
2606 this was missing, in the iplookup router, which I don't think anybody now
2607 uses, but I've fixed it anyway.
2609 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2610 well as to direct DNS lookups. Otherwise the handling of names in host
2611 lists is inconsistent and therefore confusing.
2617 PH/01 Two changes to the default runtime configuration:
2619 (1) Move the checks for relay_from_hosts and authenticated clients from
2620 after to before the (commented out) DNS black list checks.
2622 (2) Add control=submission to the relay_from_hosts and authenticated
2623 clients checks, on the grounds that messages accepted by these
2624 statements are most likely to be submissions.
2626 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2628 (1) Generate an error if the third argument for the ${prvs expansion is
2631 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2634 (3) Reset the variables that are obtained from the first argument of
2635 ${prvscheck and used in the second argument before leaving the code,
2636 because their memory is reclaimed, so using them afterwards may do
2639 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2640 one (it's much easier than Tom thought :-).
2642 (5) Because of (4), we can now allow for the use of $prvscheck_result
2643 inside the third argument.
2645 PH/03 For some reason, the default setting of PATH when running a command from
2646 a pipe transport was just "/usr/bin". I have changed it to
2649 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2650 anything to be listed in the output from -bV.
2652 PH/05 When a filter generated an autoreply, the entire To: header line was
2653 quoted in the delivery log line, like this:
2655 => >A.N.Other <ano@some.domain> <original@ddress> ...
2657 This has been changed so that it extracts the operative address. There
2658 may be more than one such address. If so, they are comma-separated, like
2661 => >ano@some.domain,ona@other.domain <original@ddress> ...
2663 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2664 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2665 not being looked up in the rDNS to get a host name, Exim was showing the
2666 IP address twice in Received: lines, even though the IP addresses were
2667 identical. For example:
2669 Received: from [1.2.3.4] (helo=[1.2.3.4])
2671 However, if the real host name was known, it was omitting the HELO data
2672 if it matched the actual IP address. This has been tidied up so that it
2673 doesn't show the same IP address twice.
2675 PH/07 When both +timestamp and +memory debugging was on, the value given by
2676 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2677 called by the debug printing, thereby overwriting the timestamp buffer.
2678 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2680 PH/08 When the original message was included in an autoreply transport, it
2681 always said "this is a copy of the message, including all the headers",
2682 even if body_only or headers_only was set. It now gives an appropriate
2685 PH/09 Applied a patch from the Sieve maintainer which:
2687 o fixes some comments
2688 o adds the (disabled) notify extension core
2689 o adds some debug output for the result of if/elsif tests
2690 o points to the current vacation draft in the documentation
2691 and documents the missing references header update
2695 o fixes a bug in processing the envelope test (when testing
2696 multiple envelope elements, the last element determinted the
2699 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2700 Electronic Mail") by including:
2702 Auto-submitted: auto-generated
2704 in the messages that it generates (bounce messages and others, such as
2705 warnings). In the case of bounce messages for non-SMTP mesages, there was
2706 also a typo: it was using "Auto_submitted" (underscore instead of
2707 hyphen). Since every message generated by Exim is necessarily in response
2708 to another message, thes have all been changed to:
2710 Auto-Submitted: auto-replied
2712 in accordance with these statements in the RFC:
2714 The auto-replied keyword:
2716 - SHOULD be used on messages sent in direct response to another
2717 message by an automatic process,
2719 - MUST NOT be used on manually-generated messages,
2721 - MAY be used on Delivery Status Notifications (DSNs) and Message
2722 Disposition Notifications (MDNs),
2724 - MUST NOT be used on messages generated by automatic or periodic
2725 processes, except for messages which are automatic responses to
2728 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2729 to the default Received: header definition.
2731 PH/12 Added log selector acl_warn_skipped (default on).
2733 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2734 variables because (a) they are in the wrong storage pool and (b) even if
2735 they were copied, it wouldn't work properly because of the caching.
2737 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2738 checking when decoding. Apparently there are clients that generate
2739 overlong encoded strings. Why am I not surprised?
2741 PH/15 If the first argument of "${if match_address" was not empty, but did not
2742 contain an "@" character, Exim crashed. Now it writes a panic log message
2743 and treats the condition as false.
2745 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2747 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2748 extension "envelope-auth". The code is finished and in agreement with
2749 other implementations, but there is no documentation so far and in fact,
2750 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2751 not changing the active code.
2753 Print executed "if" and "elsif" statements when debugging is used. This
2754 helps a great deal to understand what a filter does.
2756 Document more things not specified clearly in RFC3028. I had all this
2757 sorted out, when out of a sudden new issues came to my mind. Oops."
2759 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2762 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2763 canonical form (as documented). However, after a host name lookup from
2764 the IP address, check_host() was doing a simple string comparison with
2765 addresses acquired from the DNS when checking that the found name did
2766 have the original IP as one of its addresses. Since any found IPv6
2767 addresses are likely to be in abbreviated form, the comparison could
2768 fail. Luckily, there already exists a function for doing the comparison
2769 by converting both addresses to binary, so now that is used instead of
2770 the text comparison.
2772 PH/20 There was another similar case to PH/19, when a complete host name was
2773 given in a host list; looking up its IP address could give an abbreviated
2774 form, whereas the current host's name might or might not be abbreviated.
2775 The same fix has been applied.
2781 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2782 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2785 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2786 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2788 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2790 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2791 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2792 Also, for "Auto-Submitted", treat anything other than "no" as
2793 non-personal, in accordance with RFC 3834. (Previously it treated
2794 anything starting "auto-" as non-personal.)
2796 TF/01 The control=submission/name=... option had a problem with syntax
2797 errors if the name included a slash character. The /name= option
2798 now slurps the rest of the string, so it can include any characters
2799 but it must come last in the list of options (after /sender_retain
2802 PH/05 Some modifications to the interface to the fake nameserver for the new
2810 TK/01 Added the "success_on_redirect" address verification option. See
2811 NewStuff for rationale and an example.
2813 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2815 PH/02 Patch to exigrep to allow it to work on syslog lines.
2817 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2818 fread() to skip over the body file's header line, because in Cygwin the
2819 header line is locked and is inaccessible.
2821 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2822 co-exist for some time) to make it clear that it is the Exim ID that is
2823 referenced, not the Message-ID: header line.
2825 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2826 string_format() function, because snprintf() does not exist on all
2829 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2830 ${stat: expansion item.
2832 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2833 protocol synchronization error", to keep the pedants happy.
2835 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2836 well as for IRIX systems, when gcc is being used. See the host.c source
2839 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2841 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2844 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2845 transport and to the smtp transport in LMTP mode.
2847 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2849 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2850 run a filter in a subprocess. This could lead to confusion in subsequent
2851 lookups in the parent process. There should also be a search_tidyup() at
2852 the end of the subprocess.
2854 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2855 only if the host matched helo_try_verify_hosts, which caused the
2856 verification to occur when the EHLO/HELO command was issued. The ACL just
2857 tested the remembered result. Now, if a previous verification attempt has
2858 not happened, "verify = helo" does it there and then.
2860 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2862 TK/03 Fix log output including CR from clamd.
2864 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2865 exist provoked a memory error which could cause a segfault.
2867 PH/15 Installed PCRE 6.2
2869 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2871 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2872 of the problem. Specifically, suggested +O2 rather than +O1 for the
2875 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2877 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2878 record, the port was not correctly logged when the outgoing_port log
2879 selector was set (it logged the transort's default port).
2881 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2882 fallback_hosts, and "hosts" in the smtp transport.
2884 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2885 host errors such as "Connection refused".
2887 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2888 authentication with radiusclient 0.4.9:
2890 - Error returned from rc_read_config was caught wrongly
2891 - Username/password not passed on to radius server due to wrong length.
2893 The presumption is that some radiusclient API changes for 4.51/PH/17
2894 were not taken care of correctly. The code is still untested by me (my
2895 Linux distribution still has 0.3.2 of radiusclient), but it was
2896 contributed by a Radius user.
2898 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2899 expanding the "port" option of the smtp transport.
2901 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2902 while reading a MIME header. Thanks to Tom Hughes for a patch.
2904 PH/24 Include config.h inside local_scan.h so that configuration settings are
2907 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2908 that in an ACL for RCPT (for example), you can examine exactly what was
2911 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2912 commands, but it was not correctly comparing the address with the actual
2913 client host address. Thus, it would show the EHLO address in Received:
2914 header lines when this was not necessary.
2916 PH/27 Added the % operator to ${eval:}.
2918 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2919 it should be ignoring failures (because with -C, for example, it has lost
2920 privilege). It wasn't ignoring creation failures other than "already
2923 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2926 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2927 because an input line was too long, either on its own, or by virtue of
2928 too many continuations, the temporary file was not being removed, and the
2929 return code was incorrect.
2931 PH/31 Missing "BOOL" in function definition in filtertest.c.
2933 PH/32 Applied Sieve patches from the maintainer.
2935 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2937 PH/33 Added "verify = not_blind".
2939 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2940 Local/Makefile (with some defaults set). These are used in built scripts
2941 such as exicyclog, but they have never been used in the exim_install
2942 script (though there are many overriding facilities there). I have
2943 arranged that the exim_install script now takes note of these two
2946 PH/35 Installed configuration files for Dragonfly.
2948 PH/36 When a locally submitted message by a trusted user did not contain a
2949 From: header, and the sender address was obtained from -f or from an SMTP
2950 MAIL command, and the trusted user did not use -F to supply a sender
2951 name, $originator_name was incorrectly used when constructing a From:
2952 header. Furthermore, $originator_name was used for submission mode
2953 messages from external hosts without From: headers in a similar way,
2954 which is clearly wrong.
2956 PH/37 Added control=suppress_local_fixups.
2958 PH/38 When log_selector = +received_sender was set, and the addition of the
2959 sender made the log line's construction buffer exactly full, or one byte
2960 less than full, an overflow happened when the terminating "\n" was
2963 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2964 when the result of a list match is failure because a DNS lookup failed.
2966 PH/40 RM_COMMAND is now used in the building process.
2968 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2969 the "build-* directories that it finds.
2971 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2972 address in a domain literal was a prefix of an interface address.
2974 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2975 when verifying a sender address, unless rewrite_headers is false.
2977 PH/44 Wrote a long comment about why errors_to addresses are verified as
2978 recipients, not senders.
2980 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2981 the ratelimit ACL was added.
2983 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2985 PH/46 Added extra information about PostgreSQL errors to the error string.
2987 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2988 avoiding the need to install special zones in a real server. This is
2989 backwards compatible; if it can't find the fake resolver, it drops back.
2990 Thus, both old and new test suites can be run.
2992 TF/02 Added util/ratelimit.pl
2994 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2995 clock is set back in time.
2997 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2998 Candler <B.Candler@pobox.com>.
3000 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3001 OK for addresses that are the result of redirections.
3003 PH/48 A number of further additions for the benefit of the new test suite,
3004 including a fake gethostbyname() that interfaces to the fake DNS resolver
3007 TF/06 The fix for widen_domains has also been applied to qualify_single and
3008 search_parents which are the other dnslookup options that can cause
3011 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3014 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3016 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3017 service" if it encountered an MX record with an empty target hostname.
3018 The message is now "an MX or SRV record indicated no SMTP service".
3020 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3021 if the DNS of the sending site is misconfigured. This is quite a
3022 common situation. This change restores the behaviour of treating a
3023 helo verification defer as a failure.
3025 PH/51 If self=fail was set on a router, the bounce message did not include the
3026 actual error message.
3032 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3034 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3035 command itself ended in error, the underlying message about the transport
3036 filter timeout was being overwritten with the pipe command error. Now the
3037 underlying error message should be appended to the second error message.
3039 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3041 PH/02 Reduce the amount of output that "make" produces by default. Full output
3042 can still be requested.
3044 PH/03 The warning log line about a condition test deferring for a "warn" verb
3045 was being output only once per connection, rather than after each
3046 occurrence (because it was using the same function as for successful
3047 "warn" verbs). This seems wrong, so I have changed it.
3049 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3050 it should not have, which might have caused a crash in the right
3051 circumstances, but probably never did.
3053 PH/04 Installed a modified version of Tony Finch's patch to make submission
3054 mode fix the return path as well as the Sender: header line, and to
3055 add a /name= option so that you can make the user's friendly name appear
3058 TF/03 Added the control = fakedefer ACL modifier.
3060 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3061 Mark Lowes for thorough testing.
3063 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3065 TK/03 Merged latest SRS patch from Miles Wilton.
3067 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3068 with the definition in sysexits.h (which is #included earlier).
3069 Fortunately, Exim does not actually use EX_OK. The code used to try to
3070 preserve the sysexits.h value, by assumimg that macro definitions were
3071 scanned for macro replacements. I have been disabused of this notion,
3072 so now the code just undefines EX_OK before #including unistd.h.
3074 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3075 in the smtp transport. When a block could not be written in a single
3076 write() function, the timeout was being re-applied to each part-write.
3077 This seems wrong - if the receiver was accepting one byte at a time it
3078 would take for ever. The timeout is now adjusted when this happens. It
3079 doesn't have to be particularly precise.
3081 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3082 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3084 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3085 without a domain if the check to <postmaster@domain> fails.
3087 SC/01 Eximstats: added -xls and the ability to specify output files
3088 (patch written by Frank Heydlauf).
3090 SC/02 Eximstats: use FileHandles for outputing results.
3092 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3094 SC/04 Eximstats: fixed display of large numbers with -nvr option
3096 SC/05 Eximstats: fixed merging of reports with empty tables.
3098 SC/06 Eximstats: added the -include_original_destination flag
3100 SC/07 Eximstats: removed tabs and trailing whitespace.
3102 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3104 TK/06 MBOX spool code: Add real "From " MBOX separator line
3105 so the .eml file is really in mbox format (even though
3106 most programs do not really care). Patch from Alex Miller.
3108 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3109 The latter is generated from $received_to and is only set if the
3110 message has one envelope recipient. SA can use these headers,
3111 obviously out-of-the-box. Patch from Alex Miller.
3113 PH/08 The ${def test on a variable was returning false if the variable's
3114 value was "0", contrary to what the specification has always said!
3115 The result should be true unless the variable is empty.
3117 PH/09 The syntax error of a character other than { following "${if
3118 def:variable_name" (after optional whitespace) was not being diagnosed.
3119 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3120 accidental colon was present, for example, could give incorrect results.
3122 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3123 result is used (not including appendfile, where other changes are about
3126 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3127 This involved changing a lot of size variables from int to off_t. It
3128 should work with maildirs and everything.
3130 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3131 spamd dying while we are connected to it.
3133 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3136 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3137 function for BDB 4.3.
3139 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3141 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3142 the sources for PCRE have been re-arranged and more files are now
3145 PH/15 The code I had for printing potentially long long variables in PH/11
3146 above was not the best (it lost precision). The length of off_t variables
3147 is now inspected at build time, and an appropriate printing format (%ld
3148 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3149 to be "long long int" or "long int". This is needed for the internal
3150 formatting function string_vformat().
3152 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3153 the configuration file to be ":syslog", then the script "guesses" where
3154 the logs files are, rather than using the compiled in default. In our
3155 case the guess is not the same as the compiled default, so the script
3156 suddenly stopped working when I started to use syslog. The patch checks
3157 to see if log_file_path is "". If so, it attempts to read it from exim
3158 with no configuration file to get the compiled in version, before it
3159 falls back to the previous guessing code."
3161 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3162 implementing BATV in an Exim configuration. See NewStuff for the gory
3165 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3166 Makefile that are specific to HP-UX.
3168 PH/18 If the "use_postmaster" option was set for a recipient callout together
3169 with the "random" option, the postmaster address was used as the MAIL
3170 FROM address for the random test, but not for the subsequent recipient
3171 test. It is now used for both.
3173 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3174 patch removes a few documentation additions to RFC 3028, because the
3175 latest draft now contains them. It adds the new en;ascii-case comparator
3176 and a new error check for 8bit text in MIME parts. Comparator and
3177 require names are now matched exactly. I enabled the subaddress
3178 extension, but it is not well tested yet (read: it works for me)."
3180 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3181 rework some of the code of TK/09 above to avoid the hardwired use of
3182 "%lld" and "long long". Replaced the call to snprintf() with a call to
3185 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3186 records point to non-existent hosts", "retry timeout exceeded", and
3187 "retry time not reached for any host after a long failure period".
3189 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3190 experimental DomainKeys support:
3192 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3193 (2) On an error such as an illegally used "control", the wrong name for
3194 the control was given.
3196 These problems did NOT occur unless DomainKeys support was compiled.
3198 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3200 PH/24 Added ${if match_ip condition.
3202 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3203 databases so that it will be absolutely obvious if a crash occurs in the
3204 DB library. This is a regular occurrence (often caused by mis-matched
3207 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3208 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3209 picked up on a user's system that detects such things. There doesn't seem
3210 to be a gcc warning option for this - only an attribute that has to be
3211 put on the function's prototype. It seems that in Fedora Core 4 they have
3212 set this on a number of new functions. No doubt there will be more in due
3215 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3216 specify a transport. However, if an address that was verified by such a
3217 router was the subject of a callout, Exim crashed because it tried to
3218 read the rcpt_include_affixes from the non-existent transport. Now it
3219 just assumes that the setting of that option is false. This bug was
3220 introduced by 4.51/PH/31.
3222 PH/28 Changed -d+all to exclude +memory, because that information is very
3223 rarely of interest, but it makes the output a lot bigger. People tend to
3224 do -d+all out of habit.
3226 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3227 code in os-type was giving problems when libc.so lives in lib64, like on
3230 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3231 aren't the modern standard, and it seems that some systems' include files
3232 don't always have them. Exim was already checking for some of the newer
3233 ones like T_AAAA, and defining it itself. I've added checks for all the
3234 record types that Exim uses.
3236 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3237 not automatically generating a new one, as it is supposed to. This
3238 prevented TLS from working. If the file did exist, but contained invalid
3239 data, a new version was generated, as expected. It was only the case of a
3240 non-existent file that was broken.
3242 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3243 with a change in libdomainkeys > 0.64.
3245 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3246 from DNS. If the selector record carries the flag, it now has
3247 precedence over the domain-wide flag.
3249 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3251 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3252 the use of a port name that isn't defined in /etc/services) occurred, the
3253 message was deferred as in a normal delivery, and thus remained on the
3254 spool, instead of being failed because of the mua_wrapper setting. This
3255 is now fixed, and I tidied up some of the mua_wrapper messages at the
3258 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3259 the messages in a hash of arrays rather than using individual hashes.
3260 This is a bit cleaner and results in dramatic memory savings, albeit
3261 at a slight CPU cost.
3263 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3264 as requested by Marc Sherman.
3266 SC/10 Eximstats: added histograms for user specified patterns as requested
3269 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3271 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3272 fopen() in the content-scanning modules that did not already have it.
3278 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3279 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3281 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3283 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3285 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3286 to test Sieve filters that use "vacation".
3288 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3289 that changes the way the GnuTLS parameters are stored in the cache file.
3290 The new format can be generated externally. For backward compatibility,
3291 if the data in the cache doesn't make sense, Exim assumes it has read an
3292 old-format file, and it generates new data and writes a new file. This
3293 means that you can't go back to an older release without removing the
3296 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3297 work if there are any delivery delays because "one_time" forces the
3298 parent to be marked "delivered", so its unseen clone is never tried
3299 again. For this reason, Exim now forbids the simultaneous setting of
3302 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3303 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3304 entry if you want to know the details. The fix, however, neglected to
3305 consider the case when local delivery batching is involved. The test for
3306 "previously delivered" was not happening when checking to see if an
3307 address could be batched with a previous (undelivered) one; under
3308 certain circumstances this could lead to multiple deliveries to the same
3311 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3312 in its include files, and this causes problems building Exim.
3314 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3315 header_syntax) but Exim was just ignoring anything given after a slash.
3316 In particular, this caused confusion with an attempt to use "verify =
3317 reverse_host_lookup/defer_ok". An error is now given when options are
3318 supplied for verify items that do not have them. (Maybe reverse_host_
3319 lookup should have a defer_ok option, but that's a different point.)
3321 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3322 defined by RFC 821) to 2048, because there were problems with some AUTH
3323 commands, and RFC 1869 says the size should be increased for extended
3324 SMTP commands that take arguments.
3326 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3329 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3330 "unknown" error; now it says that the functionality isn't in the binary.
3332 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3333 an address' error message when a string expansion fails (syntax or
3334 whatever). Otherwise the password may appear in the log. Following change
3335 PH/42 below, there is no longer a chance of it appearing in a bounce
3338 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3340 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3341 of its addresses was ever tried. (Bugzilla bug #2.)
3343 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3344 the result incorrectly in the debug output. (It correctly added a newline
3345 to what was transported.)
3347 TF/01 Added $received_time.
3349 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3350 commented out examples of how to interface to a virus scanner and to
3351 SpamAssassin. Also added commented examples of av_scanner and
3352 spamd_address settings.
3354 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3355 and controls are allowed in which ACLs. There were a couple of minor
3356 errors. Some of the entries in the conditions table (which is a table of
3357 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3358 negation of where the condition IS allowed.
3360 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3362 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3363 header file does not have a version number, so I've had to invent a new
3364 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3365 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3366 radiusclient), but it was contributed by a Radius user.
3368 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3369 files or directories, for OpenSSL.
3371 PH/19 When an Exim process that is running as root has to create an Exim log
3372 file, it does so in a subprocess that runs as exim:exim so as to get the
3373 ownership right at creation (otherwise, other Exim processes might see
3374 the file with the wrong ownership). There was no test for failure of this
3375 fork() call, which would lead to the process getting stuck as it waited
3376 for a non-existent subprocess. Forks do occasionally fail when resources
3377 run out. I reviewed all the other calls to fork(); they all seem to check
3380 PH/20 When checking for unexpected SMTP input at connect time (before writing
3381 the banner), Exim was not dealing correctly with a non-positive return
3382 from the read() function. If the client had disconnected by this time,
3383 the result was a log entry for a synchronization error with an empty
3384 string after "input=" when read() returned zero. If read() returned -1
3385 (an event I could not check), uninitialized data bytes were printed.
3386 There were reports of junk text (parts of files, etc) appearing after
3389 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3391 PH/22 Added support for macro redefinition, and (re)definition in between
3392 driver and ACL definitions.
3394 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3395 forgetting to use the resulting value; it was using the unexpanded value.
3397 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3398 hadn't been configured. The fix is from Juergen Kreileder, who
3399 understands it better than I do:
3401 "Here's what I see happening with three configured cyrus_sasl
3402 authenticators configured (plain, login, cram-md5):
3404 On startup auth_cyrus_sasl_init() gets called for each of these.
3405 This means three calls to sasl_listmech() without a specified mech_list.
3406 => SASL tests which mechs of all available mechs actually work
3407 => three warnings about OTP not working
3408 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3410 With the patch, sasl_listmech() also gets called three times. But now
3411 SASL's mech_list option is set to the server_mech specified in the the
3412 authenticator. Or in other words, the answer from sasl_listmech()
3413 gets limited to just the mech you're testing for (which is different
3415 => the return list contains just 'plain' or 'login', 'cram-md5' or
3416 nothing depending on the value of ob->server_mech.
3418 I've just tested the patch: Authentication still works fine,
3419 unavailable mechs specified in the exim configuration are still
3420 caught, and the auth.log warnings about OTP are gone."
3422 PH/25 When debugging is enabled, the contents of the command line are added
3423 to the debugging output, even when log_selector=+arguments is not
3426 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3427 answer is "GNU", and only if the return is "GNU/something" is the answer
3430 PH/27 $acl_verify_message is now set immediately after the failure of a
3431 verification in an ACL, and so is available in subsequent modifiers. In
3432 particular, the message can be preserved by coding like this:
3434 warn !verify = sender
3435 set acl_m0 = $acl_verify_message
3437 Previously, $acl_verify_message was set only while expanding "message"
3438 and "log_message" when a very denied access.
3440 PH/28 Modified OS/os.c-Linux with
3442 -#ifndef OS_LOAD_AVERAGE
3443 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3445 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3446 nomenclature these days.)
3448 PH/29 Installed patch from the Sieve maintainer that adds the options
3449 sieve_useraddress and sieve_subaddress to the redirect router.
3451 PH/30 In these circumstances:
3452 . Two addresses routed to the same list of hosts;
3453 . First host does not offer TLS;
3454 . First host accepts first address;
3455 . First host gives temporary error to second address;
3456 . Second host offers TLS and a TLS session is established;
3457 . Second host accepts second address.
3458 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3459 and peerdn, if requested) that were in fact used only for the second
3462 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3463 attention to any local part prefix or suffix that was matched by the
3464 router that accepted the address. It now behaves in the same way as it
3465 does for delivery: the affixes are removed from the local part unless
3466 rcpt_include_affixes is set on the transport.
3468 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3469 timeout during the DATA phase of an incoming message.
3471 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3472 applied a patch sanctioned by the Sieve maintainer.
3474 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3475 the uid or gid is negative. A case of a negative gid caused this to be
3476 noticed. The fix allows for either to be negative.
3478 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3479 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3482 PH/36 The change PH/12 above was broken. Fixed it.
3484 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3485 the grounds that routing the same address twice would always produce the
3486 same answer. This might have been true once, but it is certainly no
3487 longer true now. Routing a child address may depend on the previous
3488 routing that produced that child. Some complicated redirection strategies
3489 went wrong when messages had multiple recipients, and made Exim's
3490 behaviour dependent on the order in which the addresses were given.
3492 I have moved the duplicate checking until after the routing is complete.
3493 Exim scans the addresses that are assigned to local and remote
3494 transports, and removes any duplicates. This means that more work will be
3495 done, as duplicates will always all be routed, but duplicates are
3496 presumably rare, so I don't expect this is of any significance.
3498 For deliveries to pipes, files, and autoreplies, the duplicate checking
3499 still happens during the routing process, since they are not going to be
3502 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3503 It corrects a timeout issue with spamd. This is Ian's comment: "The
3504 background is that sometimes spamd either never reads data from a
3505 connection it has accepted, or it never writes response data. The exiscan
3506 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3507 blindly assumes that writes won't block so it may never time out."
3509 PH/39 Allow G after quota size as well as K and M.
3511 PH/40 The value set for $authenticated_id in an authenticator may not contain
3512 binary zeroes or newlines because the value is written to log lines and
3513 to spool files. There was no check on this. Now the value is run through
3514 the string_printing() function so that such characters are converted to
3515 printable escape sequences.
3517 PH/41 $message_linecount is a new variable that contains the total number of
3518 lines in the message. Compare $body_linecount, which is the count for the
3521 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3522 bounce and delay warning messages, except in certain special cases, which
3525 (a) An SMTP error message from a remote host;
3526 (b) A message specified in a :fail: redirection;
3527 (c) A message specified in a "fail" command in a system filter;
3528 (d) A message specified in a FAIL return from the queryprogram router;
3529 (e) A message specified by the cannot_route_message router option.
3531 In these cases only, Exim does include the error details in bounce and
3532 warning messages. There are also a few cases where bland messages such
3533 as "unrouteable address" or "local delivery error" are given.
3535 PH/43 $value is now also set for the "else" part of a ${run expansion.
3537 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3538 being worked on, but at least Exim now implements the latest version to
3541 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3542 process to complete was treated as a delivery failure, a timeout while
3543 writing the message to the pipe was logged, but erroneously treated as a
3544 successful delivery. Such timeouts include transport filter timeouts. For
3545 consistency with the overall process timeout, these timeouts are now
3546 treated as errors, giving rise to delivery failures by default. However,
3547 there is now a new Boolean option for the pipe transport called
3548 timeout_defer, which, if set TRUE, converts the failures into defers for
3549 both kinds of timeout. A transport filter timeout is now identified in
3552 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3553 systems where "make" and "gmake" are different, calling "gmake" at top
3554 level broke things. I've arranged for the value of $(MAKE) to be passed
3555 from the Makefile to this script so that it can call the same version of
3559 A note about Exim versions 4.44 and 4.50
3560 ----------------------------------------
3562 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3563 changes of various kinds. As a consequence, a big documentation update was
3564 needed. This delayed the release for rather longer than seemed good, especially
3565 in the light of a couple of (minor) security issues. Therefore, the changes
3566 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3567 release. So 4.44 and 4.50 are in effect two different branches that both start
3570 I have left the 4.50 change log unchanged; it contains all the changes since
3571 4.43. The change log for 4.44 is below; many of its items are identical to
3572 those for 4.50. This seems to be the most sensible way to preserve the
3573 historical information.
3579 1. Minor wording change to the doc/README.SIEVE file.
3581 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3582 computation of the current number of files was incorrect.
3584 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3585 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3588 4. Give more explanation in the error message when the command for a transport
3589 filter fails to execute.
3591 5. There are several places where Exim runs a non-Exim command in a
3592 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3593 was being done only for the command run by the queryprogram router. It is
3594 now done for all such subprocesses. The other cases are: ${run, transport
3595 filters, and the commands run by the lmtp and pipe transports.
3597 6. Added CONFIGURE_GROUP build-time option.
3599 7. Some older OS have a limit of 256 on the maximum number of file
3600 descriptors. Exim was using setrlimit() to set 1000 as a large value
3601 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3602 systems. I've change it so that if it can't get 1000, it tries for 256.
3604 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3605 was an oversight, and furthermore, ever since the addition of extra
3606 controls (e.g. 4.43/32), the checks on when to allow different forms of
3607 "control" were broken. There should now be diagnostics for all cases when a
3608 control that does not make sense is encountered.
3610 9. Added the /retain_sender option to "control=submission".
3612 10. $recipients is now available in the predata ACL (oversight).
3614 11. Tidy the search cache before the fork to do a delivery from a message
3615 received from the command line. Otherwise the child will trigger a lookup
3616 failure and thereby defer the delivery if it tries to use (for example) a
3617 cached ldap connection that the parent has called unbind on.
3619 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3620 of $address_data from the recipient verification was clobbered by the
3621 sender verification.
3623 13. The value of address_data from a sender verification is now available in
3624 $sender_address_data in subsequent conditions in the ACL statement.
3626 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3628 15. Added a new option "connect=<time>" to callout options, to set a different
3631 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3632 was its contents. (It was OK if the option was not defined at all.)
3634 17. A "Completed" log line is now written for messages that are removed from
3635 the spool by the -Mrm option.
3637 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3638 information about exactly what failed.
3640 19. Added -dd to debug only the daemon process.
3642 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3643 handles timeouts, both on the server side and network timeouts. Renamed the
3644 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3646 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3647 It is now set to "smtps".
3649 22. $host_address is now set to the target address during the checking of
3650 ignore_target_hosts.
3652 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3653 being passed; this would have caused $sender_host_name to have been used if
3654 matching the list had actually called for a host name (not very likely,
3655 since this list is usually IP addresses). A host name is now passed as
3658 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3659 code that specifies a non-restarting handler (typically sigaction() in
3660 modern systems) in an attempt to fix a rare and obscure crash bug.
3662 25. Narrowed the window for a race in the daemon that could cause it to ignore
3663 SIGCHLD signals. This is not a major problem, because they are used only to
3664 wake it up if nothing else does.
3666 26. A malformed maildirsize file could cause Exim to calculate negative values
3667 for the mailbox size or file count. Odd effects could occur as a result.
3668 The maildirsize information is now recalculated if the size or filecount
3671 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3672 support for a long time. Removed HAVE_SYS_VFS_H.
3674 28. Installed the latest version of exipick from John Jetmore.
3676 29. In an address list, if the pattern was not a regular expression, an empty
3677 subject address (from a bounce message) matched only if the pattern was an
3678 empty string. Non-empty patterns were not even tested. This was the wrong
3679 because it is perfectly reasonable to use an empty address as part of a
3680 database query. An empty address is now tested by patterns that are
3681 lookups. However, all the other forms of pattern expect the subject to
3682 contain a local part and a domain, and therefore, for them, an empty
3683 address still always fails if the pattern is not itself empty.
3685 30. Exim went into a mad DNS loop when attempting to do a callout where the
3686 host was specified on an smtp transport, and looking it up yielded more
3687 than one IP address.
3689 31. Re-factored the code for checking spool and log partition space into a
3690 function that finds that data and another that does the check. The former
3691 is then used to implement four new variables: $spool_space, $log_space,
3692 $spool_inodes, and $log_inodes.
3694 32. The RFC2047 encoding function was originally intended for short strings
3695 such as real names; it was not keeping to the 75-character limit for
3696 encoded words that the RFC imposes. It now respects the limit, and
3697 generates multiple encoded words if necessary. To be on the safe side, I
3698 have increased the buffer size for the ${rfc2047: expansion operator from
3701 33. It is now permitted to omit both strings after an "if" condition; if the
3702 condition is true, the result is "true". As before, when the second string
3703 is omitted, a false condition yields an empty string. This makes it less
3704 cumbersome to write custom ACL and router conditions.
3706 34. Failure to deliver a bounce message always caused it to be frozen, even if
3707 there was an errors_to setting on the router. The errors_to setting is now
3710 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3711 canonical form (fully expanded) before being placed in
3712 $sender_host_address.
3714 36. The table in the code that translates DNS record types into text (T_A to
3715 "A" for instance) was missing entries for NS and CNAME. It is just possible
3716 that this could have caused confusion if both these types were looked up
3717 for the same domain, because the text type is used as part of Exim's
3718 per-process caching. But the chance of anyone hitting this buglet seems
3721 37. The dnsdb lookup has been extended in a number of ways.
3723 (1) There is a new type, "zns", which walks up the domain tree until it
3724 finds some nameserver records. It should be used with care.
3726 (2) There is a new type, "mxh", which is like "mx" except that it returns
3727 just the host names, not the priorities.
3729 (3) It is now possible to give a list of domains (or IP addresses) to be
3730 looked up. The behaviour when one of the lookups defers can be
3731 controlled by a keyword.
3733 (4) It is now possible to specify the separator character for use when
3734 multiple records are returned.
3736 38. The dnslists ACL condition has been extended: it is now possible to supply
3737 a list of IP addresses and/or domains to be looked up in a particular DNS
3740 39. Added log_selector=+queue_time_overall.
3742 40. When running the queue in the test harness, wait just a tad after forking a
3743 delivery process, to get repeatability of debugging output.
3745 41. Include certificate and key file names in error message when GnuTLS fails
3746 to set them up, because the GnuTLS error message doesn't include the name
3747 of the failing file when there is a problem reading it.
3749 42. Allow both -bf and -bF in the same test run.
3751 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3753 44. The "Exiscan patch" is now merged into the mainline Exim source.
3755 45. Sometimes the final signoff response after QUIT could fail to get
3756 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3757 < 0 before doing a fflush(). This bug looks as though it goes back to the
3758 introduction of TLS in release 3.20, but "sometimes" must have been rare
3759 because the tests only now provoked it.
3761 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3762 (this can affect the format of dates).
3764 47. exim_tidydb, when checking for the continued existence of a message for
3765 which it has found a message-specific retry record, was not finding
3766 messages that were in split spool directories. Consequently, it was
3767 deleting retry records that should have stayed in existence.
3769 48. Steve fixed some bugs in eximstats.
3771 49. The SPA authentication driver was not abandoning authentication and moving
3772 on to the next authenticator when an expansion was forced to fail,
3773 contradicting the general specification for all authenticators. Instead it
3774 was generating a temporary error. It now behaves as specified.
3776 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3777 (the order specifies the preference for clients). The order is now AES256,
3778 AES128, 3DES, ARCFOUR128.
3780 51. Small patch to Sieve code - explicitly set From: when generating an
3783 52. Exim crashed if a remote delivery caused a very long error message to be
3784 recorded - for instance if somebody sent an entire SpamAssassin report back
3785 as a large number of 550 error lines. This bug was coincidentally fixed by
3786 increasing the size of one of Exim's internal buffers (big_buffer) that
3787 happened as part of the Exiscan merge. However, to be on the safe side, I
3788 have made the code more robust (and fixed the comments that describe what
3791 53. Now that there can be additional text after "Completed" in log lines (if
3792 the queue_time_overall log selector is set), a one-byte patch to exigrep
3793 was needed to allow it to recognize "Completed" as not the last thing in
3796 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3797 patch that reportedly fixes this has been added. I am not expert enough to
3798 create a test for it. This is what the patch creator wrote:
3800 "I found a little strange behaviour of ldap code when working with
3801 Windows 2003 AD Domain, where users was placed in more than one
3802 Organization Units. When I tried to give exim partial DN, the exit code
3803 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3804 But simultaneously result of request was absolutely normal ldap result,
3805 so I produce this patch..."
3807 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3808 REFERENCE, so I have modified the code to exclude the patch when that macro
3811 55. Some experimental protocols are using DNS PTR records for new purposes. The
3812 keys for these records are domain names, not reversed IP addresses. The
3813 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3814 leaves it alone. Component reversal etc. now happens only for IP addresses.
3817 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3819 57. Double the size of the debug message buffer (to 2048) so that more of very
3820 long debug lines gets shown.
3822 58. The exicyclog utility now does better if the number of log files to keep
3823 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3825 59. Two changes related to the smtp_active_hostname option:
3827 (1) $smtp_active_hostname is now available as a variable.
3828 (2) The default for smtp_banner uses $smtp_active_hostname instead
3829 of $primary_hostname.
3831 60. The host_aton() function is supposed to be passed a string that is known
3832 to be a valid IP address. However, in the case of IPv6 addresses, it was
3833 not checking this. This is a hostage to fortune. Exim now panics and dies
3834 if the condition is not met. A case was found where this could be provoked
3835 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3836 components; fortuitously, this particular loophole had already been fixed
3837 by change 4.50/55 above.
3839 If there are any other similar loopholes, the new check in host_aton()
3840 itself should stop them being exploited. The report I received stated that
3841 data on the command line could provoke the exploit when Exim was running as
3842 exim, but did not say which command line option was involved. All I could
3843 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3844 running as the user.
3847 61. There was a buffer overflow vulnerability in the SPA authentication code
3848 (which came originally from the Samba project). I have added a test to the
3849 spa_base64_to_bits() function which I hope fixes it.
3852 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3853 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3855 63. The daemon start-up calls getloadavg() while still root for those OS that
3856 need the first call to be done as root, but it missed one case: when
3857 deliver_queue_load_max is set with deliver_drop_privilege. This is
3858 necessary for the benefit of the queue runner, because there is no re-exec
3859 when deliver_drop_privilege is set.
3861 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3862 This has been fixed.
3864 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3865 was in use, was not putting the data itself into the right store pool;
3866 consequently, it could be overwritten for a subsequent message in the same
3867 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3870 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3872 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3873 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3874 the function were treating the return as a boolean value, which happened to
3875 work because 0=false and not-0=true, but is not correct code.
3877 68. The host_aton() function was not handling scoped IPv6 addresses (those
3878 with, for example, "%eth0" on the end) correctly.
3880 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3881 negated items (that is, ~something) in unsigned ints. Some compilers
3882 apparently mutter when there is no cast.
3884 70. If an address verification called from an ACL failed, and did not produce a
3885 user-specific message (i.e. there was only a "system" message), nothing was
3886 put in $acl_verify_message. In this situation, it now puts the system
3889 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3890 session; change 4.31/43 added the unwanted input to the log line - except
3891 that it did not do this in the start of session case. It now does.
3893 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3894 This is wrong and can cause the other end to generate a synchronization
3895 error if it is another Exim or anything else that does the synchronization
3896 check. A QUIT command is no longer sent after a timeout.
3898 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3899 during host lookups.
3901 74. The defer_ok option of callout verification was not working if it was used
3902 when verifying addresses in header lines, that is, for this case:
3904 verify = header_sender/callout=defer_ok
3906 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3907 those file descriptors could be used for SMTP connections. If anything
3908 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3909 could be sent to the SMTP client, causing chaos. The daemon now opens
3910 stdin, stdout, and stderr to /dev/null when it puts itself into the
3913 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3914 log by default. The user can override this with suitable Perl magic.
3916 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3917 the log message when discard triggers, was not working for the DATA ACL or
3918 for the non-SMTP ACL.
3920 78. Error message wording change in sieve.c.
3922 79. If smtp_accept_max_per_host was set, the number of connections could be
3923 restricted to fewer than expected, because the daemon was trying to set up
3924 a new connection before checking whether the processes handling previous
3925 connections had finished. The check for completed processes is now done
3926 earlier. On busy systems, this bug wouldn't be noticed because something
3927 else would have woken the daemon, and it would have reaped the completed
3930 80. If a message was submitted locally by a user whose login name contained one
3931 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3932 It caused a spool format error. I have fixed the spool reading code. A
3933 related problem was that the "from" clause in the Received: line became
3934 illegal because of the space(s). It is now covered by ${quote_local_part.
3936 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3939 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3941 83. Patch from Sieve maintainer to fix unterminated string problem in
3942 "vacation" handling.
3944 84. Some minor changes to the Linux configuration files to help with other
3945 OS variants using glibc.
3947 85. One more patch for Sieve to update vacation handling to latest spec.
3950 ----------------------------------------------------
3951 See the note above about the 4.44 and 4.50 releases.
3952 ----------------------------------------------------
3958 1. Change 4.43/35 introduced a bug that caused file counts to be
3959 incorrectly computed when quota_filecount was set in an appendfile
3962 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3963 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3966 3. Give more explanation in the error message when the command for a transport
3967 filter fails to execute.
3969 4. There are several places where Exim runs a non-Exim command in a
3970 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3971 was being done only for the command run by the queryprogram router. It is
3972 now done for all such subprocesses. The other cases are: ${run, transport
3973 filters, and the commands run by the lmtp and pipe transports.
3975 5. Some older OS have a limit of 256 on the maximum number of file
3976 descriptors. Exim was using setrlimit() to set 1000 as a large value
3977 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3978 systems. I've change it so that if it can't get 1000, it tries for 256.
3980 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3981 was an oversight, and furthermore, ever since the addition of extra
3982 controls (e.g. 4.43/32), the checks on when to allow different forms of
3983 "control" were broken. There should now be diagnostics for all cases when a
3984 control that does not make sense is encountered.
3986 7. $recipients is now available in the predata ACL (oversight).
3988 8. Tidy the search cache before the fork to do a delivery from a message
3989 received from the command line. Otherwise the child will trigger a lookup
3990 failure and thereby defer the delivery if it tries to use (for example) a
3991 cached ldap connection that the parent has called unbind on.
3993 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3994 of $address_data from the recipient verification was clobbered by the
3995 sender verification.
3997 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3998 was its contents. (It was OK if the option was not defined at all.)
4000 11. A "Completed" log line is now written for messages that are removed from
4001 the spool by the -Mrm option.
4003 12. $host_address is now set to the target address during the checking of
4004 ignore_target_hosts.
4006 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4007 being passed; this would have caused $sender_host_name to have been used if
4008 matching the list had actually called for a host name (not very likely,
4009 since this list is usually IP addresses). A host name is now passed as
4012 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4013 code that specifies a non-restarting handler (typically sigaction() in
4014 modern systems) in an attempt to fix a rare and obscure crash bug.
4016 15. Narrowed the window for a race in the daemon that could cause it to ignore
4017 SIGCHLD signals. This is not a major problem, because they are used only to
4018 wake it up if nothing else does.
4020 16. A malformed maildirsize file could cause Exim to calculate negative values
4021 for the mailbox size or file count. Odd effects could occur as a result.
4022 The maildirsize information is now recalculated if the size or filecount
4025 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4026 support for a long time. Removed HAVE_SYS_VFS_H.
4028 18. Updated exipick to current release from John Jetmore.
4030 19. Allow an empty sender to be matched against a lookup in an address list.
4031 Previously the only cases considered were a regular expression, or an
4034 20. Exim went into a mad DNS lookup loop when doing a callout where the
4035 host was specified on the transport, if the DNS lookup yielded more than
4038 21. The RFC2047 encoding function was originally intended for short strings
4039 such as real names; it was not keeping to the 75-character limit for
4040 encoded words that the RFC imposes. It now respects the limit, and
4041 generates multiple encoded words if necessary. To be on the safe side, I
4042 have increased the buffer size for the ${rfc2047: expansion operator from
4045 22. Failure to deliver a bounce message always caused it to be frozen, even if
4046 there was an errors_to setting on the router. The errors_to setting is now
4049 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4050 canonical form (fully expanded) before being placed in
4051 $sender_host_address.
4053 24. Updated eximstats to version 1.33
4055 25. Include certificate and key file names in error message when GnuTLS fails
4056 to set them up, because the GnuTLS error message doesn't include the name
4057 of the failing file when there is a problem reading it.
4059 26. Expand error message when OpenSSL has problems setting up cert/key files.
4062 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4063 (this can affect the format of dates).
4065 28. exim_tidydb, when checking for the continued existence of a message for
4066 which it has found a message-specific retry record, was not finding
4067 messages that were in split spool directories. Consequently, it was
4068 deleting retry records that should have stayed in existence.
4070 29. eximstats updated to version 1.35
4071 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4072 1.35 - bugfix such that pie charts by volume are generated correctly
4074 30. The SPA authentication driver was not abandoning authentication and moving
4075 on to the next authenticator when an expansion was forced to fail,
4076 contradicting the general specification for all authenticators. Instead it
4077 was generating a temporary error. It now behaves as specified.
4079 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4080 (the order specifies the preference for clients). The order is now AES256,
4081 AES128, 3DES, ARCFOUR128.
4083 31. Small patch to Sieve code - explicitly set From: when generating an
4086 32. Exim crashed if a remote delivery caused a very long error message to be
4087 recorded - for instance if somebody sent an entire SpamAssassin report back
4088 as a large number of 550 error lines. This bug was coincidentally fixed by
4089 increasing the size of one of Exim's internal buffers (big_buffer) that
4090 happened as part of the Exiscan merge. However, to be on the safe side, I
4091 have made the code more robust (and fixed the comments that describe what
4094 33. Some experimental protocols are using DNS PTR records for new purposes. The
4095 keys for these records are domain names, not reversed IP addresses. The
4096 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4097 leaves it alone. Component reversal etc. now happens only for IP addresses.
4100 34. The host_aton() function is supposed to be passed a string that is known
4101 to be a valid IP address. However, in the case of IPv6 addresses, it was
4102 not checking this. This is a hostage to fortune. Exim now panics and dies
4103 if the condition is not met. A case was found where this could be provoked
4104 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4105 components; fortuitously, this particular loophole had already been fixed
4106 by change 4.50/55 or 4.44/33 above.
4108 If there are any other similar loopholes, the new check in host_aton()
4109 itself should stop them being exploited. The report I received stated that
4110 data on the command line could provoke the exploit when Exim was running as
4111 exim, but did not say which command line option was involved. All I could
4112 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4113 running as the user.
4116 35. There was a buffer overflow vulnerability in the SPA authentication code
4117 (which came originally from the Samba project). I have added a test to the
4118 spa_base64_to_bits() function which I hope fixes it.
4121 36. The daemon start-up calls getloadavg() while still root for those OS that
4122 need the first call to be done as root, but it missed one case: when
4123 deliver_queue_load_max is set with deliver_drop_privilege. This is
4124 necessary for the benefit of the queue runner, because there is no re-exec
4125 when deliver_drop_privilege is set.
4127 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4128 was in use, was not putting the data itself into the right store pool;
4129 consequently, it could be overwritten for a subsequent message in the same
4130 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4133 38. Sometimes the final signoff response after QUIT could fail to get
4134 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4135 < 0 before doing a fflush(). This bug looks as though it goes back to the
4136 introduction of TLS in release 3.20, but "sometimes" must have been rare
4137 because the tests only now provoked it.
4143 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4144 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4145 Now it returns an integer. A number of calls were still expecting a T/F
4146 return. Fortuitously, in all cases, the tests worked in OK situations,
4147 which is the norm. However, things would have gone wrong on any write
4148 failures on the smtp file descriptor. This function is used when sending
4149 messages over SMTP and also when doing verify callouts.
4151 2. When Exim is called to do synchronous delivery of a locally submitted
4152 message (the -odf or -odi options), it no longer closes stderr before doing
4155 3. Implemented the mua_wrapper option.
4157 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4159 5. Implemented the functions header_remove(), header_testname(),
4160 header_add_at_position(), and receive_remove_recipient(), and exported them
4163 6. If an ACL "warn" statement specified the addition of headers, Exim already
4164 inserted X-ACL-Warn: at the start if there was no header name. However, it
4165 was not making this test for the second and subsequent header lines if
4166 there were newlines in the string. This meant that an invalid header could
4167 be inserted if Exim was badly configured.
4169 7. Allow an ACL "warn" statement to add header lines at the start or after all
4170 the Received: headers, as well as at the end.
4172 8. Added the rcpt_4xx retry error code.
4174 9. Added postmaster_mailfrom=xxx to callout verification option.
4176 10. Added mailfrom=xxxx to the callout verification option, for verify=
4179 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4180 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4182 12. Inserted some casts to stop certain compilers warning when using pointer
4183 differences as field lengths or precisions in printf-type calls (mostly
4184 affecting debugging statements).
4186 13. Added optional readline() support for -be (dynamically loaded).
4188 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4189 same clock tick as a message's arrival, so that its received time was the
4190 same as the "first fail" time on the retry record, and that message
4191 remained on the queue past the ultimate address timeout, every queue runner
4192 would try a delivery (because it was past the ultimate address timeout) but
4193 after another failure, the ultimate address timeout, which should have then
4194 bounced the address, did not kick in. This was a "< instead of <=" error;
4195 in most cases the first failure would have been in the next clock tick
4196 after the received time, and all would be well.
4198 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4199 being recognized when the domain list was tested by the match_domain
4200 condition in an expansion string.
4202 16. Added the ${str2b64: operator.
4204 17. Exim was always calling setrlimit() to set a large limit for the number of
4205 processes, without checking whether the existing limit was already
4206 adequate. (It did check for the limit on file descriptors.) Furthermore,
4207 errors from getrlimit() and setrlimit() were being ignored. Now they are
4208 logged to the main and panic logs, but Exim does carry on, to try to do its
4209 job under whatever limits there are.
4211 18. Imported PCRE 5.0.
4213 19. Trivial typo in log message " temporarily refused connection" (the leading
4216 20. If the log selector return_path_on_delivery was set and an address was
4217 redirected to /dev/null, the delivery process crashed because it assumed
4218 that a return path would always be set for a "successful" delivery. In this
4219 case, the whole delivery is bypassed as an optimization, and therefore no
4222 21. Internal re-arrangement: the function for sending a challenge and reading
4223 a response while authentication was assuming a zero-terminated challenge
4224 string. It's now changed to take a pointer and a length, to allow for
4225 binary data in such strings.
4227 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4229 23. Exim was not respecting finduser_retries when seeking the login of the
4230 uid under which it was called; it was always trying 10 times. (The default
4231 setting of finduser_retries is zero.) Also, it was sleeping after the final
4232 failure, which is pointless.
4234 24. Implemented tls_on_connect_ports.
4236 25. Implemented acl_smtp_predata.
4238 26. If the domain in control=submission is set empty, Exim assumes that the
4239 authenticated id is a complete email address when it generates From: or
4240 Sender: header lines.
4242 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4243 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4244 chown and chgrp in /bin and hostname in /usr/bin.
4246 28. Exim was keeping the "process log" file open after each use, just as it
4247 does for the main log. This opens the possibility of it remaining open for
4248 long periods when the USR1 signal hits a daemon. Occasional processlog
4249 errors were reported, that could have been caused by this. Anyway, it seems
4250 much more sensible not to leave this file open at all, so that is what now
4253 29. The long-running daemon process does not normally write to the log once it
4254 has entered its main loop, and it closes the log before doing so. This is
4255 so that log files can straightforwardly be renamed and moved. However,
4256 there are a couple of unusual error situations where the daemon does write
4257 log entries, and I had neglected to close the log afterwards.
4259 30. The text of an SMTP error response that was received during a remote
4260 delivery was being truncated at 512 bytes. This is too short for some of
4261 the long messages that one sometimes sees. I've increased the limit to
4264 31. It is now possible to make retry rules that apply only when a message has a
4265 specific sender, in particular, an empty sender.
4267 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4268 it possible to be selective about when SMTP synchronization is enforced.
4270 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4272 32. Implemented hosts_connection_nolog.
4274 33. Added an ACL for QUIT.
4276 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4279 35. Added mailbox_size and mailbox_filecount to appendfile.
4281 36. Added control = no_multiline_responses to ACLs.
4283 37. There was a bug in the logic of the code that waits for the clock to tick
4284 in the case where the clock went backwards by a substantial amount such
4285 that the microsecond fraction of "now" was more than the microsecond
4286 fraction of "then" (but the whole seconds number was less).
4288 38. Added support for the libradius Radius client library this is found on
4289 FreeBSD (previously only the radiusclient library was supported).
4295 1. When certain lookups returned multiple values in the form name=value, the
4296 quoting of the values was not always being done properly. Specifically:
4297 (a) If the value started with a double quote, but contained no whitespace,
4299 (b) If the value contained whitespace other than a space character (i.e.
4300 tabs or newlines or carriage returns) it was not quoted.
4301 This fix has been applied to the mysql and pgsql lookups by writing a
4302 separate quoting function and calling it from the lookup code. The fix
4303 should probably also be applied to nisplus, ibase and oracle lookups, but
4304 since I cannot test any of those, I have not disturbed their existing code.
4306 2. A hit in the callout cache for a specific address caused a log line with no
4307 reason for rejecting RCPT. Now it says "Previous (cached) callout
4308 verification failure".
4310 3. There was an off-by-one bug in the queryprogram router. An over-long
4311 return line was truncated at 256 instead of 255 characters, thereby
4312 overflowing its buffer with the terminating zero. As well as fixing this, I
4313 have increased the buffer size to 1024 (and made a note to document this).
4315 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4316 when Exim is waiting for an SMTP response from a remote server, Exim
4317 restarts its select() call on the socket, thereby resetting its timeout.
4318 This is not a problem when such interrupts are rare. Somebody set up a cron
4319 job to run exiwhat every 2 minutes, which is less than the normal select()
4320 timeout (5 or 10 minutes). This meant that the select() timeout never
4321 kicked in because it was always reset. I have fixed this by comparing the
4322 time when an interrupt arrives with the time at the start of the first call
4323 to select(). If more time than the timeout has elapsed, the interrupt is
4324 treated as a timeout.
4326 5. Some internal re-factoring in preparation for the addition of Sieve
4327 extensions (by MH). In particular, the "personal" test is moved to a
4328 separate function, and given an option for scanning Cc: and Bcc: (which is
4329 not set for Exim filters).
4331 6. When Exim created an email address using the login of the caller as the
4332 local part (e.g. when creating a From: or Sender: header line), it was not
4333 quoting the local part when it contained special characters such as @.
4335 7. Installed new OpenBSD configuration files.
4337 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4338 try to make them clearer.
4340 9. Callout options, other than the timeout value, were being ignored when
4341 verifying sender addresses in header lines. For example, when using
4343 verify = header_sender/callout=no_cache
4345 the cache was (incorrectly) being used.
4347 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4348 only the Cygwin environment.
4350 11. When return_path_on_delivery was set as a log selector, if different remote
4351 addresses in the same message used different return paths and parallel
4352 remote delivery occurred, the wrong values would sometimes be logged.
4353 (Whenever a remote delivery process finished, the return path value from
4354 the most recently started remote delivery process was logged.)
4356 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4357 lines when AUTH and/or TLS are in use. This is the "received protocol"
4358 field. Exim used to use "asmtp" for authenticated SMTP, without any
4359 indication (in the protocol name) for TLS use. Now it follows the RFC and
4360 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4361 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4362 names appear in log lines as well as in Received: header lines.
4364 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4365 extensions, and comparison tests, and to fix some bugs.
4367 14. Changes to the "personal" filter test:
4369 (1) The test was buggy in that it was just doing the equivalent of
4370 "contains" tests on header lines. For example, if a user's address was
4371 anne@some.where, the "personal" test would incorrectly be true for
4373 To: susanne@some.where
4375 This test is now done by extracting each address from the header in turn,
4376 and checking the entire address. Other tests that are part of "personal"
4377 are now done using regular expressions (for example, to check local parts
4378 of addresses in From: header lines).
4380 (2) The list of non-personal local parts in From: addresses has been
4381 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4382 taken from the Sieve specification recommendations.
4384 (3) If the message contains any header line starting with "List-" it is
4385 treated as non-personal.
4387 (4) The test for "circular" in the Subject: header line has been removed
4388 because it now seems ill-conceived.
4390 15. Minor typos in src/EDITME comments corrected.
4392 16. Installed latest exipick from John Jetmore.
4394 17. If headers_add on a router specified a text string that was too long for
4395 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4396 of string_sprintf() is now avoided.
4398 18. $message_body_size was not set (it was always zero) when running the DATA
4399 ACL and the local_scan() function.
4401 19. For the "mail" command in an Exim filter, no default was being set for
4402 the once_repeat time, causing a random time value to be used if "once" was
4403 specified. (If the value happened to be <= 0, no repeat happened.) The
4404 default is now 0s, meaning "never repeat". The "vacation" command was OK
4405 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4406 (I found it when inspecting the code).
4408 20. There is now an overall timeout for performing a callout verification. It
4409 defaults to 4 times the callout timeout, which applies to individual SMTP
4410 commands during the callout. The overall timeout applies when there is more
4411 than one host that can be tried. The timeout is checked before trying the
4412 next host. This prevents very long delays if there are a large number of
4413 hosts and all are timing out (e.g. when the network connections are timing
4414 out). The value of the overall timeout can be changed by specifying an
4415 additional sub-option for "callout", called "maxwait". For example:
4417 verify = sender/callout=5s,maxwait=20s
4419 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4420 to the end before writing, but this should make it even safer).
4422 22. Exim was forgetting that it had advertised PIPELINING for the second and
4423 subsequent messages on an SMTP connection. It was also not resetting its
4424 memory on STARTTLS and an internal HELO.
4426 23. When Exim logs an SMTP synchronization error within a session, it now
4427 records whether PIPELINING has been advertised or not.
4429 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4430 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4431 rather than long int.
4433 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4435 26. Added the never_mail option to autoreply.
4441 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4442 crash if the getsockname() call failed; this can happen if a connection is
4443 closed very soon after it is established. The problem was simply in the
4444 order in which certain operations were done, causing Exim to try to write
4445 to the SMTP stream before it had set up the file descriptor. The bug has
4446 been fixed by making things happen in the correct order.
4452 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4453 before the connection was closed, thus losing the rejection response.
4455 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4456 some early Solaris releases, but causes trouble in current releases where
4457 socklen_t is defined.
4459 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4462 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4465 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4466 that releasing the top of it at the end releases what was used for sub-
4467 expansions (unless the block got too big). However, discard this block if
4468 the first thing is a variable or header, so that we can use its block when
4469 it is dynamic (useful for very large $message_headers, for example).
4471 6. Lookups now cache *every* query, not just the most recent. A new, separate
4472 store pool is used for this. It can be recovered when all lookup caches are
4473 flushed. Lookups now release memory at the end of their result strings.
4474 This has involved some general refactoring of the lookup sources.
4476 7. Some code has been added to the store_xxx() functions to reduce the amount
4477 of flapping under certain conditions.
4479 8. log_incoming_interface used to affect only the <= reception log lines. Now
4480 it causes the local interface and port to be added to several more SMTP log
4481 lines, for example "SMTP connection from", and rejection lines.
4483 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4485 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4487 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4488 could be overwritten at the end of the current message (or the start of a
4489 new message if it was set in a HELO ACL). The value is now preserved for
4490 the duration of the SMTP connection.
4492 12. If a transport had a headers_rewrite setting, and a matching header line
4493 contained an unqualified address, that address was qualified, even if it
4494 did not match any rewriting rules. The underlying bug was that the values
4495 of the flags that permit the existence of unqualified sender and recipient
4496 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4497 non-local messages, and by -bnq for local messages) were not being
4498 preserved with the message after it was received.
4500 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4501 "next input=" as part of the text comprising the host identity instead of
4502 the correct text. The code was using the same buffer for two different
4503 strings. However, depending on which order the printing function evaluated
4504 its arguments, the bug did not always show up. Under Linux, for example, my
4505 test suite worked just fine.
4507 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4508 doesn't work with some older versions of Perl. It has been changed to "my",
4509 which in any case is probably the better facility to use.
4511 15. A really picky compiler found some instances of statements for creating
4512 error messages that either had too many or two few arguments for the format
4515 16. The size of the buffer for calls to the DNS resolver has been increased
4516 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4517 for addresses that have a lot of PTR records. This alleviates a problem; it
4518 does not fully solve it.
4520 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4521 buffer now truncates the list and logs the incident, which is the same
4522 action as happens when Exim is looking up a host name and its aliases.
4523 Previously in this situation something unpredictable would happen;
4524 sometimes it was "internal error: store_reset failed".
4526 18. If a server dropped the connection unexpectedly when an Exim client was
4527 using GnuTLS and trying to read a response, the client delivery process
4528 crashed while trying to generate an error log message.
4530 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4531 string, for example:
4533 warn message = H1: something\nH2: something
4535 the text was added as a single header line from Exim's point of view
4536 though it ended up OK in the delivered message. However, searching for the
4537 second and subsequent header lines using $h_h2: did not work. This has been
4538 fixed. Similarly, if a system filter added multiple headers in this way,
4539 the routers could not see them.
4541 20. Expanded the error message when iplsearch is called with an invalid key to
4542 suggest using net-iplsearch in a host list.
4544 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4545 ACLs are no longer actually imposed (and a message to that effect is
4548 22. If a "gecos" field in a passwd entry contained escaped characters, in
4549 particular, if it contained a \" sequence, Exim got it wrong when building
4550 a From: or a Sender: header from that name. A second bug also caused
4551 incorrect handling when an unquoted " was present following a character
4552 that needed quoting.
4554 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4555 was not being matched caselessly.
4557 24. Arranged for all hyphens in the exim.8 source to be escaped with
4560 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4561 properly. Recipient callout cache records were still being keyed to include
4562 the sender, even when use_sender was set false. This led to far more
4563 callouts that were necessary. The sender is no longer included in the key
4564 when use_sender is false.
4566 26. Added "control = submission" modifier to ACLs.
4568 27. Added the ${base62d: operator to decode base 62 numbers.
4570 28. dnsdb lookups can now access SRV records.
4572 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4573 the configuration file.
4575 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4576 (-v) mode. This makes the output for a verbose queue run more intelligible.
4578 31. Added a use_postmaster feature to recipient callouts.
4580 32. Added the $body_zerocount variable, containing the number of binary zero
4581 bytes in the message body.
4583 33. The time of last modification of the "new" subdirectory is now used as the
4584 "mailbox time last read" when there is a quota error for a maildir
4587 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4589 35. Added +ignore_unknown as a special item in host lists.
4591 36. Code for decoding IPv6 addresses in host lists is now included, even if
4592 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4593 address was recognized as an IP address, but was then not correctly decoded
4594 into binary, causing unexpected and incorrect effects when compared with
4601 1. Very minor rewording of debugging text in manualroute to say "list of
4602 hosts" instead of "hostlist".
4604 2. If verify=header_syntax was set, and a header line with an unqualified
4605 address (no domain) and a large number of spaces between the end of the
4606 name and the colon was received, the reception process suffered a buffer
4607 overflow, and (when I tested it) crashed. This was caused by some obsolete
4608 code that should have been removed. The fix is to remove it!
4610 3. When running in the test harness, delay a bit after writing a bounce
4611 message to get a bit more predictability in the log output.
4613 4. Added a call to search_tidyup() just before forking a reception process. In
4614 theory, someone could use a lookup in the expansion of smtp_accept_max_
4615 per_host which, without the tidyup, could leave open a database connection.
4617 5. Added the variables $recipient_data and $sender_data which get set from a
4618 lookup success in an ACL "recipients" or "senders" condition, or a router
4619 "senders" option, similar to $domain_data and $local_part_data.
4621 6. Moved the writing of debug_print from before to after the "senders" test
4624 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4625 problems for message scanning, either using a data ACL, or using
4626 local_scan() because the Received: header was not generated till after they
4627 were called (in order to set the time as the time of reception completion).
4628 I have revised the way this works. The header is now generated after the
4629 body is received, but before the ACL or local_scan() are called. After they
4630 are run, the timestamp in the header is updated.
4636 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4637 before starting a queue runner without re-exec. This happened only when
4638 deliver_drop_privilege was set or when the Exim user was set to root. The
4639 effect of the bug was that timeouts during subsequent deliveries caused
4640 crashes instead of being properly handled. The handler is now left at its
4641 default (and expected) setting.
4643 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4644 message, again when deliver_drop_privilege is set or Exim is run as root.
4645 The bug described in (1) was not present in this case, but the tidying up
4646 of the other signals was missing. I have made the two cases consistent.
4648 3. The ignore_target_hosts setting on a manualroute router was being ignored
4649 for hosts that were looked up using the /MX notation.
4651 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4654 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4655 operated on the sender address. After changing the $sender_address to <>
4656 for the sender address verify, Exim was re-instated it as the original
4657 (before rewriting) address, but remembering that it had rewritten it, so it
4658 wasn't rewriting it again. This bug also had the effect of breaking the
4659 sender address verification caching when the sender address was rewritten.
4661 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4662 This has been changed so that if the ip literal address matches
4663 ignore_target_hosts, the router declines.
4665 7. Added expansion conditions match_domain, match_address, and match_local_
4666 part (NOT match_host).
4668 8. The placeholder for the Received: header didn't have a length field set.
4670 9. Added code to Exim itself and to exim_lock to test for a specific race
4671 condition that could lead to file corruption when using MBX delivery. The
4672 issue is with the lockfile that is created in /tmp. If this file is removed
4673 after a process has opened it but before that process has acquired a lock,
4674 there is the potential for a second process to recreate the file and also
4675 acquire a lock. This could lead to two Exim processes writing to the file
4676 at the same time. The added code performs the same test as UW imapd; it
4677 checks after acquiring the lock that its file descriptor still refers to
4678 the same named file.
4680 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4681 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4684 11. Added the smtp_active_hostname option. If used, this will typically be made
4685 to depend on the incoming interface address. Because $interface_address is
4686 not set up until the daemon has forked a reception process, error responses
4687 that can happen earlier (such as "too many connections") no longer contain
4690 12. If an expansion in a condition on a "warn" statement fails because a lookup
4691 defers, the "warn" statement is abandoned, and the next ACL statement is
4692 processed. Previously this caused the whole ACL to be aborted.
4694 13. Added the iplsearch lookup type.
4696 14. Added ident_timeout as a log selector.
4698 15. Added tls_certificate_verified as a log selector.
4700 16. Added a global option tls_require_ciphers (compare the smtp transport
4701 option of the same name). This controls incoming TLS connections.
4703 17. I finally figured out how to make tls_require_ciphers do a similar thing
4704 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4705 before starting the TLS session.
4707 18. Tabs are now shown as \t in -bP output.
4709 19. If the log selector return_path_on_delivery was set, Exim crashed when
4710 bouncing a message because it had too many Received: header lines.
4712 20. If two routers both had headers_remove settings, and the first one included
4713 a superfluous trailing colon, the final name in the first list and the
4714 first name in the second list were incorrectly joined into one item (with a
4715 colon in the middle).
4721 1. Added -C and -D options to the exinext utility, mainly to make it easier
4722 to include in the automated testing, but these could be helpful when
4723 multiple configurations are in use.
4725 2. The exinext utility was not formatting the output nicely when there was
4726 an alternate port involved in the retry record key, nor when there was a
4727 message id as well (for retries that were specific to a specific message
4728 and a specific host). It was also confused by IPv6 addresses, because of
4729 the additional colons they contain. I have fixed the IPv4 problem, and
4730 patched it up to do a reasonable job for IPv6.
4732 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4733 delivery, the log line now contains "pipelined" if PIPELINING was used.
4735 4. An SMTP transport process used to panic and die if the bind() call to set
4736 an explicit outgoing interface failed. This has been changed; it is now
4737 treated in the same way as a connect() failure.
4739 5. A reference to $sender_host_name in the part of a conditional expansion
4740 that was being skipped was still causing a DNS lookup. This no longer
4743 6. The def: expansion condition was not recognizing references to header lines
4744 that used bh_ and bheader_.
4746 7. Added the _cache feature to named lists.
4748 8. The code for checking quota_filecount in the appendfile transport was
4749 allowing one more file than it should have been.
4751 9. For compatibility with Sendmail, the command line option
4759 and sets the incoming protocol and host name (for trusted callers). The
4760 host name and its colon can be omitted when only the protocol is to be set.
4761 Note the Exim already has two private options, -pd and -ps, that refer to
4762 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4763 "s", but I don't think that's a major issue.
4765 10. A number of refactoring changes to the code, none of which should affect
4768 (a) The number of logging options was getting close to filling up the
4769 32-bit word that was used as a bit map. I have split them into two classes:
4770 those that are passed in the argument to log_write(), and those that are
4771 only ever tested independently outside of that function. These are now in
4772 separate 32-bit words, so there is plenty of room for expansion again.
4773 There is no change in the user interface or the logging behaviour.
4775 (b) When building, for example, log lines, the code previously used a
4776 macro that called string_cat() twice, in order to add two strings. This is
4777 not really sufficiently general. Furthermore, there was one instance where
4778 it was actually wrong because one of the argument was used twice, and in
4779 one call a function was used. (As it happened, calling the function twice
4780 did not affect the overall behaviour.) The macro has been replaced by a
4781 function that can join an arbitrary number of extra strings onto a growing
4784 (c) The code for expansion conditions now uses a table and a binary chop
4785 instead of a serial search (which was left over from when there were very
4786 few conditions). Also, it now recognizes conditions like "pam" even when
4787 the relevant support is not compiled in: a suitably worded error message is
4788 given if an attempt is made to use such a condition.
4790 11. Added ${time_interval:xxxxx}.
4792 12. A bug was causing one of the ddress fields not to be passed back correctly
4793 from remote delivery subprocesses. The field in question was not being
4794 subsequently used, so this caused to problems in practice.
4796 13. Added new log selectors queue_time and deliver_time.
4798 14. Might have fixed a bug in maildirsizefile handling that threw up
4799 "unexpected character" debug warnings, and recalculated the data
4800 unnecessarily. In any case, I expanded the warning message to give more
4803 15. Added the message "Restricted characters in address" to the statements in
4804 the default ACL that block characters like @ and % in local parts.
4806 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4807 Three changes have been made:
4809 (a) There was a serious bug; a negative response to MAIL caused the whole
4810 recipient domain to be cached as invalid, thereby blocking all messages
4811 to all local parts at the same domain, from all senders. This bug has
4812 been fixed. The domain is no longer cached after a negative response to
4813 MAIL if the sender used is not empty.
4815 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4818 (c) A new callout option, "use_sender" has been added for people who want
4819 the modified behaviour.
4825 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4828 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4829 indeed breaks things for older releases.
4831 3. Added additional logging to the case where there is a problem reading data
4832 from a filter that is running in a subprocess using a pipe, in order to
4833 try to track down a specific problem.
4835 4. Testing facility fudge: when running in the test harness and attempting
4836 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4837 getting "No route to host". Convert this to a timeout.
4839 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4842 6. Some OS don't have socklen_t but use size_t instead. This affects the
4843 fifth argument of getsockopt() amongst other things. This is now
4844 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4845 can be set for individual OS. I have set it for SunOS5, OSF1, and
4846 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4847 some earlier ones do not.
4849 7. Change 4.30/15 was not doing the test caselessly.
4851 8. The standard form for an IPv6 address literal was being rejected by address
4852 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4853 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4854 this, as well as the form without the "IPv6" on the front (but only when
4855 address literals are enabled, of course).
4857 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4859 10. Exim crashed if a message with an empty sender address specified by -f
4860 encountered a router with an errors_to setting. This could be provoked only
4861 by a command such as
4865 where an empty string was supplied; "<>" did not hit this bug.
4867 11. Installed PCRE release 4.5.
4869 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4870 remained set. It is now erased.
4872 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4873 times from message ids (which are base 36 rather than the normal 62).
4875 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4876 were being counted as actual protocol errors, and logged if the log
4877 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4878 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4879 and DATA following a set of rejected RCPTs do not count as protocol errors.
4880 In other words, Exim assumes they were pipelined, though this may not
4881 actually be the case. Of course, in all cases the client gets an
4882 appropriate error code.
4884 15. If a lookup fails in an ACL condition, a message about the failure may
4885 be available; it is used if testing the ACL cannot continue, because most
4886 such messages specify what the cause of the deferral is. However, some
4887 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4888 that caused an old message to be retained and used if a later statement
4889 caused a defer, replacing the real cause of the deferral.
4891 16. If an IP address had so many PTR records that the DNS lookup buffer
4892 was not large enough to hold them, Exim could crash while trying to process
4893 the truncated data. It now detects and logs this case.
4895 17. Further to 4.21/58, another change has been made: if (and only if) the
4896 first line of a message (the first header line) ends with CRLF, a bare LF
4897 in a subsequent header line has a space inserted after it, so as not to
4898 terminate the header.
4900 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4901 unnecessarily, used atoi() instead of strtol(), and didn't check the
4902 termination when getting file sizes from file names by regex.
4904 19. Completely re-implemented the support for maildirsize files, in the light
4905 of a number of problems with the previous contributed implementation
4906 (4.30/29). In particular:
4908 . If the quota is zero, the maildirsize file is maintained, but no quota is
4911 . If the maildir directory does not exist, it is created before any attempt
4912 to write a maildirsize file.
4914 . The quota value in the file is just a cache; if the quota is changed in
4915 the transport, the new value overrides.
4917 . A regular expression is available for excluding directories from the
4920 20. The autoreply transport checks the characters in options that define the
4921 message's headers; it allows continued headers, but it was checking with
4922 isspace() after an embedded newline instead of explicitly looking for a
4925 21. If all the "regular" hosts to which an address was routed had passed their
4926 expiry times, and had not reached their retry times, the address was
4927 bounced, even if fallback hosts were defined. Now Exim should go on to try
4930 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4931 equivalent code in the SMTP transport. Some hosts send humungous responses
4932 to HELO/EHLO, more than 1024 it seems.
4934 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4935 gives compiler warnings in some environments. I've now done it "properly",
4938 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4939 (because of problems with the built-in one) was declared to return uschar *
4940 instead of char *, causing compiler failure.
4942 25. Fixed a file descriptor leak when processing alias/forward files.
4944 26. Fixed a minor format string issue in dbfn.c.
4946 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4948 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4949 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4952 29. When a sender address is verified, it is cached, to save repeating the test
4953 when there is more than one recipient in a message. However, when the
4954 verification involves a callout, it is possible for different callout
4955 options to be set for different recipients. It is too complicated to keep
4956 track of this in the cache, so now Exim always runs a verification when a
4957 callout is required, relying on the callout cache for the optimization.
4958 The overhead is duplication of the address routing, but this should not be
4961 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4962 to be verified with callout=postmaster, and the main callout worked but the
4963 postmaster check failed, the verification correctly failed. However, if a
4964 subsequent RCPT command asked for sender verification *without* the
4965 postmaster check, incorrect caching caused this verification also to fail,
4968 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4969 it was not caching the DNS options (qualify_single, search_parents) that
4970 were used when the lookup failed. A subsequent lookup with different
4971 options therefore always gave the same answer, though there were cases
4972 where it should not have. (Example: a "domains = !$mx_any" option on a
4973 dnslookup router: the "domains" option is always processed without any
4974 widening, but the router might have qualify_single set.) Now Exim uses the
4975 cached value only when the same options are set.
4977 32. Added John Jetmore's "exipick" utility to the distribution.
4979 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4980 than a timeout (e.g. a certificate is required, and is not provided), an
4981 Exim server now closes the connection immediately. Previously it waited for
4982 the client to close - but if the client is SSL, it seems that they each
4983 wait for each other, leading to a delay before one of them times out.
4985 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4986 maintained 0.8.x compatibility because I don't think many are using it, and
4987 it is clearly obsolete.
4989 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4992 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4993 client certificate was expired. A simple patch fixes this, though I don't
4994 understand the full logic of why the verify callback is called multiple
4997 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4998 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4999 which causes problems with some clients (such as the Certicom SSL Plus
5000 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5001 disables the coutermeasure allowing Eudora to connect."
5003 38. Exim was not checking that a write() to a log file succeeded. This could
5004 lead to Bad Things if a log got too big, in particular if it hit a file
5005 size limit. Exim now panics and dies if it cannot write to a log file, just
5006 as it does if it cannot open a log file.
5008 39. Modified OS/Makefile-Linux so that it now contains
5010 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5012 The two -D definitions ensure that Exim is compiled with large file
5013 support, which makes it possible to handle log files that are bigger than
5016 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5017 instance) a domain was checked against a named list that involved a lookup,
5018 causing $domain_data to be set, then another domain was checked against the
5019 same list, then the first domain was re-checked, the value of $domain_data
5020 after the final check could be wrong. In particular, if the second check
5021 failed, it could be set empty. This bug probably also applied to
5024 41. The strip_trailing_dot option was not being applied to the address given
5025 with the -f command-line option.
5027 42. The code for reading a message's header from the spool was incrementing
5028 $received_count, but never initializing it. This meant that the value was
5029 incorrect (doubled) while delivering a message in the same process in which
5030 it was received. In the most common configuration of Exim, this never
5031 happens - a fresh exec is done - but it can happen when
5032 deliver_drop_privilege is set.
5034 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5035 it now includes up to 150 characters of the unexpected data in the log
5038 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5039 and building data strings. The size of both of these buffers was 10 000
5040 bytes - far larger than anybody would *ever* want, thought I. Needless to
5041 say, somebody hit the limit. I have increased the maximum line length to
5042 20 000 and the maximum data length of concatenated lines to 100 000. I have
5043 also fixed two bugs, because there was no checking on these buffers. Tsk,
5044 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5045 buffer is too small.
5047 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5048 lsearch lookups. Now it does.
5050 46. When parsing a route_list item in a manualroute router, a fixed-length
5051 buffer was used for the list of hosts. I made this 1024 bytes long,
5052 thinking that nobody would ever have a list of hosts that long. Wrong.
5053 Somebody had a whole pile of complicated expansion conditions, and the
5054 string was silently truncated, leading to an expansion error. It turns out
5055 that it is easier to change to an unlimited length (owing to other changes
5056 that have happened since this code was originally written) than to build
5057 structure for giving a limitation error. The length of the item that
5058 expands into the list of hosts is now unlimited.
5060 47. The lsearch lookup could not handle data where the length of text line was
5061 more than 4095 characters. Such lines were truncated, leading to shortened
5062 data being returned. It should now handle lines of any length.
5064 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5065 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5068 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5069 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5070 no longer happens in comment lines. A list of replacements is now placed
5071 at the head of all of the source files, except those whose only change is
5072 to replace PERL_COMMAND in the very first #! line.
5074 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5075 messages on the queue, with a bottom-up merge sort, using code contributed
5076 by Michael Haardt. This should make operations like -bp somewhat faster on
5077 large queues. It won't affect queue runners, except when queue_run_in_order
5080 51. Installed eximstats 1.31 in the distribution.
5082 52. Added support for SRV lookups to the dnslookup router.
5084 53. If an ACL referred to $message_body or $message_body_end, the value was not
5085 reset for any messages that followed in the same SMTP session.
5087 54. The store-handling optimization for building very long strings was not
5088 differentiating between the different store pools. I don't think this
5089 actually made any difference in practice, but I've tidied it.
5091 55. While running the routers to verify a sender address, $sender_address
5092 was still set to the sender address. This is wrong, because when routing to
5093 send a bounce to the sender, it would be empty. Therefore, I have changed
5094 it so that, while verifying a sender address, $sender_address is set to <>.
5095 (There is no change to what happens when verifying a recipient address.)
5097 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5098 target A or AAAA records (if not already returned) without resetting the
5099 qualify_single or search_parents options of the DNS resolver. These are
5100 inappropriate in this case because the targets of MX and SRV records must
5101 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5102 target that, when qualified, matched something in the local domain. These
5103 two options are now turned off when doing these lookups.
5105 57. It seems that at least some releases of Reiserfs (which does not have the
5106 concept of a fixed number of inodes) returns zero and not -1 for the
5107 number of available inodes. This interacted badly with check_spool_inodes,
5108 which assumed that -1 was the "no such thing" setting. What I have done is
5109 to check that the total number of inodes is greater than zero before doing
5110 the test of how many are available.
5112 58. When a "warn" ACL statement has a log_message modifier, the message is
5113 remembered, and not repeated. This is to avoid a lot of repetition when a
5114 message has many recipients that cause the same warning to be written.
5115 Howewer, Exim was preserving the list of already written lines for an
5116 entire SMTP session, which doesn't seem right. The memory is now reset if a
5117 new message is started.
5119 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5120 debugging output unless log_rewrite was also set.
5122 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5123 of (int)(handle) when we know that handle contains (void *)(-1).
5125 61. The Exim daemon panic-logs an error return when it closes the incoming
5126 connection. However "connection reset by peer" seems to be common, and
5127 isn't really an error worthy of noting specially, so that particular error
5130 62. When Exim is trying to find all the local interfaces, it used to panic and
5131 die if the ioctl to get the interface flags failed. However, it seems that
5132 on at least one OS (Solaris 9) it is possible to have an interface that is
5133 included in the list of interfaces, but for which you get a failure error
5134 for this call. This happens when the interface is not "plumbed" into a
5135 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5136 failure of the "get flags" call assumes that the interface is down.
5138 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5139 makes life easier for people who are doing arithmetic on fields extracted
5140 from dates, where you often get leading zeros that should not be
5141 interpreted as octal.
5143 64. Added qualify_domain to the redirect router, to override the global
5146 65. If a pathologically long header line contained very many addresses (the
5147 report of this problem mentioned 10 000) and each of them was rewritten,
5148 Exim could use up a very large amount of memory. (It kept on making new
5149 copies of the header line as it rewrote, and never released the old ones.)
5150 At the expense of a bit more processing, the header rewriting function has
5151 been changed so that it no longer eats memory in this way.
5153 66. The generation of the Received: header has been moved from the time that a
5154 message starts to be received, to the time that it finishes. The timestamp
5155 in the Received: header should now be very close to that of the <= log
5156 line. There are two side-effects of this change:
5158 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5159 logged header lines no longer include the local Received: line, because
5160 it has not yet been created. The same applies to a copy of the message
5161 that is returned to a non-SMTP sender when a message is rejected.
5163 (b) When a filter file is tested using -bf, no additional Received: header
5164 is added to the test message. After some thought, I decided that this
5167 This change does not affect the value of $received_for. It is still set
5168 after address rewriting, but before local_scan() is called.
5170 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5172 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5173 gave an unhelpful panic error message, and a defer error. I have managed to
5174 change this behaviour so that it now rejects any supplied certificate,
5175 which seems right, as the list of acceptable certificates is empty.
5177 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5178 gave an unhelpful defer error. I have not managed to make this reject any
5179 supplied certificates, but the error message it gives is "no certificate
5180 supplied", which is not helpful.
5182 70. exigrep's output now also includes lines that are not associated with any
5183 message, but which match the given pattern. Implemented by a patch from
5184 Martin Sluka, which also tidied up the Perl a bit.
5186 71. Recipient callout verification, like sender verification, was using <> in
5187 the MAIL FROM command. This isn't really the right thing, since the actual
5188 sender may affect whether the remote host accepts the recipient or not. I
5189 have changed it to use the actual sender in the callout; this means that
5190 the cache record is now keyed on a recipient/sender pair, not just the
5191 recipient address. There doesn't seem to be a real danger of callout loops,
5192 since a callout by the remote host to check the sender would use <>.
5193 [SEE ABOVE: changed after hitting problems.]
5195 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5196 temporary errors. However, in the case of such a code being given after
5197 the end of a data transmission (i.e. after ".") Exim was failing to write
5198 a retry record for the message. (Yes, there was some broken host that was
5199 actually sending 8xx at this point.)
5201 73. An unknown lookup type in a host list could cause Exim to panic-die when
5202 the list was checked. (An example that provoked this was putting <; in the
5203 middle of a list instead of at the start.) If this happened during a DATA
5204 ACL check, a -D file could be left lying around. This kind of configuration
5205 error no longer causes Exim to die; instead it causes a defer errror. The
5206 incident is still logged to the main and panic logs.
5208 74. Buglet left over from Exim 3 conversion. The message "too many messages
5209 in one connection" was written to the rejectlog but not the mainlog, except
5210 when address rewriting (yes!) was being logged.
5212 75. Added write_rejectlog option.
5214 76. When a system filter was run not as root (that is, when system_filter_user
5215 was set), the values of the $n variables were not being returned to the
5216 main process; thus, they were not subsequently available in the $sn
5219 77. Added +return_path_on_delivery log selector.
5221 78. A connection timeout was being treated differently from recipients deferred
5222 when testing hosts_max_try with a message that was older than the host's
5223 retry timeout. (The host should not be counted, thus allowing all hosts to
5224 be tried at least once before bouncing.) This may have been the cause of an
5225 occasionally reported bug whereby a message would remain on the queue
5226 longer than the retry timeout, but would be bounced if a delivery was
5227 forced. I say "may" because I never totally pinned down the problem;
5228 setting up timeout/retry tests is difficult. See also the next item.
5230 79. The ultimate address timeout was not being applied to errors that involved
5231 a combination of host plus message (for example, a timeout on a MAIL
5232 command). When an address resolved to a number of possible hosts, and they
5233 were not all tried for each delivery (e.g. because of hosts_max_try), a
5234 message could remain on the queue longer than the retry timeout.
5236 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5239 81. Fixed an obscure SMTP outgoing bug which required at least the following
5240 conditions: (a) there was another message waiting for the same server;
5241 (b) the server returned 5xx to all RCPT commands in the first message so
5242 that the message was not completed; (c) the server dropped the connection
5243 or gave a negative response to the RSET that Exim sends to abort the
5244 transaction. The observed case was a dropped connection after DATA that had
5245 been sent in pipelining mode. That is, the server had advertised PIPELINING
5246 but was not implementing it correctly. The effect of the bug was incorrect
5247 behaviour, such as trying another host, and this could lead to a crash.
5253 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5254 and daemon.c were passed as pointers to ints; they should have been
5255 pointers to socklen_t variables (which are typically unsigned ints).
5257 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5260 3. Fixed a really odd bug that affected only the testing scheme; patching a
5261 certain fixed string in the binary changed the value of another string that
5262 happened to be identical to the end of the original first string.
5264 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5265 name", it returns that address as the IP address. On some operating
5266 systems (e.g. Solaris), it also passes back the IP address string as the
5267 "host name". However, on others (e.g. Linux), it passes back an empty
5268 string. Exim wasn't checking for this, and was changing the host name to an
5269 empty string, assuming it had been canonicized.
5271 5. Although rare, it is permitted to have more than one PTR record for a given
5272 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5273 all the names associated with an address, because they do in Solaris.
5274 However, it seems that they do not in Linux for data that comes from the
5275 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5276 I found this out when I moved to a new Linux workstation and tried to run
5277 the Exim test suite.
5279 To get round this problem I have changed the code so that it now does its
5280 own call to the DNS to look up PTR records when searching for a host name.
5281 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5282 addresses that are only in /etc/hosts are still found.
5284 This behaviour is, however, controlled by an option called host_lookup_
5285 order, which defaults to "bydns:byaddr". If people want to use the other
5286 order, or indeed, just use one or the other means of lookup, they can
5287 specify it in this variable.
5289 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5290 some operating systems, this comes back from gethostbyaddr() as an empty
5291 string, and this is what Exim used to test for. However, it seems that in
5292 other systems, "." is yielded. Exim now tests for this case too.
5294 7. The values of check_spool_space and check_log_space are now held internally
5295 as a number of kilobytes instead of an absolute number of bytes. If a
5296 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5297 kilobyte. This means that much larger values can be stored.
5299 8. Exim monitor: an attempt to get the action menu when not actually pointing
5300 at a message produces an empty menu entitled "No message selected". This
5301 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5302 no entries in it ("Shell widget menu has zero width and/or height"). So I
5303 have added a single, blank menu entry in this case.
5305 9. Added ${quote_local_part.
5307 10. MIME decoding is now applied to the contents of Subject: header lines when
5310 11. Now that a reference to $sender_host_address automatically causes a reverse
5311 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5312 host lookup before query-style lookups in lists that might use this
5313 variable. This has therefore been abolished, and the "net-" prefix is no
5314 longer necessary for query-style lookups.
5316 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5317 have been a typo for LFLAGS, so it has been changed.
5319 13. The install script calls Exim with "-C /dev/null" in order to find the
5320 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5321 to be output. Howeve, since Exim outputs its version number before the
5322 error, it didn't break the script. It just looked ugly. I fixed this by
5323 always allowing "-C /dev/null" if the caller is root.
5325 14. Ignore overlarge ACL variable number when reading spool file - insurance
5326 against a later release with more variables having written the file.
5328 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5329 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5330 this, as well as the form without the "IPv6" on the front.
5332 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5333 OS/Makefile-Darwin file.
5335 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5336 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5338 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5339 variable. However, some OS use ints for inodes. I've added cast to long int
5340 to get rid of the compiler warning.
5342 19. I had forgotten to lock out "/../" in configuration file names when
5343 ALT_CONFIG_PREFIX was set.
5345 20. Routers used for verification do not need to specify transports. However,
5346 if such a router generated a host list, and callout was configured, Exim
5347 crashed, because it could not find a port number from the (non-existent)
5348 transport. It now assumes port 25 in this circumstance.
5350 21. Added the -t option to exigrep.
5352 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5353 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5354 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5355 them is set, LOOKUP_LSEARCH is forced.
5357 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5359 24. Added sender and host information to the "rejected by local_scan()" log
5360 line; previously there was no indication of these.
5362 25. Added .include_if_exists.
5364 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5365 Linux. It turns out that not all file systems support this. Apparently some
5366 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5367 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5368 not supported on the file descriptor, is now ignored when Exim is trying to
5369 sync a directory. This applies only to Linux.
5371 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5373 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5374 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5377 29. Added support for maildirsize files from supplied patch (modified a bit).
5379 30. The use of :fail: followed by an empty string could lead Exim to respond to
5380 sender verification failures with (e.g.):
5382 550 Verification failed for <xxx>
5383 550 Sender verify failed
5385 where the first response line was missing the '-' that indicates it is not
5386 the final line of the response.
5388 31. The loop for finding the name of the user that called Exim had a hardwired
5389 limit of 10; it now uses the value of finduser_retries, which is used for
5390 all other user lookups.
5392 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5395 33. Exim was neglecting to zero errno before one call of strtol() when
5396 expanding a string and expecting an integer value. On some systems this
5397 resulted in spurious "integer overflow" errors. Also, it was casting the
5398 result into an int without checking.
5400 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5401 did not work. The code looks as if it has *never* worked, though it appears
5402 to have been documented since at least releast 1.62. I have made it work.
5404 35. The "timeout_DNS" error in retry rules, also documented since at least
5405 1.62, also never worked. As it isn't clear exactly what this means, and
5406 clearly it isn't a major issue, I have abolished the feature by treating it
5407 as "timeout", and writing a warning to the main and panic logs.
5409 36. The display of retry rules for -brt wasn't always showing the error code
5412 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5413 timeout_connect_A, timeout_connect_MX.
5415 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5416 to the empty sender.
5418 39. The daemon was not analysing the content of -oX till after it had closed
5419 stderr and disconnected from the controlling terminal. This meant that any
5420 syntax errors were only noted on the panic log, and the return code from
5421 the command was 0. By re-arranging the code a little, I've made the
5422 decoding happen first, so such errors now appear on stderr, and the return
5423 code is 1. However, the actual setting up of the sockets still happens in
5424 the disconnected process, so errors there are still only recorded on the
5427 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5428 connections (as happens on some IP stacks) was logged at start up time as
5429 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5430 it from "IPv6 and IPv4", which means that two separate sockets are being
5433 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5434 says whether AF_INET or AF_INET6 was passed as an argument.
5436 42. Exiwhat output was messed up when time zones were included in log
5439 43. Exiwhat now gives more information about the daemon's listening ports,
5440 and whether -tls-on-connect was used.
5442 44. The "port" option of the smtp transport is now expanded.
5444 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5445 silently ignored. Now an error message is written to the main and panic
5448 46. There's a new ACL modifier called "logwrite" which writes to a log file
5449 as soon as it is encountered.
5451 47. Added $local_user_uid and $local_user_gid at routing time.
5453 48. Exim crashed when trying to verify a sender address that was being
5456 49. Exim was recognizing only a space character after ".include". It now also
5457 recognizes a tab character.
5459 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5460 extracting the relevant information from the specification. The man page no
5461 longer contains scrambled data for the -d option, and I've added a section
5462 at the front about calling Exim under different names.
5464 51. Added "extra_headers" argument to the "mail" command in filter files.
5466 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5469 53. Installed eximstats 1.29.
5471 54. Added transport_filter_timeout as a generic transport option.
5473 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5474 Cc: header lines. This was required by RFC 822, but it not required by RFC
5477 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5478 incoming messages that did not have them. Now it does so only if the
5479 message originates locally, that is, if there is no associated remote host
5480 address. When Resent- header lines are present, this applies to the Resent-
5481 lines rather than the non-Resent- lines.
5483 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5484 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5486 58. Messages for configuration errors now include the name of the main
5487 configuration file - useful now that there may be more than one file in a
5488 list (.included file names were always shown).
5490 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5491 for those rare installations that do not start the daemon as root or run it
5492 setuid root. I've cut out the call to initgroups() if the daemon is not
5495 60. The Exim user and group can now be bound into the binary as text strings
5496 that are looked up at the start of Exim's processing.
5498 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5500 62. Added $mailstore_basename variable.
5502 63. Installed patch to sieve.c from Michael Haardt.
5504 64. When Exim failed to open the panic log after failing to open the main log,
5505 the original message it was trying to log was written to stderr and debug
5506 output, but if they were not available (the usual case in production), it
5507 was lost. Now it is written to syslog before the two lines that record the
5508 failures to open the logs.
5510 65. Users' Exim filters run in subprocesses under the user's uid. It is
5511 possible for a "deliver" command or an alias in a "personal" command to
5512 provoke an address rewrite. If logging of address rewriting is configured,
5513 this fails because the process is not running as root or exim. There may be
5514 a better way of dealing with this, but for the moment (because 4.30 needs
5515 to be released), I have disabled address rewrite logging when running a
5516 filter in a non-root, non-exim process.
5522 1. The buildconfig auxiliary program wasn't quoting the value set for
5523 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5524 not defined. This bug was masked in 4.22 by the effect that was fixed in
5527 2. Some messages that were rejected after a message id was allocated were
5528 shown as "incomplete" by exigrep. It no longer does this for messages that
5529 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5531 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5532 have allow_domain_literals set, the ID did not get logged in the <= line.
5533 Domain literals are now always recognized in Message-ID: header lines.
5535 4. The first argument for a ${extract expansion item is the key name or field
5536 number. Leading and trailing spaces in this item were not being ignored,
5537 causing some misleading effects.
5539 5. When deliver_drop_privilege was set, single queue runner processes started
5540 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5541 same command in the process it spins off) were not dropping privilege.
5543 6. When the daemon running as "exim" started a queue runner, it always
5544 re-executed Exim in the spun-off process. This is a waste of effort when
5545 deliver_drop_privilege is set. The new process now just calls the
5546 queue-runner function directly.
5552 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5555 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5556 ignored. Though the use of -oP was forcing the writing of a pid file, it
5557 was always written to the default place.
5559 3. If the message "no IP address found for host xxxx" is generated during
5560 incoming verification, it is now followed by identification of the incoming
5561 connection (so you can more easily find what provoked it).
5563 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5565 5. Added some features to "harden" Exim a bit more against certain attacks:
5567 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5568 be put in Local/Makefile. This is like the never_users runtime option,
5569 but it cannot be overridden. The default setting is "root".
5571 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5572 prefix string with which any file named in a -C command line option
5575 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5576 is retained for -C and -D only if the caller of Exim is root. Without
5577 it, the exim user may also use -C and -D and retain privilege.
5579 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5580 command line option is disabled.
5582 6. Macro names set by the -D option must start with an upper case letter, just
5583 like macro names defined in the configuration file.
5585 7. Added "dereference=" facility to LDAP.
5587 8. Two instances of the typo "uknown" in the source files are fixed.
5589 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5590 the Configure-Makefile script screwed up while processing it.
5592 10. Incorporated PCRE 4.4.
5594 11. The SMTP synchronization check was not operating right at the start of an
5595 SMTP session. For example, it could not catch a HELO sent before the client
5596 waited for the greeting. There is now a check for outstanding input at the
5597 point when the greeting is written. Because of the duplex, asynchronous
5598 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5599 way, but not yet received, when the check is performed.
5601 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5602 on TCP/IP sockets, because this apparently causes some broken clients to
5605 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5606 unchanged) from the Cygwin maintainer.
5608 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5609 was supported instead of testing for mbx. Effectively a typo.
5611 15. The spa authenticator server code was not checking that the input it
5612 received was valid base64.
5614 16. The debug output line for the "set" modifier in ACLs was not showing the
5615 name of the variable that was being set.
5617 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5619 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5620 lookup on its own. Something else was needed to trigger the lookup. For
5621 example, a match in host_lookup or the need for a host name in a host list.
5622 Now, if $sender_host_name is referenced and the host name has not yet been
5623 looked up, a lookup is performed. If the lookup fails, the variable remains
5624 empty, and $host_lookup_failed is set to "1".
5626 19. Added "eqi" as a case-independent comparison operator.
5628 20. The saslauthd authentication condition could segfault if neither service
5629 nor realm was specified.
5631 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5632 error message that was logged was misleading, and incoming SMTP
5633 connections were dropped. The message is now more accurate, and temporary
5634 errors are given to SMTP connections.
5636 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5637 (except RSET) with a 421 error, until QUIT is received. However, it was
5638 failing to send a response to QUIT.
5640 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5641 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5642 ACL is now run after the helo_try_verify_hosts code.
5644 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5645 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5646 case-independent, so other case variants are also recognized). Apparently
5647 some systems use these upper case variants.
5649 25. If more than two messages were waiting for the same host, and a transport
5650 filter was specified for the transport, Exim sent two messages over the
5651 same TCP/IP connection, and then failed with "socket operation on non-
5652 socket" when it tried to send the third.
5654 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5656 27. The extern definition of crypt16() in expand.c was not being excluded when
5657 the OS had its own crypt16() function.
5659 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5660 as a preferred synonym for return_size_limit, both as an option and as an
5663 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5665 30. Changed the default configuration ACL to relax the local part checking rule
5666 for addresses that are not in any local domains. For these addresses,
5667 slashes and pipe symbols are allowed within local parts, but the sequence
5668 /../ is explicitly forbidden.
5670 31. SPA server authentication was not clearing the challenge buffer before
5673 32. log_message in a "warn" ACL statement was writing to the reject log as
5674 well as to the main log, which contradicts the documentation and doesn't
5675 seem right (because no rejection is happening). So I have stopped it.
5677 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5678 However, I am unable to do any testing of this.
5680 34. Fixed an infelicity in the appendfile transport. When checking directories
5681 for a mailbox, to see if any needed to be created, it was accidentally
5682 using path names with one or more superfluous leading slashes; tracing
5683 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5685 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5686 contents are added to the log line that is written for every discarded
5687 recipient. (Previously a log_message setting was ignored.)
5689 36. The ${quote: operator now quotes the string if it is empty.
5691 37. The install script runs exim in order to find its version number. If for
5692 some reason other than non-existence or emptiness, which it checks, it
5693 could not run './exim', it was installing it with an empty version number,
5694 i.e. as "exim-". This error state is now caught, and the installation is
5697 38. An argument was missing from the function that creates an error message
5698 when Exim fails to connect to the socket for saslauthd authentication.
5699 This could cause Exim to crash, or give a corrupted message.
5701 39. Added isip, isip4, and isip6 to ${if conditions.
5703 40. The ACL variables $acl_xx are now saved with the message, and can be
5704 accessed later in routers, transports, and filters.
5706 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5707 strings in the file are not string-expanded.
5709 42. If a MAIL command specified a SIZE value that was too large to fit into an
5710 int variable, the check against message_size_limit failed. Such values are
5711 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5712 day this will have to be increased, but I don't think I want to be around
5713 when emails are that large.
5720 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5721 iconv() is not standard in FreeBSD.
5723 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5724 IPv6 enabled. The observed symptom was a segmentation fault on return from
5725 the function os_common_find_running_interfaces() in src/os.c.
5727 3. In the check_special_case() function in daemon.c I had used "errno" as an
5728 argument name, which causes warnings on some systems. This was basically a
5729 typo, since it was named "eno" in the comments!
5731 4. The code that waits for the clock to tick (at a resolution of some fraction
5732 of a second) so as to ensure message-id uniqueness was always waiting for
5733 at least one whole tick, when it could have waited for less. [This is
5734 almost certainly not relevant at current processor speeds, where it is
5735 unlikely to ever wait at all. But we try to future-proof.]
5737 5. The function that sleeps for a time interval that includes fractions of a
5738 second contained a race. It did not block SIGALRM between setting the
5739 timer, and suspending (a couple of lines later). If the interval was short
5740 and the sigsuspend() was delayed until after it had expired, the suspension
5741 never ended. On busy systems this could lead to processes getting stuck for
5744 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5745 process, before it forks any delivery processes. The open lookup caching
5746 mechanism meant that the open file or database connection was passed into
5747 the delivery process. The problem was that delivery processes always tidy
5748 up cached lookup data. This could cause a problem for the next delivery
5749 process started by the queue runner, because the external queue runner
5750 process does not know about the closure. So the next delivery process
5751 still has data in the lookup cache. In the case of a file lookup, there was
5752 no problem because closing a file descriptor in a subprocess doesn't affect
5753 the parent. However, if the lookup was caching a connection to a database,
5754 the connection was closed, and the second delivery process was likely to
5755 see errors such as "PGSQL: query failed: server closed the connection
5756 unexpectedly". The problem has been fixed by closing all cached lookups
5757 in a queue runner before running a delivery process.
5759 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5760 seem to have the "const" qualifier which it has on other OS. I've
5763 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5764 *of the same type* (client or server) with the same public name that an
5765 error should be diagnosed.
5767 9. When Exim looked up a host name for an IP address, but failed to find the
5768 original IP address when looking up the host name (a safety check), it
5769 output the message "<ip address> does not match any IP for NULL", which was
5770 confusing, to say the least. The bug was that the host name should have
5771 appeared instead of "NULL".
5773 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5774 user that is built into the binary, and the -C or -D options is used, root
5775 privilege is dropped before the configuration file is read. In addition,
5776 logging is switched to stderr instead of the normal log files. If the
5777 configuration then re-defines the Exim user, the unprivileged environment
5778 is probably not what is expected, so Exim logs a panic warning message (but
5781 However, if deliver_drop_privilege is set, the unprivileged state may well
5782 be exactly what is intended, so the warning has been cut out in that case,
5783 and Exim is allowed to try to write to its normal log files.
5789 1. smtp_return_error_details was not giving details for temporary sender
5790 or receiver verification errors.
5792 2. Diagnose a configuration error if two authenticators have the same public
5795 3. Exim used not to create the message log file for a message until the first
5796 delivery attempt. This could be confusing when incoming messages were held
5797 for policy or load reasons. The message log file is now created at the time
5798 the message is received, and an initial "Received" line is written to it.
5800 4. The automatically generated man page for command line options had a minor
5801 bug that caused no ill effects; however, a more serious problem was that
5802 the procedure for building the man page automatically didn't always
5803 operate. Consequently, release 4.20 contains an out-of-date version. This
5804 shouldn't happen again.
5806 5. When building Exim with embedded Perl support, the script that builds the
5807 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5808 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5810 6. The freeze_tell option was not being used for messages that were frozen on
5811 arrival, either by an ACL or by local_scan().
5813 7. Added the smtp_incomplete_transaction log selector.
5815 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5816 was accepting AUTH without a new EHLO.
5818 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5819 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5822 10. Logging of TCP/IP connections (when configured) now happens in the main
5823 daemon process instead of the child process, so that the TCP/IP connection
5824 count is more accurate (but it can never be perfect).
5826 11. The use of "drop" in a nested ACL was not being handled correctly in the
5827 outer ACL. Now, if condition failure induced by the nested "drop" causes
5828 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5829 or "require"), the connection is dropped.
5831 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5832 that yield "discard" can now be used with an "accept" or a "discard" verb,
5833 but an error is generated for any others (because I can't see a useful way
5834 to define what should happen).
5836 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5837 are now processed in the same way as lines in the Exim configuration file.
5838 In particular, continuation lines are supported.
5840 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5842 15. Added -ti meaning -t -i.
5844 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5845 domains, and warn by logging if others are found.
5847 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5848 returned from GIFCONF when Exim is trying to find the list of interfaces on
5849 a host. The code in os.c has been modified to copy each ifreq to an aligned
5850 structure in all cases.
5852 Also, in some cases, the returned ifreq's were being copied to a 'struct
5853 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5854 means the last couple of bytes of an IPv6 address could be chopped if the
5855 ifreq contained only a normal sockaddr (14 bytes storage).
5857 18. Named domain lists were not supported in the hosts_treat_as_local option.
5858 An entry such as +xxxx was not recognized, and was treated as a literal
5861 19. Ensure that header lines added by a DATA ACL are included in the reject log
5862 if the ACL subsequently rejects the message.
5864 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5865 MD5 (which is deprecated).
5867 21. When testing a filter file using -bf, Exim was writing a message when it
5868 took the sender from a "From " line in the message, but it was not doing so
5869 when it took $return_path from a Return-Path: header line. It now does.
5871 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5872 with a valid header line field name (a series of printing characters
5873 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5875 23. Changed "disc" in the source to "disk" to conform to the documentation and
5876 the book and for uniformity.
5878 24. Ignore Sendmail's -Ooption=value command line item.
5880 25. When execve() failed while trying to run a command in a pipe transport,
5881 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5882 could be confused with a return value of 69 from the command itself. This
5883 has been changed to 127, the value the shell returns if it is asked to run
5884 a non-existent command. The wording for the related log line suggests a
5885 non-existent command as the problem.
5887 26. If received_header_text expands to an empty string, do not add a Received:
5888 header line to the message. (Well, it adds a token one on the spool, but
5889 marks it "old" so that it doesn't get used or transmitted.)
5891 27. Installed eximstats 1.28 (addition of -nt option).
5893 28. There was no check for failure on the call to getsockname() in the daemon
5894 code. This can fail if there is a shortage of resources on the system, with
5895 ENOMEM, for example. A temporary error is now given on failure.
5897 29. Contrary to the C standard, it seems that in some environments, the
5898 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5899 program. Exim now does this explicitly; it affects the formatting of
5900 timestamps using strftime().
5902 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5903 complaints. I've now initialized all the variables, to avoid this.
5905 32. Header lines added by a system filter were not being "seen" during
5906 transport-time rewrites.
5908 33. The info_callback() function passed to OpenSSL is set up with type void
5909 (*)(SSL *, int, int), as described somewhere. However, when calling the
5910 function (actually a macro) that sets it up, the type void(*)() is
5911 expected. I've put in a cast to prevent warnings from picky compilers.
5913 34. If a DNS black list lookup found a CNAME record, but there were no A
5914 records associated with the domain it pointed at, Exim crashed.
5916 35. If a DNS black list lookup returned more than one A record, Exim ignored
5917 all but the first. It now scans all returned addresses if a particular IP
5918 value is being sought. In this situation, the contents of the
5919 $dnslist_value variable are a list of all the addresses, separated by a
5922 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5923 to accept a host name and all its aliases if the forward lookup for any of
5924 them yielded the IP address of the incoming connection. Now it accepts only
5925 those names whose forward lookup yields the correct IP address. Any other
5926 names are discarded. This closes a loophole whereby a rogue DNS
5927 administrator could create reverse DNS records to break through a
5928 wildcarded host restriction in an ACL.
5930 37. If a user filter or a system filter that ran in a subprocess used any of
5931 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5932 the wrong values were passed to the pipe command ($thisaddress had the
5933 value of $0, $0 had the value of $1, etc). This bug was introduced by
5934 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5936 38. Improved the line breaking for long SMTP error messages from ACLs.
5937 Previously, if there was no break point between 40 and 75 characters, Exim
5938 left the rest of the message alone. Two changes have been made: (a) I've
5939 reduced the minimum length to 35 characters; (b) if it can't find a break
5940 point between 35 and 75 characters, it looks ahead and uses the first one
5941 that it finds. This may give the occasional overlong line, but at least the
5942 remaining text gets split now.
5944 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5945 file descriptors might be low, and that setting 1000 would always raise it.
5946 It turns out that in some environments, the limit is already over 1000 and
5947 that lowering it causes trouble. So now Exim takes care not to decrease it.
5949 40. When delivering a message, the value of $return_path is set to $sender_
5950 address at the start of routing (routers may change the value). By an
5951 oversight, this default was not being set up when an address was tested by
5952 -bt or -bv, which affected the outcome if any router or filter referred to
5955 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5956 log only when "message" or "log_message" are set. However, if one of the
5957 conditions was an address verification, or a call to a nested ACL, the
5958 messages generated by the underlying test were being passed through. This
5959 no longer happens. The underlying message is available in $acl_verify_
5960 message for both "message" and "log_message" expansions, so it can be
5961 passed through if needed.
5963 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5964 new expansion $bh_ to give the encoded byte string without charset
5965 translation. Translation happens only if iconv() is available; HAVE_ICONV
5966 indicates this at build time. HEADERS_CHARSET gives the charset to
5967 translate to; headers_charset can change it in the configuration, and
5968 "headers charset" can change it in an individual filter file.
5970 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5971 that creates RFC 2047 encoded "words" labels them as that charset instead
5972 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5973 expansion operator; (ii) when Exim creates a From: line for a local
5974 message; (iii) when a header line is rewritten to include a "phrase" part.
5976 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5977 buggy, causing it to skip the first lines of messages whose message ID
5978 ended in 'D'. This would not have bitten before Exim release 4.14, because
5979 message IDs were unlikely to end in 'D' before then. The effect was to have
5980 incorrect size information for certain domains.
5982 45. #include "config.h" was missing at the start of the crypt16.c module. This
5983 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5986 46. If there was a timeout during a "random" callout check, Exim treated it as
5987 a failure of the random address, and carried on sending RSET and the real
5988 address. If the delay was just some slowness somewhere, the response to the
5989 original RCPT would be taken as a response to RSET and so on, causing
5990 mayhem of various kinds.
5992 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5993 when I implemented it. It didn't allow for the fact that some option values
5994 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5995 the right test for positive values.
5997 48. Domain names in DNS records are case-independent. Exim always looks them up
5998 in lower case. Some resolvers return domain names in exactly the case they
5999 appear in the zone file, that is, they may contain uppercase letters. Not
6000 all resolvers do this - some return always lower case. Exim was treating a
6001 change of case by a resolver as a change of domain, similar to a widening
6002 of a domain abbreviation. This triggered its re-routing code and so it was
6003 trying to route what was effectively the same domain again. This normally
6004 caused routing to fail (because the router wouldn't handle the domain
6005 twice). Now Exim checks for this case specially, and just changes the
6006 casing of the domain that it ultimately uses when it transmits the message
6009 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6012 50. If a filter generated a file delivery with a non-absolute name (possible if
6013 no home directory exists for the router), the forbid_file option was not
6016 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6017 the existing equality matching.
6019 52. Exim was using ints instead of ino_t variables in some places where it was
6020 dealing with inode numbers.
6022 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6023 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6024 called TMPDIR, and if it finds it is different, it changes its value.
6026 54. The smtp_printf() function is now made available to local_scan() so
6027 additional output lines can be written before returning. There is also an
6028 smtp_fflush() function to enable the detection of a dropped connection.
6029 The variables smtp_input and smtp_batched_input are exported to
6032 55. Changed the default runtime configuration: the message "Unknown user"
6033 has been removed from the ACL, and instead placed on the localuser router,
6034 using the cannot_route_message feature. This means that any verification
6035 failures that generate their own messages won't get overridden. Similarly,
6036 the "Unrouteable address" message that was in the ACL for unverifiable
6037 relay addresses has also been removed.
6039 56. Added hosts_avoid_esmtp to the smtp transport.
6041 57. The exicyclog script was not checking for the esoteric option
6042 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6043 will work only if exicyclog is run under the appropriate euid.
6045 58. Following a discussion on the list, the rules by which Exim recognises line
6046 endings on incoming messages have been changed. The -dropcr and drop_cr
6047 options are now no-ops, retained only for backwards compatibility. The
6048 following line terminators are recognized: LF CRLF CR. However, special
6049 processing applies to CR:
6051 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6052 nor a local message in the state where . is a terminator.
6054 (ii) If a bare CR is encountered in a header line, an extra space is added
6055 after the line terminator so as not to end the header. The reasoning
6056 behind this is that bare CRs in header lines are most likely either
6057 to be mistakes, or people trying to play silly games.
6059 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6060 was being incorrectly given as 18 bytes larger than it should have been.
6061 This is a VOB (very old bug).
6063 60. This may never have affected anything current, but just in case it has:
6064 When the local host is found other than at the start of a list of hosts,
6065 the local host, those with the same MX, and any that follow, are discarded.
6066 When the list in question was part of a longer list of hosts, the following
6067 hosts (not currently being processed) were also being discarded. This no
6068 longer happens. I'm not sure if this situation could ever has previously
6071 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6074 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6075 header. This is something that is recommended in a new Internet Draft, and
6076 is something that is documented as being done by Sendmail. There are two
6077 possible values. For messages generated by the autoreply transport, Exim
6080 Auto-Submitted: auto-replied
6082 whereas for all other generated messages (e.g. bounces) it adds
6084 Auto-Submitted: auto-generated
6086 63. The "personal" condition in filters now includes a test for the
6087 Auto-Submitted: header. If it contains the string "auto-" the message it
6088 not considered personal.
6090 64. Added rcpt_include_affixes as a generic transport option.
6092 65. Added queue_only_override (default true).
6094 66. Added the syslog_duplication option.
6096 67. If what should have been the first header line of a message consisted of
6097 a space followed by a colon, Exim was mis-interpreting it as a header line.
6098 It isn't of course - it is syntactically invalid and should therefore be
6099 treated as the start of the message body. The misbehaviour could have
6100 caused a number of strange effects, including loss of data in subsequent
6101 header lines, and spool format errors.
6103 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6104 client host had authenticated. This control can now be exercised by an ACL
6105 for more flexibility.
6107 69. By default, callouts do not happen when testing with -bh. There is now a
6108 variant, -bhc, which does actually run the callout code, including
6109 consulting and updating the callout cache.
6111 70. Added support for saslauthd authentication, courtesy of Alexander
6114 71. If statvfs() failed on the spool or log directories while checking their
6115 size for availability, Exim confusingly gave the error "space shortage".
6116 Furthermore, in debugging mode it crashed with a floating point exception.
6117 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6118 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6119 serious problem, Exim now writes to the main and panic logs when this
6120 happens, with details of the failure. It then refuses to accept the
6121 incoming message, giving the message "spool directory problem" or "log
6122 directory problem" with a 421 code for SMTP messages.
6124 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6125 0, 1, and 2 exist, because some OS complain for execs without them (see
6126 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6127 descriptors. However, the code omitted to check that the open succeeded,
6128 causing mysterious errors if for some reason the permissions on /dev/null
6129 got screwed. Now Exim writes a message to the main and panic logs, and
6130 bombs out if it can't open /dev/null.
6132 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6133 interact so that it is all more flexible. It is supposed to remain
6134 backwards compatible. Also added extra_local_interfaces.
6136 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6137 to bomb out with an assertion failure - to the client this appears as a
6138 connection drop. This problem occurs in the part of the code that was taken
6139 from the Samba project. Fortunately, the assertion is in a very simple
6140 function, so I have fixed this by reproducing the function inline in the
6141 one place where it is called, and arranging for authentication to fail
6142 instead of killing the process with assert().
6144 75. The SPA client code was not working when the server requested OEM rather
6145 than Unicode encoding.
6147 76. Added code to make require_files with a specific uid setting more usable in
6148 the case where statting the file as root fails - usually a non-root-mounted
6149 NFS file system. When this happens and the failure is EACCES, Exim now
6150 forks a subprocess and does the per-uid checking as the relevant uid.
6152 77. Added process_log_path.
6154 78. If log_file_path was not explicitly set, a setting of check_log_space or
6155 check_log_inodes was ignored.
6157 79. If a space check for the spool or log partitions fails, the incident is now
6158 logged. Of course, in the latter case the data may get lost...
6160 80. Added the %p formatting code to string_format() so that it can be used to
6161 print addresses in debug_print(). Adjusted all the address printing in the
6162 debugging in store.c to use %p rather than %d.
6164 81. There was a concern that a line of code in smtp_in.c could overflow a
6165 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6166 initially expressed, the concern was not well-founded, because trailing
6167 spaces are removed early. However, if the trailing spaces were followed by
6168 a NULL, they did not get removed, so the overflow was possible. Two fixes
6171 (a) I re-wrote the offending code in a cleaner fashion.
6172 (b) If an incoming SMTP command contains a NULL character, it is rejected
6175 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6176 runs initgroups(), so that if the Exim user is in any additional groups,
6177 they will be used during message reception.
6183 The change log for 4.20 and earlier releases has been archived.
projects / user / henk / code / exim.git / blob