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
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and backup-only
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
67 TL/01 When running the test suite, the README says that variables such as
68 no_msglog_check are global and can be placed anywhere in a specific
69 test's script, however it was observed that placement needed to be near
70 the beginning for it to behave that way. Changed the runtest perl
71 script to read through the entire script once to detect and set these
72 variables, reset to the beginning of the script, and then run through
73 the script parsing/test process like normal.
75 TL/02 The BSD's have an arc4random API. One of the functions to induce
76 adding randomness was arc4random_stir(), but it has been removed in
77 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
78 function when detected.
80 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
81 cause callback expansion.
83 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
84 syntax errors in an expansion can be treated as a string instead of
85 logging or causing an error, due to the internal use of bool_lax
86 instead of bool when processing it.
88 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
89 server certificates when making smtp deliveries.
91 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
93 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
95 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
97 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
98 Merged patch from Sebastian Wiedenroth.
100 JH/05 Fix results-pipe from transport process. Several recipients, combined
101 with certificate use, exposed issues where response data items split
102 over buffer boundaries were not parsed properly. This eventually
103 resulted in duplicates being sent. This issue only became common enough
104 to notice due to the introduction of conection certificate information,
105 the item size being so much larger. Found and fixed by Wolfgang Breyha.
107 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
108 size buffer was used, resulting in syntax errors when an expansion
111 JH/07 Add support for directories of certificates when compiled with a GnuTLS
112 version 3.3.6 or later.
114 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
115 is EXPERIMENTAL_EVENT, the main-configuration and transport options
116 both become "event_action", the variables become $event_name, $event_data
117 and $event_defer_errno. There is a new variable $verify_mode, usable in
118 routers, transports and related events. The tls:cert event is now also
119 raised for inbound connections, if the main configuration event_action
122 TL/06 In test suite, disable OCSP for old versions of openssl which contained
123 early OCSP support, but no stapling (appears to be less than 1.0.0).
125 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
126 server certificate names available under the smtp transport option
127 "tls_verify_cert_hostname" now do not permit multi-component wildcard
130 JH/10 Time-related extraction expansions from certificates now use the main
131 option "timezone" setting for output formatting, and are consistent
132 between OpenSSL and GnuTLS compilations. Bug 1541.
134 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
135 encoded parameter in the incoming message. Bug 1558.
137 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
138 include certificate info, eximon was claiming there were spoolfile
141 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
143 JH/14 Log delivery-related information more consistently, using the sequence
144 "H=<name> [<ip>]" wherever possible.
146 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
147 are problematic for Debian distribution, omit them from the release
150 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
152 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
155 JH/17 Fix a null-indirection in certextract expansions when a nondefault
156 output list separator was used.
161 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
162 checkers that were complaining about end of non-void function with no
165 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
166 This was a regression intruduced in 4.83 by another bugfix.
168 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
170 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
171 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
177 TF/01 Correctly close the server side of TLS when forking for delivery.
179 When a message was received over SMTP with TLS, Exim failed to clear up
180 the incoming connection properly after forking off the child process to
181 deliver the message. In some situations the subsequent outgoing
182 delivery connection happened to have the same fd number as the incoming
183 connection previously had. Exim would try to use TLS and fail, logging
184 a "Bad file descriptor" error.
186 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
187 utilities have not been installed.
189 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
190 temporary space as the ACL may create new global variables.
192 TL/01 LDAP support uses per connection or global context settings, depending
193 upon the detected version of the libraries at build time.
195 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
196 to extract and use the src ip:port in logging and expansions as if it
197 were a direct connection from the outside internet. PPv2 support was
198 updated based on HAProxy spec change in May 2014.
200 JH/02 Add ${listextract {number}{list}{success}{fail}}.
202 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
203 Properly escape header and check for NULL return.
205 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
208 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
210 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
211 characters in header names, implemented as a verify condition.
212 Contributed by Michael Fischer v. Mollard.
214 TL/05 Rename SPF condition results err_perm and err_temp to standardized
215 results permerror and temperror. Previous values are deprecated but
216 still accepted. In a future release, err_perm and err_temp will be
217 completely removed, which will be a backward incompatibility if the
218 ACL tests for either of these two old results. Patch contributed by
219 user bes-internal on the mailing list.
221 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
223 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
224 selectors, in both main and reject logs.
226 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
229 JH/07 Add malware type "sock" for talking to simple daemon.
231 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
233 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
234 routers/transports under cutthrough routing.
236 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
237 numbers. Touch up "bool" conditional to keep the same definition.
239 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
241 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
243 JH/12 Expand items in router/transport headers_add or headers_remove lists
244 individually rather than the list as a whole. Bug 1452.
246 Required for reasonable handling of multiple headers_ options when
247 they may be empty; requires that headers_remove items with embedded
248 colons must have them doubled (or the list-separator changed).
250 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
251 view the policy declared in the DMARC record. Currently, $dmarc_status
252 is a combined value of both the record presence and the result of the
255 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
257 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
258 dnslookup router and the smtp transport (applying to the forward
261 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
262 of ldap servers used for a specific lookup. Patch provided by Heiko
265 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
266 New variable $lookup_dnssec_authenticated for observability.
268 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
269 Patch submitted by Lars Timman.
271 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
273 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
274 Requires trusted mode and valid format message id, aborts otherwise.
275 Patch contributed by Heiko Schlichting.
277 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
278 certextract with support for various fields. Bug 1358.
280 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
281 is requested by default, modifiable by smtp transport option
284 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
285 operate on certificate variables to give certificate fingerprints
286 Also new ${sha256:cert_variable}.
288 JH/23 The PRDR feature is moved from being Experimental into the mainline.
290 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
291 Christian Aistleitner.
293 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
295 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
296 file. Patch from Wolfgang Breyha.
298 JH/25 Expand the coverage of the delivery $host and $host_address to
299 client authenticators run in verify callout. Bug 1476.
301 JH/26 Port service names are now accepted for tls_on_connect_ports, to
302 align with daemon_smtp_ports. Bug 72.
304 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
305 support and error reporting did not work properly.
307 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
308 and is readable. Patch from Andrew Colin Kissa.
310 TL/14 Enhance documentation of ${run expansion and how it parses the
311 commandline after expansion, particularly in the case when an
312 unquoted variable expansion results in an empty value.
314 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
316 PP/02 Fix internal collision of T_APL on systems which support RFC3123
317 by renaming away from it. Addresses GH issue 15, reported by
320 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
322 TL/15 SECURITY: prevent double expansion in math comparison functions
323 (can expand unsanitized data). Not remotely exploitable.
330 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
332 PP/02 Make -n do something, by making it not do something.
333 When combined with -bP, the name of an option is not output.
335 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
338 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
339 $sender_host_name and config options to manage this, and basic check
342 PP/05 DSCP support for outbound connections and control modifier for inbound.
344 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
345 (Only plugin which currently uses this is kerberos4, which nobody should
346 be using, but we should make it available and other future plugins might
347 conceivably use it, even though it would break NAT; stuff *should* be
348 using channel bindings instead).
350 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
351 name; added for Sendmail compatibility; requires admin caller.
352 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
353 just ignore it); requires trusted caller.
354 Also parse but ignore: -Ac -Am -X<logfile>
357 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
359 TL/02 Add +smtp_confirmation as a default logging option.
361 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
362 Patch by Magnus Holmgren from 2007-02-20.
364 TL/04 Bugzilla 1281 - Spec typo.
365 Bugzilla 1283 - Spec typo.
366 Bugzilla 1290 - Spec grammar fixes.
368 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
370 TL/06 Add Experimental DMARC support using libopendmarc libraries.
372 TL/07 Fix an out of order global option causing a segfault. Reported to dev
373 mailing list by by Dmitry Isaikin.
375 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
377 JH/02 Support "G" suffix to numbers in ${if comparisons.
379 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
381 NM/01 Bugzilla 1197 - Spec typo
382 Bugzilla 1196 - Spec examples corrections
384 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
386 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
387 gnutls_enable_pkcs11, but renamed to more accurately indicate its
390 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
391 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
393 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
394 "acl {{name}{arg}...}", and optional args on acl condition
397 JH/05 Permit multiple router/transport headers_add/remove lines.
399 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
401 JH/07 Avoid using a waiting database for a single-message-only transport.
402 Performance patch from Paul Fisher. Bugzilla 1262.
404 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
407 JH/09 Add $headers_added variable, with content from use of ACL modifier
408 add_header (but not yet added to the message). Bugzilla 199.
410 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
411 Pulled from Bugzilla 817 by Wolfgang Breyha.
413 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
415 (nb: this is the same fix as in Exim 4.80.1)
417 JH/11 Add A= logging on delivery lines, and a client_set_id option on
420 JH/12 Add optional authenticated_sender logging to A= and a log_selector
423 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
425 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
426 advertise SMTP AUTH mechanism to us, instead of a generic
427 protocol violation error. Also, make Exim more robust to bad
428 data from the Dovecot auth socket.
430 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
432 When a queue runner is handling a message, Exim first routes the
433 recipient addresses, during which it prunes them based on the retry
434 hints database. After that it attempts to deliver the message to
435 any remaining recipients. It then updates the hints database using
438 So if a recipient address works intermittently, it can get repeatedly
439 deferred at routing time. The retry hints record remains fresh so the
440 address never reaches the final cutoff time.
442 This is a fairly common occurrence when a user is bumping up against
443 their storage quota. Exim had some logic in its local delivery code
444 to deal with this. However it did not apply to per-recipient defers
445 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
447 This change adds a proper retry rule check during routing so that the
448 final cutoff time is checked against the message's age. We only do
449 this check if there is an address retry record and there is not a
450 domain retry record; this implies that previous attempts to handle
451 the address had the retry_use_local_parts option turned on. We use
452 this as an approximation for the destination being like a local
453 delivery, as in LMTP.
455 I suspect this new check makes the old local delivery cutoff check
456 redundant, but I have not verified this so I left the code in place.
458 TF/02 Correct gecos expansion when From: is a prefix of the username.
460 Test 0254 submits a message to Exim with the header
464 When I ran the test suite under the user fanf2, Exim expanded
465 the header to contain my full name, whereas it should have added
466 a Resent-Sender: header. It erroneously treats any prefix of the
467 username as equal to the username.
469 This change corrects that bug.
471 GF/01 DCC debug and logging tidyup
472 Error conditions log to paniclog rather than rejectlog.
473 Debug lines prefixed by "DCC: " to remove any ambiguity.
475 TF/03 Avoid unnecessary rebuilds of lookup-related code.
477 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
478 Bug spotted by Jeremy Harris; was flawed since initial commit.
479 Would have resulted in OCSP responses post-SNI triggering an Exim
480 NULL dereference and crash.
482 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
484 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
485 Bug detection, analysis and fix by Samuel Thibault.
486 Bugzilla 1331, Debian bug #698092.
488 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
490 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
491 Server implementation by Todd Lyons, client by JH.
492 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
493 config variable "prdr_enable" controls whether the server
494 advertises the facility. If the client requests PRDR a new
495 acl_data_smtp_prdr ACL is called once for each recipient, after
496 the body content is received and before the acl_smtp_data ACL.
497 The client is controlled by bolth of: a hosts_try_prdr option
498 on the smtp transport, and the server advertisement.
499 Default client logging of deliveries and rejections involving
500 PRDR are flagged with the string "PRDR".
502 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
503 fclose(). Diagnosis by Todd Lyons.
505 PP/17 Update configure.default to handle IPv6 localhost better.
506 Patch by Alain Williams (plus minor tweaks).
509 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
510 This is now consistent with GnuTLS, and is now documented: the
511 previous undocumented portable approach to treating the option as
512 unset was to force an expansion failure. That still works, and
513 an empty string is now equivalent.
515 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
516 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
517 not performing validation itself.
519 PP/20 Added force_command boolean option to pipe transport.
520 Patch from Nick Koston, of cPanel Inc.
522 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
525 TF/04 Added udpsend ACL modifer and hexquote expansion operator
527 PP/21 Fix eximon continuous updating with timestamped log-files.
528 Broken in a format-string cleanup in 4.80, missed when I repaired the
529 other false fix of the same issue.
530 Report and fix from Heiko Schlichting.
533 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
534 Report from Prashanth Katuri.
536 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
537 It's SecureTransport, so affects any MacOS clients which use the
538 system-integrated TLS libraries, including email clients.
540 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
541 using a MIME ACL for non-SMTP local injection.
542 Report and assistance in diagnosis by Warren Baker.
544 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
546 JH/16 Fix comparisons for 64b. Bugzilla 1385.
548 TL/09 Add expansion variable $authenticated_fail_id to keep track of
549 last id that failed so it may be referenced in subsequent ACL's.
551 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
554 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
555 ldap library initialization, allowing self-signed CA's to be
556 used. Also properly sets require_cert option later in code by
557 using NULL (global ldap config) instead of ldap handle (per
558 session). Bug diagnosis and testing by alxgomz.
560 TL/12 Enhanced documentation in the ratelimit.pl script provided in
561 the src/util/ subdirectory.
563 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
564 renamed to Transport Post Delivery Action by Jeremy Harris, as
567 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
568 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
569 redis_servers = needs to be configured which will be used by the redis
570 lookup. Patch from Warren Baker, of The Packet Hub.
572 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
574 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
575 hostname or reverse DNS when processing a host list. Used suggestions
576 from multiple comments on this bug.
578 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
580 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
581 Missed a few lines, added it to make the runtest require no keyboard
584 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
585 contains upper case chars. Make router use caseful_local_part.
587 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
588 support when GnuTLS has been built with p11-kit.
594 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
596 This, or similar/improved, will also be change PP/11 of 4.82.
602 PP/01 Handle short writes when writing local log-files.
603 In practice, only affects FreeBSD (8 onwards).
604 Bugzilla 1053, with thanks to Dmitry Isaikin.
606 NM/01 Bugzilla 949 - Documentation tweak
608 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
611 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
613 PP/02 Implemented gsasl authenticator.
615 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
617 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
618 `pkg-config foo` for cflags/libs.
620 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
621 with rest of GSASL and with heimdal_gssapi.
623 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
624 `pkg-config foo` for cflags/libs for the TLS implementation.
626 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
627 properties get this fed in as external SSF. A number of robustness
628 and debugging improvements to the cyrus_sasl authenticator.
630 PP/08 cyrus_sasl server now expands the server_realm option.
632 PP/09 Bugzilla 1214 - Log authentication information in reject log.
633 Patch by Jeremy Harris.
635 PP/10 Added dbmjz lookup type.
637 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
639 PP/12 MAIL args handles TAB as well as SP, for better interop with
640 non-compliant senders.
641 Analysis and variant patch by Todd Lyons.
643 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
644 Bug report from Lars Müller <lars@samba.org> (via SUSE),
645 Patch from Dirk Mueller <dmueller@suse.com>
647 PP/13 tls_peerdn now print-escaped for spool files.
648 Observed some $tls_peerdn in wild which contained \n, which resulted
649 in spool file corruption.
651 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
652 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
653 or write after TLS renegotiation, which otherwise led to messages
656 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
657 as a tracking header (ie: a signed header comes before the signature).
658 Patch from Wolfgang Breyha.
660 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
661 comma-sep list; embedded commas doubled.
663 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
665 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
667 Report and patch from Dmitry Banschikov.
669 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
670 Removed SSL_clear() after SSL_new() which led to protocol negotiation
671 failures. We appear to now support TLS1.1+ with Exim.
673 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
674 lets Exim select keys and certificates based upon TLS SNI from client.
675 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
676 before an outbound SMTP session. New log_selector, +tls_sni.
678 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
679 NULL dereference. Report and patch from Alun Jones.
681 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
682 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
683 Not seeing resolver debug output on NetBSD, but suspect this is a
684 resolver implementation change.
686 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
687 Left warnings. Added "eximon gdb" invocation mode.
689 PP/21 Defaulting "accept_8bitmime" to true, not false.
691 PP/22 Added -bw for inetd wait mode support.
693 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
694 locate the relevant includes and libraries. Made this the default.
696 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
697 Bugzilla 1246, report and most of solution from Tomasz Kusy.
699 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
700 This may cause build issues on older platforms.
702 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
703 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
704 gnutls_require_mac & gnutls_require_protocols (no longer supported).
705 Added SNI support via GnuTLS too.
706 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
708 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
710 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
711 Applied second patch from Janne, implementing suggestion to default
712 multiple-strings-in-record handling to match SPF spec.
714 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
716 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
717 read-only, out of scope).
718 Patch from Wolfgang Breyha, report from Stuart Northfield.
720 PP/29 Fix three issues highlighted by clang analyser static analysis.
721 Only crash-plausible issue would require the Cambridge-specific
722 iplookup router and a misconfiguration.
723 Report from Marcin Mirosław.
725 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
727 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
728 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
729 As part of this, removing so much warning spew let me fix some minor
730 real issues in debug logging.
732 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
733 assignment on my part. Fixed.
735 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
736 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
737 Janne Snabb (who went above and beyond: thank you).
739 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
740 string otherwise requires a connection and a bunch more work and it's
741 relatively easy to get wrong. Should also expose TLS library linkage
744 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
745 64-bit ${eval} (JH/03).
747 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
748 GNU libc to support some of the 64-bit stuff, should not lead to
749 conflicts. Defined before os.h is pulled in, so if a given platform
750 needs to override this, it can.
752 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
753 protection layer was required, which is not implemented.
754 Bugzilla 1254, patch from Wolfgang Breyha.
756 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
757 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
758 tls_dhparam take prime identifiers. Also unbreak combination of
759 OpenSSL+DH_params+TLSSNI.
761 PP/39 Disable SSLv2 by default in OpenSSL support.
767 PP/01 Solaris build fix for Oracle's LDAP libraries.
768 Bugzilla 1109, patch from Stephen Usher.
770 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
772 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
775 TF/02 Fix a couple more cases where we did not log the error message
776 when unlink() failed. See also change 4.74-TF/03.
778 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
779 lock up or crash if it happened to be inside a call to libc when it
780 got a SIGUSR1 from exiwhat.
782 The SIGUSR1 handler appends the current process status to the process
783 log which is later printed by exiwhat. It used to use the general
784 purpose logging code to do this, but several functions it calls are
785 not safe for signals.
787 The new output code in the SIGUSR1 handler is specific to the process
788 log, and simple enough that it's easy to inspect for signal safety.
789 Removing some special cases also simplifies the general logging code.
790 Removing the spurious timestamps from the process log simplifies
793 TF/04 Improved ratelimit ACL condition.
795 The /noupdate option has been deprecated in favour of /readonly which
796 has clearer semantics. The /leaky, /strict, and /readonly update modes
797 are mutually exclusive. The update mode is no longer included in the
798 database key; it just determines when the database is updated. (This
799 means that when you upgrde Exim will forget old rate measurements.)
801 Exim now checks that the per_* options are used with an update mode that
802 makes sense for the current ACL. For example, when Exim is processing a
803 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
804 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
805 must specify per_mail/readonly. If you omit the update mode it defaults to
806 /leaky where that makes sense (as before) or /readonly where required.
808 The /noupdate option is now undocumented but still supported for
809 backwards compatibility. It is equivalent to /readonly except that in
810 ACLs where /readonly is required you may specify /leaky/noupdate or
811 /strict/noupdate which are treated the same as /readonly.
813 A useful new feature is the /count= option. This is a generalization
814 of the per_byte option, so that you can measure the throughput of other
815 aggregate values. For example, the per_byte option is now equivalent
816 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
818 The per_rcpt option has been generalized using the /count= mechanism
819 (though it's more complicated than the per_byte equivalence). When it is
820 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
821 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
822 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
823 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
824 Note that using per_rcpt with a non-readonly update mode in more than
825 one ACL will cause the recipients to be double-counted. (The per_mail
826 and per_byte options don't have this problem.)
828 The handling of very low rates has changed slightly. If the computed rate
829 is less than the event's count (usually one) then this event is the first
830 after a long gap. In this case the rate is set to the same as this event's
831 count, so that the first message of a spam run is counted properly.
833 The major new feature is a mechanism for counting the rate of unique
834 events. The new per_addr option counts the number of different
835 recipients that someone has sent messages to in the last time period. It
836 behaves like per_rcpt if all the recipient addresses are different, but
837 duplicate recipient addresses do not increase the measured rate. Like
838 the /count= option this is a general mechanism, so the per_addr option
839 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
840 example, measure the rate that a client uses different sender addresses
841 with the options per_mail/unique=$sender_address. There are further
842 details in the main documentation.
844 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
846 TF/06 Removed a few PCRE remnants.
848 TF/07 Automatically extract Exim's version number from tags in the git
849 repository when doing development or release builds.
851 PP/02 Raise smtp_cmd_buffer_size to 16kB.
852 Bugzilla 879. Patch from Paul Fisher.
854 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
855 Heavily based on revision 40f9a89a from Simon Arlott's tree.
858 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
860 PP/05 Variable $av_failed, true if the AV scanner deferred.
861 Bugzilla 1078. Patch from John Horne.
863 PP/06 Stop make process more reliably on build failure.
864 Bugzilla 1087. Patch from Heiko Schlittermann.
866 PP/07 Make maildir_use_size_file an _expandable_ boolean.
867 Bugzilla 1089. Patch from Heiko Schlittermann.
869 PP/08 Handle ${run} returning more data than OS pipe buffer size.
870 Bugzilla 1131. Patch from Holger Weiß.
872 PP/09 Handle IPv6 addresses with SPF.
873 Bugzilla 860. Patch from Wolfgang Breyha.
875 PP/10 GnuTLS: support TLS 1.2 & 1.1.
877 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
880 PP/11 match_* no longer expand right-hand-side by default.
881 New compile-time build option, EXPAND_LISTMATCH_RHS.
882 New expansion conditions, "inlist", "inlisti".
884 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
886 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
888 PP/14 fix log_write() format string regression from TF/03.
889 Bugzilla 1152. Patch from Dmitry Isaikin.
895 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
897 PP/02 Harmonised TLS library version reporting; only show if debugging.
898 Layout now matches that introduced for other libraries in 4.74 PP/03.
900 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
902 PP/04 New "dns_use_edns0" global option.
904 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
907 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
908 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
910 TK/01 Updated PolarSSL code to 0.14.2.
911 Bugzilla 1097. Patch from Andreas Metzler.
913 PP/07 Catch divide-by-zero in ${eval:...}.
916 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
919 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
920 format-string attack -- SECURITY: remote arbitrary code execution.
922 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
923 time unintentionally subject to list matching rules, letting the header
924 cause arbitrary Exim lookups (of items which can occur in lists, *not*
925 arbitrary string expansion). This allowed for information disclosure.
927 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
928 INT_MIN/-1 -- value coerced to INT_MAX.
934 NM/01 Workround for PCRE version dependency in version reporting
937 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
938 This fixes portability to compilers other than gcc, notably
939 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
941 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
942 makefiles for portability to HP-UX and POSIX correctness.
944 PP/01 Permit LOOKUP_foo enabling on the make command-line.
945 Also via indented variable definition in the Makefile.
946 (Debugging by Oliver Heesakkers).
948 PP/02 Restore caching of spamd results with expanded spamd_address.
949 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
951 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
952 Improves build reliability. Fix from: Frank Elsner
954 NM/02 Fix wide character breakage in the rfc2047 coding
955 Fixes bug 1064. Patch from Andrey N. Oktyabrski
957 NM/03 Allow underscore in dnslist lookups
958 Fixes bug 1026. Patch from Graeme Fowler
960 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
961 Code patches from Adam Ciarcinski of NetBSD.
963 NM/04 Fixed exiqgrep to cope with mailq missing size issue
966 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
967 is logged, to avoid truncation. Patch from John Horne.
969 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
970 Patch from Jakob Hirsch.
972 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
973 SQL string expansion failure details.
974 Patch from Andrey Oktyabrski.
976 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
977 Patch from Simon Arlott.
979 PP/09 New lookups functionality failed to compile on old gcc which rejects
980 extern declarations in function scope.
981 Patch from Oliver Fleischmann
983 PP/10 Use sig_atomic_t for flags set from signal handlers.
984 Check getgroups() return and improve debugging.
985 Fixed developed for diagnosis in bug 927 (which turned out to be
988 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
989 Patch from Mark Zealey.
991 PP/12 Bugzilla 1056: Improved spamd server selection.
992 Patch from Mark Zealey.
994 PP/13 Bugzilla 1086: Deal with maildir quota file races.
995 Based on patch from Heiko Schlittermann.
997 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
998 Patch from Uwe Doering, sign-off by Michael Haardt.
1000 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1001 variable declaration deep within a block. Bug and patch from
1004 PP/15 lookups-Makefile IRIX compatibilty coercion.
1006 PP/16 Make DISABLE_DKIM build knob functional.
1008 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1009 Patch by Simon Arlott
1011 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1012 variable argument macros. Our copy now differs from upstream.
1018 TF/01 Failure to get a lock on a hints database can have serious
1019 consequences so log it to the panic log.
1021 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1022 controlled using the smtp_confirmation log selector.
1024 TF/03 Include the error message when we fail to unlink a spool file.
1026 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1027 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1028 for maintaining out-of-tree patches for some time.
1030 PP/01 Bugzilla 139: Documentation and portability issues.
1031 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1032 Handle per-OS dynamic-module compilation flags.
1034 PP/02 Let /dev/null have normal permissions.
1035 The 4.73 fixes were a little too stringent and complained about the
1036 permissions on /dev/null. Exempt it from some checks.
1037 Reported by Andreas M. Kirchwitz.
1039 PP/03 Report version information for many libraries, including
1040 Exim version information for dynamically loaded libraries. Created
1041 version.h, now support a version extension string for distributors
1042 who patch heavily. Dynamic module ABI change.
1044 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1045 privilege escalation vulnerability whereby the Exim run-time user
1046 can cause root to append content of the attacker's choosing to
1049 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1052 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1053 If dropping privileges for untrusted macros, we disabled normal logging
1054 on the basis that it would fail; for the Exim run-time user, this is not
1055 the case, and it resulted in successful deliveries going unlogged.
1056 Fixed. Reported by Andreas Metzler.
1062 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1063 only prepend for the Resent-* case. Fixes regression introduced in
1064 Exim 4.70 by NM/22 for Bugzilla 607.
1066 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1067 increasing numbers of administrators be bitten by this.
1069 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1071 PP/03 Bugzilla 994: added openssl_options main configuration option.
1073 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1075 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1077 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1079 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1080 without a peer certificate, leading to a segfault because of an
1081 assumption that peers always have certificates. Be a little more
1082 paranoid. Problem reported by Martin Tscholak.
1084 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1085 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1086 NB: ClamAV planning to remove STREAM in "middle of 2010".
1087 CL also introduces -bmalware, various -d+acl logging additions and
1088 more caution in buffer sizes.
1090 PP/09 Implemented reverse_ip expansion operator.
1092 PP/10 Bugzilla 937: provide a "debug" ACL control.
1094 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1096 PP/12 Bugzilla 973: Implement --version.
1098 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1100 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1102 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1104 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1105 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1106 ignore trailing whitespace.
1108 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1110 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1113 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1114 Notification from Dr Andrew Aitchison.
1116 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1117 ExtendedDetectionInfo response format.
1118 Notification from John Horne.
1120 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1123 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1124 XSL and documented dependency on system catalogs, with examples of how
1127 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1130 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1131 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1132 configuration file which is writeable by the Exim user or group.
1134 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1135 of configuration files to cover files specified with the -C option if
1136 they are going to be used with root privileges, not just the default
1139 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1140 option (effectively making it always true).
1142 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1143 files to be used while preserving root privileges.
1145 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1146 that rogue child processes cannot use them.
1148 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1149 run-time user, instead of root.
1151 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1152 Exim run-time user without dropping privileges.
1154 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1155 result string, instead of calling string_vformat() twice with the same
1158 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1159 for other users. Others should always drop root privileges if they use
1160 -C on the command line, even for a whitelisted configure file.
1162 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1164 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1170 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1171 $data_path, and $header_path variables; fixed documentation bugs and
1174 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1175 exipick to access non-standard spools, including the "frozen" queue
1178 NM/01 Bugzilla 965: Support mysql stored procedures.
1179 Patch from Alain Williams
1181 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1183 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1184 Patch from Andreas Metzler
1186 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1187 Patch from Kirill Miazine
1189 NM/05 Bugzilla 671: Added umask to procmail example.
1191 JJ/03 installed exipick 20100323.0, fixing doc bug
1193 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1194 directory. Notification and patch from Dan Rosenberg.
1196 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1198 TK/02 Improve log output when DKIM signing operation fails.
1200 MH/01 Treat the transport option dkim_domain as a colon separated
1201 list, not as a single string, and sign the message with each element,
1202 omitting multiple occurences of the same signer.
1204 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1205 Bugzilla 985, 986. Patch by Simon Arlott
1207 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1208 Patch by Simon Arlott
1210 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1211 MBX locking. Notification from Dan Rosenberg.
1217 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1219 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1221 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1223 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1225 NM/04 Bugzilla 915: Flag broken perl installation during build.
1231 TK/01 Added patch by Johannes Berg that expands the main option
1232 "spamd_address" if it starts with a dollar sign.
1234 TK/02 Write list of recipients to X-Envelope-Sender header when building
1235 the mbox-format spool file for content scanning (suggested by Jakob
1238 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1239 (http://www.dcc-servers.net/) support via dccifd. Activated by
1240 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1242 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1243 by Mark Daniel Reidel <mr@df.eu>.
1245 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1246 When building exim an external PCRE library is now needed -
1247 PCRE is a system library on the majority of modern systems.
1248 See entry on PCRE_LIBS in EDITME file.
1250 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1251 conversation. Added nologin parameter to request.
1252 Patch contributed by Kirill Miazine.
1254 TF/01 Do not log submission mode rewrites if they do not change the address.
1256 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1258 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1259 log files in place. Contributed by Roberto Lima.
1261 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1263 TF/03 Bugzilla 615: When checking the local_parts router precondition
1264 after a local_part_suffix or local_part_prefix option, Exim now
1265 does not use the address's named list lookup cache, since this
1266 contains cached lookups for the whole local part.
1268 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1269 Robert Millan. Documentation is in experimental-spec.txt.
1271 TF/04 Bugzilla 668: Fix parallel build (make -j).
1273 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1275 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1276 Patch provided by Jan Srzednicki.
1278 TF/05 Leading white space used to be stripped from $spam_report which
1279 wrecked the formatting. Now it is preserved.
1281 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1282 that they are available at delivery time.
1284 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1286 TF/08 TLS error reporting now respects the incoming_interface and
1287 incoming_port log selectors.
1289 TF/09 Produce a more useful error message if an SMTP transport's hosts
1290 setting expands to an empty string.
1292 NM/06 Bugzilla 744: EXPN did not work under TLS.
1293 Patch provided by Phil Pennock.
1295 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1296 Patch provided by Richard Godbee.
1298 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1299 acl_smtp_notquit, added index entry.
1301 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1302 Patch provided by Eugene Bujak.
1304 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1305 accept(). Patch provided by Maxim Dounin.
1307 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1308 Patch provided by Phil Pennock.
1310 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1312 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1313 Patch provided by Brad "anomie" Jorsch.
1315 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1316 Patch provided by Dean Brooks.
1318 TK/05 Add native DKIM support (does not depend on external libraries).
1320 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1321 Patch provided by Graeme Fowler.
1323 NM/16 Bugzilla 851: Documentation example syntax fix.
1325 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1327 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1330 NM/19 Bugzilla 745: TLS version reporting.
1331 Patch provided by Phil Pennock.
1333 NM/20 Bugzilla 167: bool: condition support.
1334 Patch provided by Phil Pennock.
1336 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1337 clients. Patch provided by Phil Pennock.
1339 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1340 Patch provided by Brad "anomie" Jorsch.
1342 NM/23 Bugzilla 687: Fix misparses in eximstats.
1343 Patch provided by Heiko Schlittermann.
1345 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1346 Patch provided by Heiko Schlittermann.
1348 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1349 plus update to original patch.
1351 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1353 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1354 Patch provided by David Brownlee.
1356 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1358 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1360 NM/30 Bugzilla 888: TLS documentation bugfixes.
1362 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1364 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1365 Unlike the original bugzilla I have changed all shell scripts in src tree.
1367 NM/33 Bugzilla 898: Transport filter timeout fix.
1368 Patch by Todd Rinaldo.
1370 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1371 Patch by Serge Demonchaux.
1373 NM/35 Bugzilla 39: Base64 decode bug fixes.
1374 Patch by Jakob Hirsch.
1376 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1378 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1380 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1382 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1388 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1389 ALT-N's libdkim that I have put here:
1390 http://duncanthrax.net/exim-experimental/
1392 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1393 were called 'true' and it seems that C99 defines that as a reserved
1394 keyword to be used with 'bool' variable types. That means you could
1395 not include C99-style headers which use bools without triggering
1396 build errors in sieve.c.
1398 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1399 as mailq or other aliases. Changed the --help handling significantly
1400 to do whats expected. exim_usage() emits usage/help information.
1402 SC/01 Added the -bylocaldomain option to eximstats.
1404 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1406 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1408 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1414 PH/01 Another patch from the Sieve maintainer.
1416 PH/02 When an IPv6 address is converted to a string for single-key lookup
1417 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1418 dots are used instead of colons so that keys in lsearch files need not
1419 contain colons. This was done some time before quoting was made available
1420 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1421 (notated using the quote facility) so as to distinguish them from IPv4
1422 keys. This meant that lookups for IP addresses in host lists did not work
1423 for iplsearch lookups.
1425 This has been fixed by arranging for IPv6 addresses to be expressed with
1426 colons if the lookup type is iplsearch. This is not incompatible, because
1427 previously such lookups could never work.
1429 The situation is now rather anomolous, since one *can* have colons in
1430 ordinary lsearch keys. However, making the change in all cases is
1431 incompatible and would probably break a number of configurations.
1433 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1436 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1437 conversion specification without a maximum field width, thereby enabling
1438 a rogue spamd server to cause a buffer overflow. While nobody in their
1439 right mind would setup Exim to query an untrusted spamd server, an
1440 attacker that gains access to a server running spamd could potentially
1441 exploit this vulnerability to run arbitrary code as the Exim user.
1443 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1444 $primary_hostname instead of what libspf2 thinks the hosts name is.
1446 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1447 a directory entry by the name of the lookup key. Previously, if a
1448 symlink pointed to a non-existing file or a file in a directory that
1449 Exim lacked permissions to read, a lookup for a key matching that
1450 symlink would fail. Now it is enough that a matching directory entry
1451 exists, symlink or not. (Bugzilla 503.)
1453 PH/03 The body_linecount and body_zerocount variables are now exported in the
1456 PH/04 Added the $dnslist_matched variable.
1458 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1459 This means they are set thereafter only if the connection becomes
1462 PH/06 Added the client_condition to authenticators so that some can be skipped
1463 by clients under certain conditions.
1465 PH/07 The error message for a badly-placed control=no_multiline_responses left
1466 "_responses" off the end of the name.
1468 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1470 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1471 (without spaces) instead of just copying the configuration text.
1473 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1475 PH/11 Added $max_received_linelength.
1477 PH/12 Added +ignore_defer and +include_defer to host lists.
1479 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1480 way in which PCRE > 7.0 is built.
1482 PH/14 Implemented queue_only_load_latch.
1484 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1485 MAIL command. The effect was to mangle the value on 64-bit systems.
1487 PH/16 Another patch from the Sieve maintainer.
1489 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1491 PH/18 If a system quota error occurred while trying to create the file for
1492 a maildir delivery, the message "Mailbox is full" was not appended to the
1493 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1494 only to a quota excession during the actual writing of the file.
1496 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1497 characters?) which causes problems in log lines. The DN values are now
1498 passed through string_printing() before being added to log lines.
1500 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1501 and InterBase are left for another time.)
1503 PH/21 Added message_body_newlines option.
1505 PH/22 Guard against possible overflow in moan_check_errorcopy().
1507 PH/23 POSIX allows open() to be a macro; guard against that.
1509 PH/24 If the recipient of an error message contained an @ in the local part
1510 (suitably quoted, of course), incorrect values were put in $domain and
1511 $local_part during the evaluation of errors_copy.
1517 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1518 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1521 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1522 issue a MAIL command.
1524 PH/02 In an ACL statement such as
1526 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1528 if a client was not listed at all, or was listed with a value other than
1529 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1530 the condition was not true (as it should be), so access was not denied.
1531 The bug was that the ! inversion was incorrectly passed on to the second
1532 item. This has been fixed.
1534 PH/03 Added additional dnslists conditions == and =& which are different from
1535 = and & when the dns lookup returns more than one IP address.
1537 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1538 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1540 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1541 FSYNC, which compiles an option called disable_fsync that allows for
1542 bypassing fsync(). The documentation is heavily laced with warnings.
1544 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1546 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1547 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1548 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1549 including adding "make clean"; (3) Added -fPIC when compiling the test
1550 dynamically loaded module, to get rid of a warning.
1552 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1553 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1554 = 0s. The bug is otherwise harmless.
1556 PH/07 There was a bug in the dovecot authenticator such that the value of
1557 $auth1 could be overwritten, and so not correctly preserved, after a
1558 successful authentication. This usually meant that the value preserved by
1559 the server_setid option was incorrect.
1561 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1563 PH/09 Installed PCRE release 7.0.
1565 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1566 run for batched SMTP input. It is now run at the start of every message
1567 in the batch. While fixing this I discovered that the process information
1568 (output by running exiwhat) was not always getting set for -bs and -bS
1569 input. This is fixed, and it now also says "batched" for BSMTP.
1571 PH/11 Added control=no_pipelining.
1573 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1574 patch, slightly modified), and move the expansion of helo_data till after
1575 the connection is made in the smtp transport (so it can use these
1578 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1580 PH/14 Added log_selector = +pid.
1582 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1584 PH/16 Add ${if forany and ${if forall.
1586 PH/17 Added dsn_from option to vary the From: line in DSNs.
1588 PH/18 Flush SMTP output before performing a callout, unless control =
1589 no_callout_flush is set.
1591 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1592 was true (the default) a successful delivery failed to delete the retry
1593 item, thus causing premature timeout of the address. The bug is now
1596 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1598 PH/21 Long custom messages for fakedefer and fakereject are now split up
1599 into multiline reponses in the same way that messages for "deny" and
1600 other ACL rejections are.
1602 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1603 with slight modification.
1605 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1606 draft, changing the syntax and factoring some duplicate code".
1608 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1609 for deliveries of the second and subsequent messages over the same SMTP
1612 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1613 ${reduce, with only minor "tidies".
1615 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1617 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1618 expansion side effects.
1620 PH/27 When a message times out after an over-quota error from an Exim-imposed
1621 quota, the bounce message says "mailbox is full". This message was not
1622 being given when it was a system quota that was exceeded. It now should
1625 MH/03 Made $recipients available in local_scan(). local_scan() already has
1626 better access to the recipient list through recipients_list[], but
1627 $recipients can be useful in postmaster-provided expansion strings.
1629 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1630 in the case of a MAIL command with additional options following the
1631 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1632 were accidentally chopped off.
1634 PH/29 SMTP synchronization checks are implemented when a command is read -
1635 there is a check that no more input is waiting when there shouldn't be
1636 any. However, for some commands, a delay in an ACL can mean that it is
1637 some time before the response is written. In this time, more input might
1638 arrive, invalidly. So now there are extra checks after an ACL has run for
1639 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1640 pipelining has not been advertised.
1642 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1644 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1645 returned in the SMTP response when smtp_return_error_details was false.
1646 This has been fixed.
1648 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1649 instead of the C I/O that was originally supplied, because problems were
1650 reported on Solaris.
1652 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1653 Exim which did not show up earlier: it was assuming that a call to
1654 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1655 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1656 was a macro that became an assignment, so it seemed to work. This has
1657 changed to a proper function call with a void return, hence the compile
1658 error. Exim's code has been fixed.
1660 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1663 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1665 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1667 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1668 because it turns out to be too big or there is a timeout) there is no
1669 "Completed" line in the log. When some messages of this type were
1670 selected by exigrep, they were listed as "not completed". Others were
1671 picked up by some special patterns. I have improved the selection
1672 criteria to be more general.
1674 PH/38 The host_find_failed option in the manualroute router can now be set
1675 to "ignore", to completely ignore a host whose IP address cannot be
1676 found. If all hosts are ignored, the behaviour is controlled by the new
1677 host_all_ignored option.
1679 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1680 homing or because of multiple MX records with /mx) generated more than
1681 one IP address, and the following item turned out to be the local host,
1682 all the secondary addresses of the first item were incorrectly removed
1683 from the list, along with the local host and any following hosts (which
1684 is what is supposed to happen).
1686 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1687 whoever called Exim into the -H file. In the case of the daemon it was
1688 behaving confusingly. When first started, it used values for whoever
1689 started the daemon, but after a SIGHUP it used the Exim user (because it
1690 calls itself on a restart). I have changed the code so that it now always
1693 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1694 message are rejected with the same error (e.g. no authentication or bad
1695 sender address), and a DATA command is nevertheless sent (as can happen
1696 with PIPELINING or a stupid MUA), the error message that was given to the
1697 RCPT commands is included in the rejection of the DATA command. This is
1698 intended to be helpful for MUAs that show only the final error to their
1701 PH/42 Another patch from the Sieve maintainer.
1703 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1704 Eximstats - Fixed some broken HTML links and added missing column headers
1706 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1707 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1709 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1711 PH/43 Yet another patch from the Sieve maintainer.
1713 PH/44 I found a way to check for a TCP/IP connection going away before sending
1714 the response to the final '.' that terminates a message, but only in the
1715 case where the client has not sent further data following the '.'
1716 (unfortunately, this is allowed). However, in many cases there won't be
1717 any further data because there won't be any more messages to send. A call
1718 to select() can be used: if it shows that the input is "ready", there is
1719 either input waiting, or the socket has been closed. An attempt to read
1720 the next input character can distinguish the two cases. Previously, Exim
1721 would have sent an OK response which the client would never have see.
1722 This could lead to message repetition. This fix should cure that, at
1723 least in a lot of common cases.
1725 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1726 advertised in response to EHLO.
1732 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1733 fixed by 4.65/MH/01 (is this a record?) are fixed:
1735 (i) An empty string was always treated as zero by the numeric comparison
1736 operators. This behaviour has been restored.
1738 (ii) It is documented that the numeric comparison operators always treat
1739 their arguments as decimal numbers. This was broken in that numbers
1740 starting with 0 were being interpreted as octal.
1742 While fixing these problems I realized that there was another issue that
1743 hadn't been noticed. Values of message_size_limit (both the global option
1744 and the transport option) were treated as octal if they started with 0.
1745 The documentation was vague. These values are now always treated as
1746 decimal, and I will make that clear in the documentation.
1752 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1753 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1756 MH/01 Don't check that the operands of numeric comparison operators are
1757 integers when their expansion is in "skipping" mode (fixes bug
1758 introduced by 4.64-PH/07).
1760 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1761 child addresses, Exim now panics and dies. Previously, because the count
1762 is held in a short int, deliveries were likely to be lost. As such a
1763 large number of recipients for a single message is ridiculous
1764 (performance will be very, very poor), I have chosen to impose a limit
1765 rather than extend the field.
1771 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1772 leftover -K file (the existence of which was triggered by #402).
1773 While we were at it, introduced process PID as part of the -K
1774 filename. This should rule out race conditions when creating
1777 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1778 processing considerably. Previous code took too long for large mails,
1779 triggering a timeout which in turn triggers #401.
1781 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1782 in the DK code in transports.c. sendfile() is not really portable,
1783 hence the _LINUX specificness.
1785 TF/01 In the add_headers option to the mail command in an Exim filter,
1786 there was a bug that Exim would claim a syntax error in any
1787 header after the first one which had an odd number of characters
1790 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1791 callout verification, Exim cached a "reject" for the entire domain. This
1792 is correct for most verifications, but it is not correct for a recipient
1793 verification with use_sender or use_postmaster set, because in that case
1794 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1795 case of MAIL FROM:<> rejection from other early rejections (e.g.
1796 rejection of HELO). When verifying a recipient using a non-null MAIL
1797 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1798 Whatever the result of the callout, the value of the domain cache is
1799 left unchanged (for any other kind of callout, getting as far as trying
1800 RCPT means that the domain itself is ok).
1802 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1805 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1806 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1807 session, the daemon ignores ECONNECT errors and logs others; it now
1808 ignores EPIPE as well.
1810 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1811 (quoted-printable decoding).
1813 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1814 later the small subsequent patch to fix an introduced bug.
1816 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1818 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1820 PH/08 An error is now given if message_size_limit is specified negative.
1822 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1823 to be given (somewhat) arbitrary names.
1825 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1828 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1829 miscellaneous code fixes
1831 PH/10 Added the log_reject_target ACL modifier to specify where to log
1834 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1835 hostname. This is wrong, because it relates to the incoming message (and
1836 probably the interface on which it is arriving) and not to the outgoing
1837 callout (which could be using a different interface). This has been
1838 changed to use the value of the helo_data option from the smtp transport
1839 instead - this is what is used when a message is actually being sent. If
1840 there is no remote transport (possible with a router that sets up host
1841 addresses), $smtp_active_hostname is used.
1843 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1844 tweaks were necessary in order to get it to work (see also 21 below):
1845 (a) The code assumed that strncpy() returns a negative number on buffer
1846 overflow, which isn't the case. Replaced with Exim's string_format()
1848 (b) There were several signed/unsigned issues. I just did the minimum
1849 hacking in of casts. There is scope for a larger refactoring.
1850 (c) The code used strcasecmp() which is not a standard C function.
1851 Replaced with Exim's strcmpic() function.
1852 (d) The code set only $1; it now sets $auth1 as well.
1853 (e) A simple test gave the error "authentication client didn't specify
1854 service in request". It would seem that Dovecot has changed its
1855 interface. Fortunately there's a specification; I followed it and
1856 changed what the client sends and it appears to be working now.
1858 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1861 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1862 address A is aliased to B and C, where B exists and C does not. Without
1863 -v the output is "A verified" because verification stops after a
1864 successful redirection if more than one address is generated. However,
1865 with -v the child addresses are also verified. Exim was outputting "A
1866 failed to verify" and then showing the successful verification for C,
1867 with its parentage. It now outputs "B failed to verify", showing B's
1868 parentage before showing the successful verification of C.
1870 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1871 look up a TXT record in a specific list after matching in a combined
1874 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1875 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1876 they consult the DNS. I had assumed they would set it the way they
1877 wanted; and indeed my experiments on Linux seem to show that in some
1878 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1879 To be on the safe side, however, I have now made the interface to
1880 host_find_byname() similar to host_find_bydns(), with an argument
1881 containing the DNS resolver options. The host_find_byname() function now
1882 sets these options at its start, just as host_find_bydns() does. The smtp
1883 transport options dns_qualify_single and dns_search_parents are passed to
1884 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1885 of host_find_byname() use the default settings of RES_DEFNAMES
1886 (qualify_single) but not RES_DNSRCH (search_parents).
1888 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1889 spool_read_header() do less string testing, by means of a preliminary
1890 switch on the second character of optional "-foo" lines. (This is
1891 overdue, caused by the large number of possibilities that now exist.
1892 Originally there were few.) While I was there, I also converted the
1893 str(n)cmp tests so they don't re-test the leading "-" and the first
1894 character, in the hope this might squeeze out yet more improvement.
1896 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1897 flag allowing group syntax was set by the header_syntax check but not
1898 turned off, possible causing trouble later; (2) The flag was not being
1899 set at all for the header_verify test, causing "group"-style headers to
1900 be rejected. I have now set it in this case, and also caused header_
1901 verify to ignore an empty address taken from a group. While doing this, I
1902 came across some other cases where the code for allowing group syntax
1903 while scanning a header line wasn't quite right (mostly, not resetting
1904 the flag correctly in the right place). These bugs could have caused
1905 trouble for malformed header lines. I hope it is now all correct.
1907 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1908 with the "reply" argument non-NULL. The code, however (which originally
1909 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1910 but it didn't always do it. This confused somebody who was copying the
1911 code for some other use. I have removed all the tests.
1913 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1914 feature that was used to support insecure browsers during the U.S. crypto
1915 embargo. It requires special client support, and Exim is probably the
1916 only MTA that supported it -- and would never use it because real RSA is
1917 always available. This code has been removed, because it had the bad
1918 effect of slowing Exim down by computing (never used) parameters for the
1919 RSA_EXPORT functionality.
1921 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1922 authenticator to fail if there's a tab character in the incoming data
1923 (there should never be unless someone is messing about, as it's supposed
1924 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1925 if the connection is using TLS or if the remote IP is the same as the
1926 local IP, and the "valid-client-cert option" if a client certificate has
1929 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1930 authenticators. This can be used for authorization after authentication
1931 succeeds. (In the case of plaintext, it servers for both authentication
1934 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1935 if any retry times were supplied.
1937 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1938 connection, where there is no client IP address to check. In this
1939 situation, the verify now always succeeds.
1941 PH/25 Applied John Jetmore's -Mset patch.
1943 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1945 PH/27 In a string expansion for a processed (not raw) header when multiple
1946 headers of the same name were present, leading whitespace was being
1947 removed from all of them, but trailing whitespace was being removed only
1948 from the last one. Now trailing whitespace is removed from each header
1949 before concatenation. Completely empty headers in a concatenation (as
1950 before) are ignored.
1952 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1953 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1955 PH/29 [Removed. This was a change that I later backed out, and forgot to
1956 correct the ChangeLog entry (that I had efficiently created) before
1957 committing the later change.]
1959 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1960 address errors (4xx response to RCPT) over the same connection as other
1961 messages routed to the same hosts. Such deliveries are always "forced",
1962 so retry times are not inspected. This resulted in far too many retries
1963 for the affected addresses. The effect occurred only when there were more
1964 hosts than the hosts_max_try setting in the smtp transport when it had
1965 the 4xx errors. Those hosts that it had tried were not added to the list
1966 of hosts for which the message was waiting, so if all were tried, there
1967 was no problem. Two fixes have been applied:
1969 (i) If there are any address or message errors in an SMTP delivery, none
1970 of the hosts (tried or untried) are now added to the list of hosts
1971 for which the message is waiting, so the message should not be a
1972 candidate for sending over the same connection that was used for a
1973 successful delivery of some other message. This seems entirely
1974 reasonable: after all the message is NOT "waiting for some host".
1975 This is so "obvious" that I'm not sure why it wasn't done
1976 previously. Hope I haven't missed anything, but it can't do any
1977 harm, as the worst effect is to miss an optimization.
1979 (ii) If, despite (i), such a delivery is accidentally attempted, the
1980 routing retry time is respected, so at least it doesn't keep
1981 hammering the server.
1983 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1984 in ${readsocket because some servers need this prod.
1986 PH/32 Added some extra debug output when updating a wait-xxx database.
1988 PH/33 The hint "could be header name not terminated by colon", which has been
1989 given for certain expansion errors for a long time, was not being given
1990 for the ${if def:h_colon_omitted{... case.
1992 PH/34 The spec says: "With one important exception, whenever a domain list is
1993 being scanned, $domain contains the subject domain." There was at least
1994 one case where this was not true.
1996 PH/35 The error "getsockname() failed: connection reset by peer" was being
1997 written to the panic log as well as the main log, but it isn't really
1998 panic-worthy as it just means the connection died rather early on. I have
1999 removed the panic log writing for the ECONNRESET error when getsockname()
2002 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2003 runs only) independently of the message's sender address. This meant
2004 that, if the 4xx error was in fact related to the sender, a different
2005 message to the same recipient with a different sender could confuse
2006 things. In particualar, this can happen when sending to a greylisting
2007 server, but other circumstances could also provoke similar problems.
2008 I have changed the default so that the retry time for these errors is now
2009 based a combination of the sender and recipient addresses. This change
2010 can be overridden by setting address_retry_include_sender=false in the
2013 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2014 remote server are returned as part of bounce messages. This was not
2015 happening for LMTP over a pipe (the lmtp transport), but now it is the
2016 same for both kinds of LMTP.
2018 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2019 in header lines that were in fact CNAMEs. This is no longer the case.
2021 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2022 and queue runs started by the daemon processed all messages. This has
2023 been fixed so that -R and -S can now usefully be given with -q<time>.
2025 PH/40 Import PCRE release 6.7 (fixes some bugs).
2027 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2029 PH/42 Give an error if -q is specified more than once.
2031 PH/43 Renamed the variables $interface_address and $interface_port as
2032 $received_ip_address and $received_port, to make it clear that these
2033 values apply to message reception, and not to the outgoing interface when
2034 a message is delivered. (The old names remain recognized, of course.)
2036 PH/44 There was no timeout on the connect() call when using a Unix domain
2037 socket in the ${readsocket expansion. There now is.
2039 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2040 be meaningful with "accept".
2042 SC/01 Eximstats V1.43
2043 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2045 SC/02 Eximstats V1.44
2046 Use a glob alias rather than an array ref in the generated
2047 parser. This improves both readability and performance.
2049 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2050 Collect SpamAssassin and rejection statistics.
2051 Don't display local sender or destination tables unless
2052 there is data to show.
2053 Added average volumes into the top table text output.
2055 SC/04 Eximstats V1.46
2056 Collect data on the number of addresses (recipients)
2057 as well as the number of messages.
2059 SC/05 Eximstats V1.47
2060 Added 'Message too big' to the list of mail rejection
2061 reasons (thanks to Marco Gaiarin).
2063 SC/06 Eximstats V1.48
2064 Mainlog lines which have GMT offsets and are too short to
2065 have a flag are now skipped.
2067 SC/07 Eximstats V1.49 (Alain Williams)
2068 Added the -emptyok flag.
2070 SC/08 Eximstats V1.50
2071 Fixes for obtaining the IP address from reject messages.
2073 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2074 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2075 whitesspace changes from 4.64-PH/27
2077 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2080 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2081 are found, allow negative numbers in numeric criteria)
2083 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2085 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2086 to match changes made in 4.64-PH/43
2088 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2090 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2091 "domain" is actually an IP address, and force a failure. This locks out
2092 those revolvers/nameservers that support "A-for-A" lookups, in
2093 contravention of the specifications.
2095 PH/48 When a host name was looked up from an IP address, and the subsequent
2096 forward lookup of the name timed out, the host name was left in
2097 $sender_host_name, contrary to the specification.
2099 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2100 restricted to single-key lookups, Exim was not diagnosing an error if
2101 * or *@ was used with a query-style lookup.
2103 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2105 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2106 long ago, but noone interested enough thought of it. Let's just say that
2107 the "1.1" means that there are some new functions that weren't there at
2108 some point in the past.
2110 PH/51 Error processing for expansion failure of helo_data from an smtp
2111 transport during callout processing was broken.
2113 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2114 tested/used via the -bh/-bhc/-bs options.
2116 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2117 bug, fixed in subsequent PCRE releases).
2119 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2120 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2122 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2128 SC/01 Use a glob alias rather than an array ref in eximstats generated
2129 parser. This improves both readability and performance.
2131 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2132 Don't display local sender or destination tables in eximstats unless
2133 there is data to show.
2134 Added average volumes into the eximstats top table text output.
2136 SC/03 Collect data on the number of addresses (recipients) as well
2137 as the number of messages in eximstats.
2139 TF/01 Correct an error in the documentation for the redirect router. Exim
2140 does (usually) call initgroups() when daemonizing.
2142 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2143 with consistent privilege compared to when running as a daemon.
2145 TF/03 Note in the spec that $authenticated_id is not set for local
2146 submissions from trusted users.
2148 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2149 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2151 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2152 by adding some example configuration directives to the default
2153 configuration file. A little bit of work is required to uncomment the
2154 directives and define how usernames and passwords are checked, but
2155 there is now a framework to start from.
2157 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2158 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2159 without this. I don't know how relevant this is to other LDAP libraries.
2161 PH/02 Add the verb name to the "unknown ACL verb" error.
2163 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2165 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2167 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2168 directory not expanded when it should be if an expanded home directory
2169 was set for the address (which is overridden by the transport).
2171 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2174 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2175 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2176 because it is too late at that time, and has no effect.
2178 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2179 security issue with \' (bugzilla #107). I could not use the
2180 PQescapeStringConn() function, because it needs a PGconn value as one of
2183 PH/08 When testing addresses using -bt, indicate those final addresses that
2184 are duplicates that would not cause an additional delivery. At least one
2185 person was confused, thinking that -bt output corresponded to deliveries.
2186 (Suppressing duplicates isn't a good idea as you lose the information
2187 about possibly different redirections that led to the duplicates.)
2189 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2190 systems where poll() doesn't work, in particular OS X.
2192 PH/10 Added more information to debugging output for retry time not reached.
2194 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2195 operations in malware.c.
2197 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2200 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2201 syslog_duplication set false, log lines that would normally be written
2202 both the the main log and to the reject log were not written to syslog at
2205 PH/14 In the default configuration, change the use of "message" in ACL warn
2206 statements to "add_header".
2208 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2209 not followed by a command (e.g. "seen endif").
2211 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2212 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2215 PH/17 Added extra conditions to the default value of delay_warning_condition
2219 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2220 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2221 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2224 The Auto-Submitted: and various List- headers are standardised, whereas I
2225 don't think Precedence: ever was.
2227 PH/18 Refactored debugging code in route_finduser() to show more information,
2228 in particular, the error code if getpwnam() issues one.
2230 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2231 This is apparently needed in addition to the PH/07 change above to avoid
2232 any possible encoding problems.
2234 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2235 but not after initializing Perl.
2237 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2238 output them only if debugging. By default they are written stderr,
2239 apparently, which is not desirable.
2241 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2244 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2247 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2249 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2250 authenticated or an ident call has been made. Suppress the default
2251 values for $authenticated_id and $authenticated_sender (but permit -oMai
2252 and -oMas) when testing with -bh.
2254 PH/24 Re-jigged the order of the tests in the default configuration so that the
2255 tests for valid domains and recipients precede the DNS black list and CSA
2256 tests, on the grounds that those ones are more expensive.
2258 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2259 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2260 This bug exists in every version of Exim that I still have, right back to
2263 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2264 However, an attempt to turn on case-sensitivity in a regex key by
2265 including (?-i) didn't work because the subject string was already
2266 lowercased, and the effects were non-intuitive. It turns out that a
2267 one-line patch can be used to allow (?-i) to work as expected.
2273 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2274 other effects) broke the use of negated acl sub-conditions.
2276 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2279 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2280 "Deny" causes Exim to reject the incoming connection with a 554 error.
2281 Unfortunately, if there is a major crisis, such as a disk failure,
2282 tcp-wrappers gives "deny", whereas what one would like would be some
2283 kind of temporary error. A kludge has been added to help with this.
2284 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2285 554 error is used if errno is still zero or contains ENOENT (which occurs
2286 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2289 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2291 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2292 errors. Otherwise a message that provokes a temporary error (when other
2293 messages do not) can cause a whole host to time out.
2295 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2296 addresses were routed directly to files or pipes from a redirect router.
2297 File deliveries just didn't batch; pipe deliveries might have suffered
2300 PH/06 A failure to get a lock for a hints database would erroneously always say
2301 "Failed to get write lock", even when it was really a read lock.
2303 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2304 of 0600. This has been changed to use the value of the lockfile_mode
2305 option (which defaults to 0600).
2307 PH/08 Applied small patch from the Sieve maintainer.
2309 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2310 folder from quota calculations, a direct delivery into this folder messed
2311 up the contents of the maildirsize file. This was because the regex was
2312 used only to exclude .Trash (or whatever) when the size of the mailbox
2313 was calculated. There was no check that a delivery was happening into an
2314 excluded directory. This bug has been fixed by ignoring all quota
2315 processing for deliveries into excluded directories.
2317 PH/10 Added the maildirfolder_create_regex option to appendfile.
2323 PH/01 The code for finding all the local interface addresses on a FreeBSD
2324 system running IPv6 was broken. This may well have applied to all BSD
2325 systems, as well as to others that have similar system calls. The broken
2326 code found IPv4 interfaces correctly, but gave incorrect values for the
2327 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2328 that it would not match correctly against @[] and not recognize the IPv6
2331 PH/02 The ipliteral router was not recognizing addresses of the form user@
2332 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2334 PH/03 Added disable_ipv6.
2336 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2337 decoded form, because it is most often used to construct To: headers
2338 lines in autoreplies, and the decoded form may well be syntactically
2339 invalid. However, $reply_address has leading white space removed, and all
2340 newlines turned into spaces so that the autoreply transport does not
2343 PH/05 If group was specified without a user on a router, and no group or user
2344 was specified on a transport, the group from the router was ignored.
2346 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2347 for visible compile-time settings that can be used to change these
2348 numbers, for those that want even more. Backwards compatibility with old
2349 spool files has been maintained. However, going back to a previous Exim
2350 release will lost any variables that are in spool files.
2352 PH/07 Two small changes when running in the test harness: increase delay when
2353 passing a TCP/IP connection to a new process, in case the original
2354 process has to generate a bounce, and remove special handling of
2355 127.0.0.2 (sic), which is no longer necessary.
2357 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2358 be the same on different OS.
2360 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2363 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2364 whether --show-vars was specified or not
2366 JJ/02 exipick: Added support for new ACL variable spool format introduced
2369 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2370 syntactically invalid From: or Reply-to: line, and a filter used this to
2371 generate an autoreply, and therefore failed to obtain an address for the
2372 autoreply, Exim could try to deliver to a non-existent relative file
2373 name, causing unrelated and misleading errors. What now happens is that
2374 it logs this as a hard delivery error, but does not attempt to create a
2377 PH/11 The exinext utility has a -C option for testing purposes, but although
2378 the given file was scanned by exinext itself; it wasn't being passed on
2379 when Exim was called.
2381 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2382 an end-of-file indication when reading a command response.
2384 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2385 compiled. In many other places in Exim, IPv6 addresses are always
2386 recognized, so I have changed this. It also means that IPv4 domain
2387 literals of the form [IPV4:n.n.n.n] are now always recognized.
2389 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2390 used if the router is not running as root, for example, when verifying at
2391 ACL time, or when using -bh. The debugging output from this situation was
2392 non-existent - all you got was a failure to exec. I have made two
2395 (a) Failures to set uid/gid, the current directory, or a process leader
2396 in a subprocess such as that created by queryprogram now generate
2397 suitable debugging ouput when -d is set.
2399 (b) The queryprogram router detects when it is not running as root,
2400 outputs suitable debugging information if -d is set, and then runs
2401 the subprocess without attempting to change uid/gid.
2403 PH/15 Minor change to Makefile for building test_host (undocumented testing
2406 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2407 additional section of a DNS packet that returns MX or SRV records.
2408 Instead, it always explicitly searches for A/AAAA records. This avoids
2409 major problems that occur when a DNS server includes only records of one
2410 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2411 fixed another bug: if SRV records were looked up and the corresponding
2412 address records were *not* found in the additional section, the port
2413 values from the SRV records were lost.
2415 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2416 using the correct key (the original address) when searching the retry
2417 rules in order to find which one to use for generating the retry hint.
2419 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2420 adding the default one. Similarly, if it contains a Reply-To: header, the
2421 errors_reply_to option, if set, is not used.
2423 PH/19 When calculating a retry time, Exim used to measure the "time since
2424 failure" by looking at the "first failed" field in the retry record. Now
2425 it does not use this if it is later than than the arrival time of the
2426 message. Instead it uses the arrival time. This makes for better
2427 behaviour in cases where some deliveries succeed, thus re-setting the
2428 "first failed" field. An example is a quota failure for a huge message
2429 when small messages continue to be delivered. Without this change, the
2430 "time since failure" will always be short, possible causing more frequent
2431 delivery attempts for the huge message than are intended.
2432 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2434 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2435 $1, $2, $3) because the numerical variables can be reset during some
2436 expansion items (e.g. "match"), thereby losing the authentication data.
2438 PH/21 Make -bV show the size of off_t variables so that the test suite can
2439 decide whether to run tests for quotas > 2G.
2441 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2442 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2443 filecount value is greater than 2G or if a quota value is greater than 2G
2444 on a system where the size of off_t is not greater than 4, a panic error
2447 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2448 never match. The debug and -bh output now contains an explicit error
2449 message indicating a malformed IPv4 address or mask.
2451 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2452 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2453 PH/23 above applies.
2455 PH/25 Do not write to syslog when running in the test harness. The only
2456 occasion when this arises is a failure to open the main or panic logs
2457 (for which there is an explicit test).
2459 PH/26 Added the /no_tell option to "control=freeze".
2461 PH/27 If a host name lookup failed very early in a connection, for example, if
2462 the IP address matched host_lookup and the reverse lookup yielded a name
2463 that did not have a forward lookup, an error message of the form "no IP
2464 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2465 could be logged. Now it outputs the IP address instead of "NULL".
2467 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2468 allows the sender and the authenticated sender to be set when
2469 submitting a message from within Exim. Since child_open_exim() is
2470 documented for local_scan(), the new function should be too.
2472 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2473 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2474 results in an empty string is now treated as unset.
2476 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2478 PH/31 Added sender_verify_fail logging option.
2480 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2481 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2482 tidied the source and removed it altogether.
2484 PH/33 When a queue run was abandoned because the load average was too high, a
2485 log line was always written; now it is written only if the queue_run log
2486 selector is set. In addition, the log line for abandonment now contains
2487 information about the queue run such as the pid. This is always present
2488 in "start" and "stop" lines but was omitted from the "abandon" line.
2490 PH/34 Omit spaces between a header name and the colon in the error message that
2491 is given when verify = headers_syntax fails (if there are lots of them,
2492 the message gets confusing).
2494 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2495 names, as there are now some PTR records that contain slashes. This check
2496 is only to protect against broken name servers that fall over on strange
2497 characters, so the fact that it applies to all lookups doesn't matter.
2499 PH/36 Now that the new test suite is complete, we can remove some of the
2500 special code in Exim that was needed for the old test suite. For example,
2501 sorting DNS records because real resolvers return them in an arbitrary
2502 order. The new test suite's fake resolver always returns records in the
2505 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2506 bounces) except when queue_only is set, to avoid logging races between
2507 the different processes.
2509 PH/38 Panic-die if .include specifies a non-absolute path.
2511 PH/39 A tweak to the "H" retry rule from its user.
2513 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2514 a label. They prevented compilation on older perls.
2516 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2517 a warning to be raised on newish perls.
2519 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2520 on queue. Changes to match documented behaviour of showing count of
2521 messages matching specified criteria.
2523 PH/40 Changed the default ident timeout from 30s to 5s.
2525 PH/41 Added support for the use of login_cap features, on those BSD systems
2526 that have them, for controlling the resources used by pipe deliveries.
2528 PH/42 The content-scanning code uses fopen() to create files in which to put
2529 message data. Previously it was not paying any attention to the mode of
2530 the files. Exim runs with umask(0) because the rest of the code creates
2531 files with open(), and sets the required mode explicitly. Thus, these
2532 files were ending up world-writeable. This was not a big issue, because,
2533 being within the spool directory, they were not world-accessible. I have
2534 created a function called modefopen, which takes an additional mode
2535 argument. It sets umask(777), creates the file, chmods it to the required
2536 mode, then resets the umask. All the relevant calls to fopen() in the
2537 content scanning code have been changed to use this function.
2539 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2540 to 24 hours. This avoids potential overflow problems when processing G
2541 and H retry rules. I suspect nobody ever tinkers with this value.
2543 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2545 PH/45 When the plaintext authenticator is running as a client, the server's
2546 challenges are checked to ensure they are valid base64 strings. By
2547 default, the authentication attempt is cancelled if an invalid string is
2548 received. Setting client_ignore_invalid_base64 true ignores these errors.
2549 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2550 they are received. Thus, the responses can be made to depend on the
2551 challenges. If an invalid string is ignored, an empty string is placed in
2554 PH/46 Messages that are created by the autoreply transport now contains a
2555 References: header, in accordance with RFCs 2822 and 3834.
2557 PH/47 Added authenticated_sender_force to the smtp transport.
2559 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2561 PH/49 Installed latest patch from the Sieve maintainer.
2563 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2564 was also set, the appendfile transport was unnecessarily scanning a
2565 directory of message files (e.g. for maildir delivery) to find the count
2566 of files (along with the size), even though it did not need this
2567 information. It now does the scan only if it needs to find either the
2568 size of the count of files.
2570 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2572 PH/52 Two bugs concerned with error handling when the smtp transport is
2575 (i) Exim was not creating retry information for temporary errors given
2576 for individual recipients after the DATA command when the smtp transport
2577 was used in LMTP mode. This meant that they could be retried too
2578 frequently, and not timed out correctly.
2580 (ii) Exim was setting the flag that allows error details to be returned
2581 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2582 recipients that were returned after the DATA command.
2584 PH/53 This is related to PH/52, but is more general: for any failing address,
2585 when detailed error information was permitted to be returned to the
2586 sender, but the error was temporary, then after the final timeout, only
2587 "retry timeout exceeded" was returned. Now it returns the full error as
2588 well as "retry timeout exceeded".
2590 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2591 do this, and (what is worse) MTAs that accept it.
2593 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2594 will now be deprecated.
2596 PH/56 New os.c-cygwin from the Cygwin maintainer.
2598 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2599 formats (previously only available in exim formats via -bpr, -bpru,
2600 and -bpra. Now also available in native and exiqgrep formats)
2602 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2603 with very large, slow to parse queues
2605 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2607 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2609 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2610 responses to authentication challenges, though it was showing the
2611 challenges; (ii) I've removed the CR characters from the debug output for
2614 PH/58 Allow for the insertion of a newline as well as a space when a string
2615 is turned into more than one encoded-word during RFC 2047 encoding. The
2616 Sieve code now uses this.
2618 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2619 data_4xx, lost_connection, tls_required.
2621 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2622 message was being sent as an SMTP response.
2624 PH/61 Add -l and -k options to exicyclog.
2626 PH/62 When verifying, if an address was redirected to one new address, so that
2627 verification continued, and the new address failed or deferred after
2628 having set something in $address_data, the value of $address_data was not
2629 passed back to the ACL. This was different to the case when no
2630 redirection occurred. The value is now passed back in both cases.
2632 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2633 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2634 use login_cap.h, so on its own it isn't the distinguishing feature. The
2635 new name refers directly to the setclassresources() function.
2637 PH/65 Added configuration files for NetBSD3.
2639 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2641 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2642 is preferred over IPv4.
2644 PH/68 The bounce_return_message and bounce_return_body options were not being
2645 honoured for bounces generated during the reception of non-SMTP messages.
2646 In particular, this applied to messages rejected by the ACL. This bug has
2647 been fixed. However, if bounce_return_message is true and bounce_return_
2648 body is false, the headers that are returned for a non-SMTP message
2649 include only those that have been read before the error was detected.
2650 (In the case of an ACL rejection, they have all been read.)
2652 PH/69 The HTML version of the specification is now built in a directory called
2653 spec_html instead of spec.html, because the latter looks like a path with
2654 a MIME-type, and this confuses some software.
2656 PH/70 Catch two compiler warnings in sieve.c.
2658 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2659 function verify_get_ident() calls ip_connect() to connect a socket, but
2660 if the "connect()" function timed out, ip_connect() used to close the
2661 socket. However, verify_get_ident() also closes the socket later, and in
2662 between Exim writes to the log, which may get opened at this point. When
2663 the socket was closed in ip_connect(), the log could get the same file
2664 descriptor number as the socket. This naturally causes chaos. The fix is
2665 not to close the socket in ip_connect(); the socket should be closed by
2666 the function that creates it. There was only one place in the code where
2667 this was missing, in the iplookup router, which I don't think anybody now
2668 uses, but I've fixed it anyway.
2670 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2671 well as to direct DNS lookups. Otherwise the handling of names in host
2672 lists is inconsistent and therefore confusing.
2678 PH/01 Two changes to the default runtime configuration:
2680 (1) Move the checks for relay_from_hosts and authenticated clients from
2681 after to before the (commented out) DNS black list checks.
2683 (2) Add control=submission to the relay_from_hosts and authenticated
2684 clients checks, on the grounds that messages accepted by these
2685 statements are most likely to be submissions.
2687 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2689 (1) Generate an error if the third argument for the ${prvs expansion is
2692 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2695 (3) Reset the variables that are obtained from the first argument of
2696 ${prvscheck and used in the second argument before leaving the code,
2697 because their memory is reclaimed, so using them afterwards may do
2700 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2701 one (it's much easier than Tom thought :-).
2703 (5) Because of (4), we can now allow for the use of $prvscheck_result
2704 inside the third argument.
2706 PH/03 For some reason, the default setting of PATH when running a command from
2707 a pipe transport was just "/usr/bin". I have changed it to
2710 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2711 anything to be listed in the output from -bV.
2713 PH/05 When a filter generated an autoreply, the entire To: header line was
2714 quoted in the delivery log line, like this:
2716 => >A.N.Other <ano@some.domain> <original@ddress> ...
2718 This has been changed so that it extracts the operative address. There
2719 may be more than one such address. If so, they are comma-separated, like
2722 => >ano@some.domain,ona@other.domain <original@ddress> ...
2724 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2725 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2726 not being looked up in the rDNS to get a host name, Exim was showing the
2727 IP address twice in Received: lines, even though the IP addresses were
2728 identical. For example:
2730 Received: from [1.2.3.4] (helo=[1.2.3.4])
2732 However, if the real host name was known, it was omitting the HELO data
2733 if it matched the actual IP address. This has been tidied up so that it
2734 doesn't show the same IP address twice.
2736 PH/07 When both +timestamp and +memory debugging was on, the value given by
2737 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2738 called by the debug printing, thereby overwriting the timestamp buffer.
2739 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2741 PH/08 When the original message was included in an autoreply transport, it
2742 always said "this is a copy of the message, including all the headers",
2743 even if body_only or headers_only was set. It now gives an appropriate
2746 PH/09 Applied a patch from the Sieve maintainer which:
2748 o fixes some comments
2749 o adds the (disabled) notify extension core
2750 o adds some debug output for the result of if/elsif tests
2751 o points to the current vacation draft in the documentation
2752 and documents the missing references header update
2756 o fixes a bug in processing the envelope test (when testing
2757 multiple envelope elements, the last element determinted the
2760 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2761 Electronic Mail") by including:
2763 Auto-submitted: auto-generated
2765 in the messages that it generates (bounce messages and others, such as
2766 warnings). In the case of bounce messages for non-SMTP mesages, there was
2767 also a typo: it was using "Auto_submitted" (underscore instead of
2768 hyphen). Since every message generated by Exim is necessarily in response
2769 to another message, thes have all been changed to:
2771 Auto-Submitted: auto-replied
2773 in accordance with these statements in the RFC:
2775 The auto-replied keyword:
2777 - SHOULD be used on messages sent in direct response to another
2778 message by an automatic process,
2780 - MUST NOT be used on manually-generated messages,
2782 - MAY be used on Delivery Status Notifications (DSNs) and Message
2783 Disposition Notifications (MDNs),
2785 - MUST NOT be used on messages generated by automatic or periodic
2786 processes, except for messages which are automatic responses to
2789 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2790 to the default Received: header definition.
2792 PH/12 Added log selector acl_warn_skipped (default on).
2794 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2795 variables because (a) they are in the wrong storage pool and (b) even if
2796 they were copied, it wouldn't work properly because of the caching.
2798 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2799 checking when decoding. Apparently there are clients that generate
2800 overlong encoded strings. Why am I not surprised?
2802 PH/15 If the first argument of "${if match_address" was not empty, but did not
2803 contain an "@" character, Exim crashed. Now it writes a panic log message
2804 and treats the condition as false.
2806 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2808 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2809 extension "envelope-auth". The code is finished and in agreement with
2810 other implementations, but there is no documentation so far and in fact,
2811 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2812 not changing the active code.
2814 Print executed "if" and "elsif" statements when debugging is used. This
2815 helps a great deal to understand what a filter does.
2817 Document more things not specified clearly in RFC3028. I had all this
2818 sorted out, when out of a sudden new issues came to my mind. Oops."
2820 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2823 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2824 canonical form (as documented). However, after a host name lookup from
2825 the IP address, check_host() was doing a simple string comparison with
2826 addresses acquired from the DNS when checking that the found name did
2827 have the original IP as one of its addresses. Since any found IPv6
2828 addresses are likely to be in abbreviated form, the comparison could
2829 fail. Luckily, there already exists a function for doing the comparison
2830 by converting both addresses to binary, so now that is used instead of
2831 the text comparison.
2833 PH/20 There was another similar case to PH/19, when a complete host name was
2834 given in a host list; looking up its IP address could give an abbreviated
2835 form, whereas the current host's name might or might not be abbreviated.
2836 The same fix has been applied.
2842 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2843 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2846 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2847 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2849 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2851 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2852 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2853 Also, for "Auto-Submitted", treat anything other than "no" as
2854 non-personal, in accordance with RFC 3834. (Previously it treated
2855 anything starting "auto-" as non-personal.)
2857 TF/01 The control=submission/name=... option had a problem with syntax
2858 errors if the name included a slash character. The /name= option
2859 now slurps the rest of the string, so it can include any characters
2860 but it must come last in the list of options (after /sender_retain
2863 PH/05 Some modifications to the interface to the fake nameserver for the new
2871 TK/01 Added the "success_on_redirect" address verification option. See
2872 NewStuff for rationale and an example.
2874 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2876 PH/02 Patch to exigrep to allow it to work on syslog lines.
2878 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2879 fread() to skip over the body file's header line, because in Cygwin the
2880 header line is locked and is inaccessible.
2882 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2883 co-exist for some time) to make it clear that it is the Exim ID that is
2884 referenced, not the Message-ID: header line.
2886 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2887 string_format() function, because snprintf() does not exist on all
2890 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2891 ${stat: expansion item.
2893 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2894 protocol synchronization error", to keep the pedants happy.
2896 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2897 well as for IRIX systems, when gcc is being used. See the host.c source
2900 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2902 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2905 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2906 transport and to the smtp transport in LMTP mode.
2908 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2910 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2911 run a filter in a subprocess. This could lead to confusion in subsequent
2912 lookups in the parent process. There should also be a search_tidyup() at
2913 the end of the subprocess.
2915 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2916 only if the host matched helo_try_verify_hosts, which caused the
2917 verification to occur when the EHLO/HELO command was issued. The ACL just
2918 tested the remembered result. Now, if a previous verification attempt has
2919 not happened, "verify = helo" does it there and then.
2921 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2923 TK/03 Fix log output including CR from clamd.
2925 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2926 exist provoked a memory error which could cause a segfault.
2928 PH/15 Installed PCRE 6.2
2930 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2932 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2933 of the problem. Specifically, suggested +O2 rather than +O1 for the
2936 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2938 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2939 record, the port was not correctly logged when the outgoing_port log
2940 selector was set (it logged the transort's default port).
2942 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2943 fallback_hosts, and "hosts" in the smtp transport.
2945 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2946 host errors such as "Connection refused".
2948 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2949 authentication with radiusclient 0.4.9:
2951 - Error returned from rc_read_config was caught wrongly
2952 - Username/password not passed on to radius server due to wrong length.
2954 The presumption is that some radiusclient API changes for 4.51/PH/17
2955 were not taken care of correctly. The code is still untested by me (my
2956 Linux distribution still has 0.3.2 of radiusclient), but it was
2957 contributed by a Radius user.
2959 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2960 expanding the "port" option of the smtp transport.
2962 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2963 while reading a MIME header. Thanks to Tom Hughes for a patch.
2965 PH/24 Include config.h inside local_scan.h so that configuration settings are
2968 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2969 that in an ACL for RCPT (for example), you can examine exactly what was
2972 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2973 commands, but it was not correctly comparing the address with the actual
2974 client host address. Thus, it would show the EHLO address in Received:
2975 header lines when this was not necessary.
2977 PH/27 Added the % operator to ${eval:}.
2979 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2980 it should be ignoring failures (because with -C, for example, it has lost
2981 privilege). It wasn't ignoring creation failures other than "already
2984 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2987 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2988 because an input line was too long, either on its own, or by virtue of
2989 too many continuations, the temporary file was not being removed, and the
2990 return code was incorrect.
2992 PH/31 Missing "BOOL" in function definition in filtertest.c.
2994 PH/32 Applied Sieve patches from the maintainer.
2996 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2998 PH/33 Added "verify = not_blind".
3000 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3001 Local/Makefile (with some defaults set). These are used in built scripts
3002 such as exicyclog, but they have never been used in the exim_install
3003 script (though there are many overriding facilities there). I have
3004 arranged that the exim_install script now takes note of these two
3007 PH/35 Installed configuration files for Dragonfly.
3009 PH/36 When a locally submitted message by a trusted user did not contain a
3010 From: header, and the sender address was obtained from -f or from an SMTP
3011 MAIL command, and the trusted user did not use -F to supply a sender
3012 name, $originator_name was incorrectly used when constructing a From:
3013 header. Furthermore, $originator_name was used for submission mode
3014 messages from external hosts without From: headers in a similar way,
3015 which is clearly wrong.
3017 PH/37 Added control=suppress_local_fixups.
3019 PH/38 When log_selector = +received_sender was set, and the addition of the
3020 sender made the log line's construction buffer exactly full, or one byte
3021 less than full, an overflow happened when the terminating "\n" was
3024 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3025 when the result of a list match is failure because a DNS lookup failed.
3027 PH/40 RM_COMMAND is now used in the building process.
3029 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3030 the "build-* directories that it finds.
3032 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3033 address in a domain literal was a prefix of an interface address.
3035 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3036 when verifying a sender address, unless rewrite_headers is false.
3038 PH/44 Wrote a long comment about why errors_to addresses are verified as
3039 recipients, not senders.
3041 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3042 the ratelimit ACL was added.
3044 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3046 PH/46 Added extra information about PostgreSQL errors to the error string.
3048 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3049 avoiding the need to install special zones in a real server. This is
3050 backwards compatible; if it can't find the fake resolver, it drops back.
3051 Thus, both old and new test suites can be run.
3053 TF/02 Added util/ratelimit.pl
3055 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3056 clock is set back in time.
3058 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3059 Candler <B.Candler@pobox.com>.
3061 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3062 OK for addresses that are the result of redirections.
3064 PH/48 A number of further additions for the benefit of the new test suite,
3065 including a fake gethostbyname() that interfaces to the fake DNS resolver
3068 TF/06 The fix for widen_domains has also been applied to qualify_single and
3069 search_parents which are the other dnslookup options that can cause
3072 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3075 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3077 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3078 service" if it encountered an MX record with an empty target hostname.
3079 The message is now "an MX or SRV record indicated no SMTP service".
3081 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3082 if the DNS of the sending site is misconfigured. This is quite a
3083 common situation. This change restores the behaviour of treating a
3084 helo verification defer as a failure.
3086 PH/51 If self=fail was set on a router, the bounce message did not include the
3087 actual error message.
3093 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3095 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3096 command itself ended in error, the underlying message about the transport
3097 filter timeout was being overwritten with the pipe command error. Now the
3098 underlying error message should be appended to the second error message.
3100 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3102 PH/02 Reduce the amount of output that "make" produces by default. Full output
3103 can still be requested.
3105 PH/03 The warning log line about a condition test deferring for a "warn" verb
3106 was being output only once per connection, rather than after each
3107 occurrence (because it was using the same function as for successful
3108 "warn" verbs). This seems wrong, so I have changed it.
3110 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3111 it should not have, which might have caused a crash in the right
3112 circumstances, but probably never did.
3114 PH/04 Installed a modified version of Tony Finch's patch to make submission
3115 mode fix the return path as well as the Sender: header line, and to
3116 add a /name= option so that you can make the user's friendly name appear
3119 TF/03 Added the control = fakedefer ACL modifier.
3121 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3122 Mark Lowes for thorough testing.
3124 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3126 TK/03 Merged latest SRS patch from Miles Wilton.
3128 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3129 with the definition in sysexits.h (which is #included earlier).
3130 Fortunately, Exim does not actually use EX_OK. The code used to try to
3131 preserve the sysexits.h value, by assumimg that macro definitions were
3132 scanned for macro replacements. I have been disabused of this notion,
3133 so now the code just undefines EX_OK before #including unistd.h.
3135 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3136 in the smtp transport. When a block could not be written in a single
3137 write() function, the timeout was being re-applied to each part-write.
3138 This seems wrong - if the receiver was accepting one byte at a time it
3139 would take for ever. The timeout is now adjusted when this happens. It
3140 doesn't have to be particularly precise.
3142 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3143 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3145 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3146 without a domain if the check to <postmaster@domain> fails.
3148 SC/01 Eximstats: added -xls and the ability to specify output files
3149 (patch written by Frank Heydlauf).
3151 SC/02 Eximstats: use FileHandles for outputing results.
3153 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3155 SC/04 Eximstats: fixed display of large numbers with -nvr option
3157 SC/05 Eximstats: fixed merging of reports with empty tables.
3159 SC/06 Eximstats: added the -include_original_destination flag
3161 SC/07 Eximstats: removed tabs and trailing whitespace.
3163 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3165 TK/06 MBOX spool code: Add real "From " MBOX separator line
3166 so the .eml file is really in mbox format (even though
3167 most programs do not really care). Patch from Alex Miller.
3169 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3170 The latter is generated from $received_to and is only set if the
3171 message has one envelope recipient. SA can use these headers,
3172 obviously out-of-the-box. Patch from Alex Miller.
3174 PH/08 The ${def test on a variable was returning false if the variable's
3175 value was "0", contrary to what the specification has always said!
3176 The result should be true unless the variable is empty.
3178 PH/09 The syntax error of a character other than { following "${if
3179 def:variable_name" (after optional whitespace) was not being diagnosed.
3180 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3181 accidental colon was present, for example, could give incorrect results.
3183 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3184 result is used (not including appendfile, where other changes are about
3187 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3188 This involved changing a lot of size variables from int to off_t. It
3189 should work with maildirs and everything.
3191 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3192 spamd dying while we are connected to it.
3194 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3197 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3198 function for BDB 4.3.
3200 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3202 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3203 the sources for PCRE have been re-arranged and more files are now
3206 PH/15 The code I had for printing potentially long long variables in PH/11
3207 above was not the best (it lost precision). The length of off_t variables
3208 is now inspected at build time, and an appropriate printing format (%ld
3209 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3210 to be "long long int" or "long int". This is needed for the internal
3211 formatting function string_vformat().
3213 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3214 the configuration file to be ":syslog", then the script "guesses" where
3215 the logs files are, rather than using the compiled in default. In our
3216 case the guess is not the same as the compiled default, so the script
3217 suddenly stopped working when I started to use syslog. The patch checks
3218 to see if log_file_path is "". If so, it attempts to read it from exim
3219 with no configuration file to get the compiled in version, before it
3220 falls back to the previous guessing code."
3222 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3223 implementing BATV in an Exim configuration. See NewStuff for the gory
3226 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3227 Makefile that are specific to HP-UX.
3229 PH/18 If the "use_postmaster" option was set for a recipient callout together
3230 with the "random" option, the postmaster address was used as the MAIL
3231 FROM address for the random test, but not for the subsequent recipient
3232 test. It is now used for both.
3234 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3235 patch removes a few documentation additions to RFC 3028, because the
3236 latest draft now contains them. It adds the new en;ascii-case comparator
3237 and a new error check for 8bit text in MIME parts. Comparator and
3238 require names are now matched exactly. I enabled the subaddress
3239 extension, but it is not well tested yet (read: it works for me)."
3241 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3242 rework some of the code of TK/09 above to avoid the hardwired use of
3243 "%lld" and "long long". Replaced the call to snprintf() with a call to
3246 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3247 records point to non-existent hosts", "retry timeout exceeded", and
3248 "retry time not reached for any host after a long failure period".
3250 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3251 experimental DomainKeys support:
3253 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3254 (2) On an error such as an illegally used "control", the wrong name for
3255 the control was given.
3257 These problems did NOT occur unless DomainKeys support was compiled.
3259 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3261 PH/24 Added ${if match_ip condition.
3263 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3264 databases so that it will be absolutely obvious if a crash occurs in the
3265 DB library. This is a regular occurrence (often caused by mis-matched
3268 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3269 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3270 picked up on a user's system that detects such things. There doesn't seem
3271 to be a gcc warning option for this - only an attribute that has to be
3272 put on the function's prototype. It seems that in Fedora Core 4 they have
3273 set this on a number of new functions. No doubt there will be more in due
3276 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3277 specify a transport. However, if an address that was verified by such a
3278 router was the subject of a callout, Exim crashed because it tried to
3279 read the rcpt_include_affixes from the non-existent transport. Now it
3280 just assumes that the setting of that option is false. This bug was
3281 introduced by 4.51/PH/31.
3283 PH/28 Changed -d+all to exclude +memory, because that information is very
3284 rarely of interest, but it makes the output a lot bigger. People tend to
3285 do -d+all out of habit.
3287 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3288 code in os-type was giving problems when libc.so lives in lib64, like on
3291 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3292 aren't the modern standard, and it seems that some systems' include files
3293 don't always have them. Exim was already checking for some of the newer
3294 ones like T_AAAA, and defining it itself. I've added checks for all the
3295 record types that Exim uses.
3297 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3298 not automatically generating a new one, as it is supposed to. This
3299 prevented TLS from working. If the file did exist, but contained invalid
3300 data, a new version was generated, as expected. It was only the case of a
3301 non-existent file that was broken.
3303 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3304 with a change in libdomainkeys > 0.64.
3306 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3307 from DNS. If the selector record carries the flag, it now has
3308 precedence over the domain-wide flag.
3310 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3312 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3313 the use of a port name that isn't defined in /etc/services) occurred, the
3314 message was deferred as in a normal delivery, and thus remained on the
3315 spool, instead of being failed because of the mua_wrapper setting. This
3316 is now fixed, and I tidied up some of the mua_wrapper messages at the
3319 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3320 the messages in a hash of arrays rather than using individual hashes.
3321 This is a bit cleaner and results in dramatic memory savings, albeit
3322 at a slight CPU cost.
3324 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3325 as requested by Marc Sherman.
3327 SC/10 Eximstats: added histograms for user specified patterns as requested
3330 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3332 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3333 fopen() in the content-scanning modules that did not already have it.
3339 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3340 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3342 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3344 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3346 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3347 to test Sieve filters that use "vacation".
3349 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3350 that changes the way the GnuTLS parameters are stored in the cache file.
3351 The new format can be generated externally. For backward compatibility,
3352 if the data in the cache doesn't make sense, Exim assumes it has read an
3353 old-format file, and it generates new data and writes a new file. This
3354 means that you can't go back to an older release without removing the
3357 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3358 work if there are any delivery delays because "one_time" forces the
3359 parent to be marked "delivered", so its unseen clone is never tried
3360 again. For this reason, Exim now forbids the simultaneous setting of
3363 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3364 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3365 entry if you want to know the details. The fix, however, neglected to
3366 consider the case when local delivery batching is involved. The test for
3367 "previously delivered" was not happening when checking to see if an
3368 address could be batched with a previous (undelivered) one; under
3369 certain circumstances this could lead to multiple deliveries to the same
3372 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3373 in its include files, and this causes problems building Exim.
3375 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3376 header_syntax) but Exim was just ignoring anything given after a slash.
3377 In particular, this caused confusion with an attempt to use "verify =
3378 reverse_host_lookup/defer_ok". An error is now given when options are
3379 supplied for verify items that do not have them. (Maybe reverse_host_
3380 lookup should have a defer_ok option, but that's a different point.)
3382 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3383 defined by RFC 821) to 2048, because there were problems with some AUTH
3384 commands, and RFC 1869 says the size should be increased for extended
3385 SMTP commands that take arguments.
3387 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3390 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3391 "unknown" error; now it says that the functionality isn't in the binary.
3393 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3394 an address' error message when a string expansion fails (syntax or
3395 whatever). Otherwise the password may appear in the log. Following change
3396 PH/42 below, there is no longer a chance of it appearing in a bounce
3399 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3401 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3402 of its addresses was ever tried. (Bugzilla bug #2.)
3404 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3405 the result incorrectly in the debug output. (It correctly added a newline
3406 to what was transported.)
3408 TF/01 Added $received_time.
3410 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3411 commented out examples of how to interface to a virus scanner and to
3412 SpamAssassin. Also added commented examples of av_scanner and
3413 spamd_address settings.
3415 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3416 and controls are allowed in which ACLs. There were a couple of minor
3417 errors. Some of the entries in the conditions table (which is a table of
3418 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3419 negation of where the condition IS allowed.
3421 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3423 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3424 header file does not have a version number, so I've had to invent a new
3425 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3426 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3427 radiusclient), but it was contributed by a Radius user.
3429 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3430 files or directories, for OpenSSL.
3432 PH/19 When an Exim process that is running as root has to create an Exim log
3433 file, it does so in a subprocess that runs as exim:exim so as to get the
3434 ownership right at creation (otherwise, other Exim processes might see
3435 the file with the wrong ownership). There was no test for failure of this
3436 fork() call, which would lead to the process getting stuck as it waited
3437 for a non-existent subprocess. Forks do occasionally fail when resources
3438 run out. I reviewed all the other calls to fork(); they all seem to check
3441 PH/20 When checking for unexpected SMTP input at connect time (before writing
3442 the banner), Exim was not dealing correctly with a non-positive return
3443 from the read() function. If the client had disconnected by this time,
3444 the result was a log entry for a synchronization error with an empty
3445 string after "input=" when read() returned zero. If read() returned -1
3446 (an event I could not check), uninitialized data bytes were printed.
3447 There were reports of junk text (parts of files, etc) appearing after
3450 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3452 PH/22 Added support for macro redefinition, and (re)definition in between
3453 driver and ACL definitions.
3455 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3456 forgetting to use the resulting value; it was using the unexpanded value.
3458 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3459 hadn't been configured. The fix is from Juergen Kreileder, who
3460 understands it better than I do:
3462 "Here's what I see happening with three configured cyrus_sasl
3463 authenticators configured (plain, login, cram-md5):
3465 On startup auth_cyrus_sasl_init() gets called for each of these.
3466 This means three calls to sasl_listmech() without a specified mech_list.
3467 => SASL tests which mechs of all available mechs actually work
3468 => three warnings about OTP not working
3469 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3471 With the patch, sasl_listmech() also gets called three times. But now
3472 SASL's mech_list option is set to the server_mech specified in the the
3473 authenticator. Or in other words, the answer from sasl_listmech()
3474 gets limited to just the mech you're testing for (which is different
3476 => the return list contains just 'plain' or 'login', 'cram-md5' or
3477 nothing depending on the value of ob->server_mech.
3479 I've just tested the patch: Authentication still works fine,
3480 unavailable mechs specified in the exim configuration are still
3481 caught, and the auth.log warnings about OTP are gone."
3483 PH/25 When debugging is enabled, the contents of the command line are added
3484 to the debugging output, even when log_selector=+arguments is not
3487 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3488 answer is "GNU", and only if the return is "GNU/something" is the answer
3491 PH/27 $acl_verify_message is now set immediately after the failure of a
3492 verification in an ACL, and so is available in subsequent modifiers. In
3493 particular, the message can be preserved by coding like this:
3495 warn !verify = sender
3496 set acl_m0 = $acl_verify_message
3498 Previously, $acl_verify_message was set only while expanding "message"
3499 and "log_message" when a very denied access.
3501 PH/28 Modified OS/os.c-Linux with
3503 -#ifndef OS_LOAD_AVERAGE
3504 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3506 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3507 nomenclature these days.)
3509 PH/29 Installed patch from the Sieve maintainer that adds the options
3510 sieve_useraddress and sieve_subaddress to the redirect router.
3512 PH/30 In these circumstances:
3513 . Two addresses routed to the same list of hosts;
3514 . First host does not offer TLS;
3515 . First host accepts first address;
3516 . First host gives temporary error to second address;
3517 . Second host offers TLS and a TLS session is established;
3518 . Second host accepts second address.
3519 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3520 and peerdn, if requested) that were in fact used only for the second
3523 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3524 attention to any local part prefix or suffix that was matched by the
3525 router that accepted the address. It now behaves in the same way as it
3526 does for delivery: the affixes are removed from the local part unless
3527 rcpt_include_affixes is set on the transport.
3529 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3530 timeout during the DATA phase of an incoming message.
3532 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3533 applied a patch sanctioned by the Sieve maintainer.
3535 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3536 the uid or gid is negative. A case of a negative gid caused this to be
3537 noticed. The fix allows for either to be negative.
3539 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3540 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3543 PH/36 The change PH/12 above was broken. Fixed it.
3545 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3546 the grounds that routing the same address twice would always produce the
3547 same answer. This might have been true once, but it is certainly no
3548 longer true now. Routing a child address may depend on the previous
3549 routing that produced that child. Some complicated redirection strategies
3550 went wrong when messages had multiple recipients, and made Exim's
3551 behaviour dependent on the order in which the addresses were given.
3553 I have moved the duplicate checking until after the routing is complete.
3554 Exim scans the addresses that are assigned to local and remote
3555 transports, and removes any duplicates. This means that more work will be
3556 done, as duplicates will always all be routed, but duplicates are
3557 presumably rare, so I don't expect this is of any significance.
3559 For deliveries to pipes, files, and autoreplies, the duplicate checking
3560 still happens during the routing process, since they are not going to be
3563 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3564 It corrects a timeout issue with spamd. This is Ian's comment: "The
3565 background is that sometimes spamd either never reads data from a
3566 connection it has accepted, or it never writes response data. The exiscan
3567 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3568 blindly assumes that writes won't block so it may never time out."
3570 PH/39 Allow G after quota size as well as K and M.
3572 PH/40 The value set for $authenticated_id in an authenticator may not contain
3573 binary zeroes or newlines because the value is written to log lines and
3574 to spool files. There was no check on this. Now the value is run through
3575 the string_printing() function so that such characters are converted to
3576 printable escape sequences.
3578 PH/41 $message_linecount is a new variable that contains the total number of
3579 lines in the message. Compare $body_linecount, which is the count for the
3582 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3583 bounce and delay warning messages, except in certain special cases, which
3586 (a) An SMTP error message from a remote host;
3587 (b) A message specified in a :fail: redirection;
3588 (c) A message specified in a "fail" command in a system filter;
3589 (d) A message specified in a FAIL return from the queryprogram router;
3590 (e) A message specified by the cannot_route_message router option.
3592 In these cases only, Exim does include the error details in bounce and
3593 warning messages. There are also a few cases where bland messages such
3594 as "unrouteable address" or "local delivery error" are given.
3596 PH/43 $value is now also set for the "else" part of a ${run expansion.
3598 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3599 being worked on, but at least Exim now implements the latest version to
3602 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3603 process to complete was treated as a delivery failure, a timeout while
3604 writing the message to the pipe was logged, but erroneously treated as a
3605 successful delivery. Such timeouts include transport filter timeouts. For
3606 consistency with the overall process timeout, these timeouts are now
3607 treated as errors, giving rise to delivery failures by default. However,
3608 there is now a new Boolean option for the pipe transport called
3609 timeout_defer, which, if set TRUE, converts the failures into defers for
3610 both kinds of timeout. A transport filter timeout is now identified in
3613 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3614 systems where "make" and "gmake" are different, calling "gmake" at top
3615 level broke things. I've arranged for the value of $(MAKE) to be passed
3616 from the Makefile to this script so that it can call the same version of
3620 A note about Exim versions 4.44 and 4.50
3621 ----------------------------------------
3623 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3624 changes of various kinds. As a consequence, a big documentation update was
3625 needed. This delayed the release for rather longer than seemed good, especially
3626 in the light of a couple of (minor) security issues. Therefore, the changes
3627 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3628 release. So 4.44 and 4.50 are in effect two different branches that both start
3631 I have left the 4.50 change log unchanged; it contains all the changes since
3632 4.43. The change log for 4.44 is below; many of its items are identical to
3633 those for 4.50. This seems to be the most sensible way to preserve the
3634 historical information.
3640 1. Minor wording change to the doc/README.SIEVE file.
3642 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3643 computation of the current number of files was incorrect.
3645 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3646 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3649 4. Give more explanation in the error message when the command for a transport
3650 filter fails to execute.
3652 5. There are several places where Exim runs a non-Exim command in a
3653 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3654 was being done only for the command run by the queryprogram router. It is
3655 now done for all such subprocesses. The other cases are: ${run, transport
3656 filters, and the commands run by the lmtp and pipe transports.
3658 6. Added CONFIGURE_GROUP build-time option.
3660 7. Some older OS have a limit of 256 on the maximum number of file
3661 descriptors. Exim was using setrlimit() to set 1000 as a large value
3662 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3663 systems. I've change it so that if it can't get 1000, it tries for 256.
3665 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3666 was an oversight, and furthermore, ever since the addition of extra
3667 controls (e.g. 4.43/32), the checks on when to allow different forms of
3668 "control" were broken. There should now be diagnostics for all cases when a
3669 control that does not make sense is encountered.
3671 9. Added the /retain_sender option to "control=submission".
3673 10. $recipients is now available in the predata ACL (oversight).
3675 11. Tidy the search cache before the fork to do a delivery from a message
3676 received from the command line. Otherwise the child will trigger a lookup
3677 failure and thereby defer the delivery if it tries to use (for example) a
3678 cached ldap connection that the parent has called unbind on.
3680 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3681 of $address_data from the recipient verification was clobbered by the
3682 sender verification.
3684 13. The value of address_data from a sender verification is now available in
3685 $sender_address_data in subsequent conditions in the ACL statement.
3687 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3689 15. Added a new option "connect=<time>" to callout options, to set a different
3692 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3693 was its contents. (It was OK if the option was not defined at all.)
3695 17. A "Completed" log line is now written for messages that are removed from
3696 the spool by the -Mrm option.
3698 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3699 information about exactly what failed.
3701 19. Added -dd to debug only the daemon process.
3703 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3704 handles timeouts, both on the server side and network timeouts. Renamed the
3705 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3707 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3708 It is now set to "smtps".
3710 22. $host_address is now set to the target address during the checking of
3711 ignore_target_hosts.
3713 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3714 being passed; this would have caused $sender_host_name to have been used if
3715 matching the list had actually called for a host name (not very likely,
3716 since this list is usually IP addresses). A host name is now passed as
3719 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3720 code that specifies a non-restarting handler (typically sigaction() in
3721 modern systems) in an attempt to fix a rare and obscure crash bug.
3723 25. Narrowed the window for a race in the daemon that could cause it to ignore
3724 SIGCHLD signals. This is not a major problem, because they are used only to
3725 wake it up if nothing else does.
3727 26. A malformed maildirsize file could cause Exim to calculate negative values
3728 for the mailbox size or file count. Odd effects could occur as a result.
3729 The maildirsize information is now recalculated if the size or filecount
3732 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3733 support for a long time. Removed HAVE_SYS_VFS_H.
3735 28. Installed the latest version of exipick from John Jetmore.
3737 29. In an address list, if the pattern was not a regular expression, an empty
3738 subject address (from a bounce message) matched only if the pattern was an
3739 empty string. Non-empty patterns were not even tested. This was the wrong
3740 because it is perfectly reasonable to use an empty address as part of a
3741 database query. An empty address is now tested by patterns that are
3742 lookups. However, all the other forms of pattern expect the subject to
3743 contain a local part and a domain, and therefore, for them, an empty
3744 address still always fails if the pattern is not itself empty.
3746 30. Exim went into a mad DNS loop when attempting to do a callout where the
3747 host was specified on an smtp transport, and looking it up yielded more
3748 than one IP address.
3750 31. Re-factored the code for checking spool and log partition space into a
3751 function that finds that data and another that does the check. The former
3752 is then used to implement four new variables: $spool_space, $log_space,
3753 $spool_inodes, and $log_inodes.
3755 32. The RFC2047 encoding function was originally intended for short strings
3756 such as real names; it was not keeping to the 75-character limit for
3757 encoded words that the RFC imposes. It now respects the limit, and
3758 generates multiple encoded words if necessary. To be on the safe side, I
3759 have increased the buffer size for the ${rfc2047: expansion operator from
3762 33. It is now permitted to omit both strings after an "if" condition; if the
3763 condition is true, the result is "true". As before, when the second string
3764 is omitted, a false condition yields an empty string. This makes it less
3765 cumbersome to write custom ACL and router conditions.
3767 34. Failure to deliver a bounce message always caused it to be frozen, even if
3768 there was an errors_to setting on the router. The errors_to setting is now
3771 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3772 canonical form (fully expanded) before being placed in
3773 $sender_host_address.
3775 36. The table in the code that translates DNS record types into text (T_A to
3776 "A" for instance) was missing entries for NS and CNAME. It is just possible
3777 that this could have caused confusion if both these types were looked up
3778 for the same domain, because the text type is used as part of Exim's
3779 per-process caching. But the chance of anyone hitting this buglet seems
3782 37. The dnsdb lookup has been extended in a number of ways.
3784 (1) There is a new type, "zns", which walks up the domain tree until it
3785 finds some nameserver records. It should be used with care.
3787 (2) There is a new type, "mxh", which is like "mx" except that it returns
3788 just the host names, not the priorities.
3790 (3) It is now possible to give a list of domains (or IP addresses) to be
3791 looked up. The behaviour when one of the lookups defers can be
3792 controlled by a keyword.
3794 (4) It is now possible to specify the separator character for use when
3795 multiple records are returned.
3797 38. The dnslists ACL condition has been extended: it is now possible to supply
3798 a list of IP addresses and/or domains to be looked up in a particular DNS
3801 39. Added log_selector=+queue_time_overall.
3803 40. When running the queue in the test harness, wait just a tad after forking a
3804 delivery process, to get repeatability of debugging output.
3806 41. Include certificate and key file names in error message when GnuTLS fails
3807 to set them up, because the GnuTLS error message doesn't include the name
3808 of the failing file when there is a problem reading it.
3810 42. Allow both -bf and -bF in the same test run.
3812 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3814 44. The "Exiscan patch" is now merged into the mainline Exim source.
3816 45. Sometimes the final signoff response after QUIT could fail to get
3817 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3818 < 0 before doing a fflush(). This bug looks as though it goes back to the
3819 introduction of TLS in release 3.20, but "sometimes" must have been rare
3820 because the tests only now provoked it.
3822 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3823 (this can affect the format of dates).
3825 47. exim_tidydb, when checking for the continued existence of a message for
3826 which it has found a message-specific retry record, was not finding
3827 messages that were in split spool directories. Consequently, it was
3828 deleting retry records that should have stayed in existence.
3830 48. Steve fixed some bugs in eximstats.
3832 49. The SPA authentication driver was not abandoning authentication and moving
3833 on to the next authenticator when an expansion was forced to fail,
3834 contradicting the general specification for all authenticators. Instead it
3835 was generating a temporary error. It now behaves as specified.
3837 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3838 (the order specifies the preference for clients). The order is now AES256,
3839 AES128, 3DES, ARCFOUR128.
3841 51. Small patch to Sieve code - explicitly set From: when generating an
3844 52. Exim crashed if a remote delivery caused a very long error message to be
3845 recorded - for instance if somebody sent an entire SpamAssassin report back
3846 as a large number of 550 error lines. This bug was coincidentally fixed by
3847 increasing the size of one of Exim's internal buffers (big_buffer) that
3848 happened as part of the Exiscan merge. However, to be on the safe side, I
3849 have made the code more robust (and fixed the comments that describe what
3852 53. Now that there can be additional text after "Completed" in log lines (if
3853 the queue_time_overall log selector is set), a one-byte patch to exigrep
3854 was needed to allow it to recognize "Completed" as not the last thing in
3857 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3858 patch that reportedly fixes this has been added. I am not expert enough to
3859 create a test for it. This is what the patch creator wrote:
3861 "I found a little strange behaviour of ldap code when working with
3862 Windows 2003 AD Domain, where users was placed in more than one
3863 Organization Units. When I tried to give exim partial DN, the exit code
3864 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3865 But simultaneously result of request was absolutely normal ldap result,
3866 so I produce this patch..."
3868 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3869 REFERENCE, so I have modified the code to exclude the patch when that macro
3872 55. Some experimental protocols are using DNS PTR records for new purposes. The
3873 keys for these records are domain names, not reversed IP addresses. The
3874 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3875 leaves it alone. Component reversal etc. now happens only for IP addresses.
3878 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3880 57. Double the size of the debug message buffer (to 2048) so that more of very
3881 long debug lines gets shown.
3883 58. The exicyclog utility now does better if the number of log files to keep
3884 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3886 59. Two changes related to the smtp_active_hostname option:
3888 (1) $smtp_active_hostname is now available as a variable.
3889 (2) The default for smtp_banner uses $smtp_active_hostname instead
3890 of $primary_hostname.
3892 60. The host_aton() function is supposed to be passed a string that is known
3893 to be a valid IP address. However, in the case of IPv6 addresses, it was
3894 not checking this. This is a hostage to fortune. Exim now panics and dies
3895 if the condition is not met. A case was found where this could be provoked
3896 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3897 components; fortuitously, this particular loophole had already been fixed
3898 by change 4.50/55 above.
3900 If there are any other similar loopholes, the new check in host_aton()
3901 itself should stop them being exploited. The report I received stated that
3902 data on the command line could provoke the exploit when Exim was running as
3903 exim, but did not say which command line option was involved. All I could
3904 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3905 running as the user.
3908 61. There was a buffer overflow vulnerability in the SPA authentication code
3909 (which came originally from the Samba project). I have added a test to the
3910 spa_base64_to_bits() function which I hope fixes it.
3913 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3914 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3916 63. The daemon start-up calls getloadavg() while still root for those OS that
3917 need the first call to be done as root, but it missed one case: when
3918 deliver_queue_load_max is set with deliver_drop_privilege. This is
3919 necessary for the benefit of the queue runner, because there is no re-exec
3920 when deliver_drop_privilege is set.
3922 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3923 This has been fixed.
3925 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3926 was in use, was not putting the data itself into the right store pool;
3927 consequently, it could be overwritten for a subsequent message in the same
3928 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3931 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3933 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3934 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3935 the function were treating the return as a boolean value, which happened to
3936 work because 0=false and not-0=true, but is not correct code.
3938 68. The host_aton() function was not handling scoped IPv6 addresses (those
3939 with, for example, "%eth0" on the end) correctly.
3941 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3942 negated items (that is, ~something) in unsigned ints. Some compilers
3943 apparently mutter when there is no cast.
3945 70. If an address verification called from an ACL failed, and did not produce a
3946 user-specific message (i.e. there was only a "system" message), nothing was
3947 put in $acl_verify_message. In this situation, it now puts the system
3950 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3951 session; change 4.31/43 added the unwanted input to the log line - except
3952 that it did not do this in the start of session case. It now does.
3954 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3955 This is wrong and can cause the other end to generate a synchronization
3956 error if it is another Exim or anything else that does the synchronization
3957 check. A QUIT command is no longer sent after a timeout.
3959 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3960 during host lookups.
3962 74. The defer_ok option of callout verification was not working if it was used
3963 when verifying addresses in header lines, that is, for this case:
3965 verify = header_sender/callout=defer_ok
3967 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3968 those file descriptors could be used for SMTP connections. If anything
3969 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3970 could be sent to the SMTP client, causing chaos. The daemon now opens
3971 stdin, stdout, and stderr to /dev/null when it puts itself into the
3974 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3975 log by default. The user can override this with suitable Perl magic.
3977 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3978 the log message when discard triggers, was not working for the DATA ACL or
3979 for the non-SMTP ACL.
3981 78. Error message wording change in sieve.c.
3983 79. If smtp_accept_max_per_host was set, the number of connections could be
3984 restricted to fewer than expected, because the daemon was trying to set up
3985 a new connection before checking whether the processes handling previous
3986 connections had finished. The check for completed processes is now done
3987 earlier. On busy systems, this bug wouldn't be noticed because something
3988 else would have woken the daemon, and it would have reaped the completed
3991 80. If a message was submitted locally by a user whose login name contained one
3992 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3993 It caused a spool format error. I have fixed the spool reading code. A
3994 related problem was that the "from" clause in the Received: line became
3995 illegal because of the space(s). It is now covered by ${quote_local_part.
3997 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4000 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4002 83. Patch from Sieve maintainer to fix unterminated string problem in
4003 "vacation" handling.
4005 84. Some minor changes to the Linux configuration files to help with other
4006 OS variants using glibc.
4008 85. One more patch for Sieve to update vacation handling to latest spec.
4011 ----------------------------------------------------
4012 See the note above about the 4.44 and 4.50 releases.
4013 ----------------------------------------------------
4019 1. Change 4.43/35 introduced a bug that caused file counts to be
4020 incorrectly computed when quota_filecount was set in an appendfile
4023 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4024 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4027 3. Give more explanation in the error message when the command for a transport
4028 filter fails to execute.
4030 4. There are several places where Exim runs a non-Exim command in a
4031 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4032 was being done only for the command run by the queryprogram router. It is
4033 now done for all such subprocesses. The other cases are: ${run, transport
4034 filters, and the commands run by the lmtp and pipe transports.
4036 5. Some older OS have a limit of 256 on the maximum number of file
4037 descriptors. Exim was using setrlimit() to set 1000 as a large value
4038 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4039 systems. I've change it so that if it can't get 1000, it tries for 256.
4041 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4042 was an oversight, and furthermore, ever since the addition of extra
4043 controls (e.g. 4.43/32), the checks on when to allow different forms of
4044 "control" were broken. There should now be diagnostics for all cases when a
4045 control that does not make sense is encountered.
4047 7. $recipients is now available in the predata ACL (oversight).
4049 8. Tidy the search cache before the fork to do a delivery from a message
4050 received from the command line. Otherwise the child will trigger a lookup
4051 failure and thereby defer the delivery if it tries to use (for example) a
4052 cached ldap connection that the parent has called unbind on.
4054 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4055 of $address_data from the recipient verification was clobbered by the
4056 sender verification.
4058 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4059 was its contents. (It was OK if the option was not defined at all.)
4061 11. A "Completed" log line is now written for messages that are removed from
4062 the spool by the -Mrm option.
4064 12. $host_address is now set to the target address during the checking of
4065 ignore_target_hosts.
4067 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4068 being passed; this would have caused $sender_host_name to have been used if
4069 matching the list had actually called for a host name (not very likely,
4070 since this list is usually IP addresses). A host name is now passed as
4073 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4074 code that specifies a non-restarting handler (typically sigaction() in
4075 modern systems) in an attempt to fix a rare and obscure crash bug.
4077 15. Narrowed the window for a race in the daemon that could cause it to ignore
4078 SIGCHLD signals. This is not a major problem, because they are used only to
4079 wake it up if nothing else does.
4081 16. A malformed maildirsize file could cause Exim to calculate negative values
4082 for the mailbox size or file count. Odd effects could occur as a result.
4083 The maildirsize information is now recalculated if the size or filecount
4086 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4087 support for a long time. Removed HAVE_SYS_VFS_H.
4089 18. Updated exipick to current release from John Jetmore.
4091 19. Allow an empty sender to be matched against a lookup in an address list.
4092 Previously the only cases considered were a regular expression, or an
4095 20. Exim went into a mad DNS lookup loop when doing a callout where the
4096 host was specified on the transport, if the DNS lookup yielded more than
4099 21. The RFC2047 encoding function was originally intended for short strings
4100 such as real names; it was not keeping to the 75-character limit for
4101 encoded words that the RFC imposes. It now respects the limit, and
4102 generates multiple encoded words if necessary. To be on the safe side, I
4103 have increased the buffer size for the ${rfc2047: expansion operator from
4106 22. Failure to deliver a bounce message always caused it to be frozen, even if
4107 there was an errors_to setting on the router. The errors_to setting is now
4110 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4111 canonical form (fully expanded) before being placed in
4112 $sender_host_address.
4114 24. Updated eximstats to version 1.33
4116 25. Include certificate and key file names in error message when GnuTLS fails
4117 to set them up, because the GnuTLS error message doesn't include the name
4118 of the failing file when there is a problem reading it.
4120 26. Expand error message when OpenSSL has problems setting up cert/key files.
4123 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4124 (this can affect the format of dates).
4126 28. exim_tidydb, when checking for the continued existence of a message for
4127 which it has found a message-specific retry record, was not finding
4128 messages that were in split spool directories. Consequently, it was
4129 deleting retry records that should have stayed in existence.
4131 29. eximstats updated to version 1.35
4132 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4133 1.35 - bugfix such that pie charts by volume are generated correctly
4135 30. The SPA authentication driver was not abandoning authentication and moving
4136 on to the next authenticator when an expansion was forced to fail,
4137 contradicting the general specification for all authenticators. Instead it
4138 was generating a temporary error. It now behaves as specified.
4140 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4141 (the order specifies the preference for clients). The order is now AES256,
4142 AES128, 3DES, ARCFOUR128.
4144 31. Small patch to Sieve code - explicitly set From: when generating an
4147 32. Exim crashed if a remote delivery caused a very long error message to be
4148 recorded - for instance if somebody sent an entire SpamAssassin report back
4149 as a large number of 550 error lines. This bug was coincidentally fixed by
4150 increasing the size of one of Exim's internal buffers (big_buffer) that
4151 happened as part of the Exiscan merge. However, to be on the safe side, I
4152 have made the code more robust (and fixed the comments that describe what
4155 33. Some experimental protocols are using DNS PTR records for new purposes. The
4156 keys for these records are domain names, not reversed IP addresses. The
4157 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4158 leaves it alone. Component reversal etc. now happens only for IP addresses.
4161 34. The host_aton() function is supposed to be passed a string that is known
4162 to be a valid IP address. However, in the case of IPv6 addresses, it was
4163 not checking this. This is a hostage to fortune. Exim now panics and dies
4164 if the condition is not met. A case was found where this could be provoked
4165 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4166 components; fortuitously, this particular loophole had already been fixed
4167 by change 4.50/55 or 4.44/33 above.
4169 If there are any other similar loopholes, the new check in host_aton()
4170 itself should stop them being exploited. The report I received stated that
4171 data on the command line could provoke the exploit when Exim was running as
4172 exim, but did not say which command line option was involved. All I could
4173 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4174 running as the user.
4177 35. There was a buffer overflow vulnerability in the SPA authentication code
4178 (which came originally from the Samba project). I have added a test to the
4179 spa_base64_to_bits() function which I hope fixes it.
4182 36. The daemon start-up calls getloadavg() while still root for those OS that
4183 need the first call to be done as root, but it missed one case: when
4184 deliver_queue_load_max is set with deliver_drop_privilege. This is
4185 necessary for the benefit of the queue runner, because there is no re-exec
4186 when deliver_drop_privilege is set.
4188 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4189 was in use, was not putting the data itself into the right store pool;
4190 consequently, it could be overwritten for a subsequent message in the same
4191 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4194 38. Sometimes the final signoff response after QUIT could fail to get
4195 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4196 < 0 before doing a fflush(). This bug looks as though it goes back to the
4197 introduction of TLS in release 3.20, but "sometimes" must have been rare
4198 because the tests only now provoked it.
4204 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4205 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4206 Now it returns an integer. A number of calls were still expecting a T/F
4207 return. Fortuitously, in all cases, the tests worked in OK situations,
4208 which is the norm. However, things would have gone wrong on any write
4209 failures on the smtp file descriptor. This function is used when sending
4210 messages over SMTP and also when doing verify callouts.
4212 2. When Exim is called to do synchronous delivery of a locally submitted
4213 message (the -odf or -odi options), it no longer closes stderr before doing
4216 3. Implemented the mua_wrapper option.
4218 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4220 5. Implemented the functions header_remove(), header_testname(),
4221 header_add_at_position(), and receive_remove_recipient(), and exported them
4224 6. If an ACL "warn" statement specified the addition of headers, Exim already
4225 inserted X-ACL-Warn: at the start if there was no header name. However, it
4226 was not making this test for the second and subsequent header lines if
4227 there were newlines in the string. This meant that an invalid header could
4228 be inserted if Exim was badly configured.
4230 7. Allow an ACL "warn" statement to add header lines at the start or after all
4231 the Received: headers, as well as at the end.
4233 8. Added the rcpt_4xx retry error code.
4235 9. Added postmaster_mailfrom=xxx to callout verification option.
4237 10. Added mailfrom=xxxx to the callout verification option, for verify=
4240 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4241 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4243 12. Inserted some casts to stop certain compilers warning when using pointer
4244 differences as field lengths or precisions in printf-type calls (mostly
4245 affecting debugging statements).
4247 13. Added optional readline() support for -be (dynamically loaded).
4249 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4250 same clock tick as a message's arrival, so that its received time was the
4251 same as the "first fail" time on the retry record, and that message
4252 remained on the queue past the ultimate address timeout, every queue runner
4253 would try a delivery (because it was past the ultimate address timeout) but
4254 after another failure, the ultimate address timeout, which should have then
4255 bounced the address, did not kick in. This was a "< instead of <=" error;
4256 in most cases the first failure would have been in the next clock tick
4257 after the received time, and all would be well.
4259 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4260 being recognized when the domain list was tested by the match_domain
4261 condition in an expansion string.
4263 16. Added the ${str2b64: operator.
4265 17. Exim was always calling setrlimit() to set a large limit for the number of
4266 processes, without checking whether the existing limit was already
4267 adequate. (It did check for the limit on file descriptors.) Furthermore,
4268 errors from getrlimit() and setrlimit() were being ignored. Now they are
4269 logged to the main and panic logs, but Exim does carry on, to try to do its
4270 job under whatever limits there are.
4272 18. Imported PCRE 5.0.
4274 19. Trivial typo in log message " temporarily refused connection" (the leading
4277 20. If the log selector return_path_on_delivery was set and an address was
4278 redirected to /dev/null, the delivery process crashed because it assumed
4279 that a return path would always be set for a "successful" delivery. In this
4280 case, the whole delivery is bypassed as an optimization, and therefore no
4283 21. Internal re-arrangement: the function for sending a challenge and reading
4284 a response while authentication was assuming a zero-terminated challenge
4285 string. It's now changed to take a pointer and a length, to allow for
4286 binary data in such strings.
4288 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4290 23. Exim was not respecting finduser_retries when seeking the login of the
4291 uid under which it was called; it was always trying 10 times. (The default
4292 setting of finduser_retries is zero.) Also, it was sleeping after the final
4293 failure, which is pointless.
4295 24. Implemented tls_on_connect_ports.
4297 25. Implemented acl_smtp_predata.
4299 26. If the domain in control=submission is set empty, Exim assumes that the
4300 authenticated id is a complete email address when it generates From: or
4301 Sender: header lines.
4303 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4304 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4305 chown and chgrp in /bin and hostname in /usr/bin.
4307 28. Exim was keeping the "process log" file open after each use, just as it
4308 does for the main log. This opens the possibility of it remaining open for
4309 long periods when the USR1 signal hits a daemon. Occasional processlog
4310 errors were reported, that could have been caused by this. Anyway, it seems
4311 much more sensible not to leave this file open at all, so that is what now
4314 29. The long-running daemon process does not normally write to the log once it
4315 has entered its main loop, and it closes the log before doing so. This is
4316 so that log files can straightforwardly be renamed and moved. However,
4317 there are a couple of unusual error situations where the daemon does write
4318 log entries, and I had neglected to close the log afterwards.
4320 30. The text of an SMTP error response that was received during a remote
4321 delivery was being truncated at 512 bytes. This is too short for some of
4322 the long messages that one sometimes sees. I've increased the limit to
4325 31. It is now possible to make retry rules that apply only when a message has a
4326 specific sender, in particular, an empty sender.
4328 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4329 it possible to be selective about when SMTP synchronization is enforced.
4331 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4333 32. Implemented hosts_connection_nolog.
4335 33. Added an ACL for QUIT.
4337 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4340 35. Added mailbox_size and mailbox_filecount to appendfile.
4342 36. Added control = no_multiline_responses to ACLs.
4344 37. There was a bug in the logic of the code that waits for the clock to tick
4345 in the case where the clock went backwards by a substantial amount such
4346 that the microsecond fraction of "now" was more than the microsecond
4347 fraction of "then" (but the whole seconds number was less).
4349 38. Added support for the libradius Radius client library this is found on
4350 FreeBSD (previously only the radiusclient library was supported).
4356 1. When certain lookups returned multiple values in the form name=value, the
4357 quoting of the values was not always being done properly. Specifically:
4358 (a) If the value started with a double quote, but contained no whitespace,
4360 (b) If the value contained whitespace other than a space character (i.e.
4361 tabs or newlines or carriage returns) it was not quoted.
4362 This fix has been applied to the mysql and pgsql lookups by writing a
4363 separate quoting function and calling it from the lookup code. The fix
4364 should probably also be applied to nisplus, ibase and oracle lookups, but
4365 since I cannot test any of those, I have not disturbed their existing code.
4367 2. A hit in the callout cache for a specific address caused a log line with no
4368 reason for rejecting RCPT. Now it says "Previous (cached) callout
4369 verification failure".
4371 3. There was an off-by-one bug in the queryprogram router. An over-long
4372 return line was truncated at 256 instead of 255 characters, thereby
4373 overflowing its buffer with the terminating zero. As well as fixing this, I
4374 have increased the buffer size to 1024 (and made a note to document this).
4376 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4377 when Exim is waiting for an SMTP response from a remote server, Exim
4378 restarts its select() call on the socket, thereby resetting its timeout.
4379 This is not a problem when such interrupts are rare. Somebody set up a cron
4380 job to run exiwhat every 2 minutes, which is less than the normal select()
4381 timeout (5 or 10 minutes). This meant that the select() timeout never
4382 kicked in because it was always reset. I have fixed this by comparing the
4383 time when an interrupt arrives with the time at the start of the first call
4384 to select(). If more time than the timeout has elapsed, the interrupt is
4385 treated as a timeout.
4387 5. Some internal re-factoring in preparation for the addition of Sieve
4388 extensions (by MH). In particular, the "personal" test is moved to a
4389 separate function, and given an option for scanning Cc: and Bcc: (which is
4390 not set for Exim filters).
4392 6. When Exim created an email address using the login of the caller as the
4393 local part (e.g. when creating a From: or Sender: header line), it was not
4394 quoting the local part when it contained special characters such as @.
4396 7. Installed new OpenBSD configuration files.
4398 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4399 try to make them clearer.
4401 9. Callout options, other than the timeout value, were being ignored when
4402 verifying sender addresses in header lines. For example, when using
4404 verify = header_sender/callout=no_cache
4406 the cache was (incorrectly) being used.
4408 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4409 only the Cygwin environment.
4411 11. When return_path_on_delivery was set as a log selector, if different remote
4412 addresses in the same message used different return paths and parallel
4413 remote delivery occurred, the wrong values would sometimes be logged.
4414 (Whenever a remote delivery process finished, the return path value from
4415 the most recently started remote delivery process was logged.)
4417 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4418 lines when AUTH and/or TLS are in use. This is the "received protocol"
4419 field. Exim used to use "asmtp" for authenticated SMTP, without any
4420 indication (in the protocol name) for TLS use. Now it follows the RFC and
4421 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4422 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4423 names appear in log lines as well as in Received: header lines.
4425 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4426 extensions, and comparison tests, and to fix some bugs.
4428 14. Changes to the "personal" filter test:
4430 (1) The test was buggy in that it was just doing the equivalent of
4431 "contains" tests on header lines. For example, if a user's address was
4432 anne@some.where, the "personal" test would incorrectly be true for
4434 To: susanne@some.where
4436 This test is now done by extracting each address from the header in turn,
4437 and checking the entire address. Other tests that are part of "personal"
4438 are now done using regular expressions (for example, to check local parts
4439 of addresses in From: header lines).
4441 (2) The list of non-personal local parts in From: addresses has been
4442 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4443 taken from the Sieve specification recommendations.
4445 (3) If the message contains any header line starting with "List-" it is
4446 treated as non-personal.
4448 (4) The test for "circular" in the Subject: header line has been removed
4449 because it now seems ill-conceived.
4451 15. Minor typos in src/EDITME comments corrected.
4453 16. Installed latest exipick from John Jetmore.
4455 17. If headers_add on a router specified a text string that was too long for
4456 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4457 of string_sprintf() is now avoided.
4459 18. $message_body_size was not set (it was always zero) when running the DATA
4460 ACL and the local_scan() function.
4462 19. For the "mail" command in an Exim filter, no default was being set for
4463 the once_repeat time, causing a random time value to be used if "once" was
4464 specified. (If the value happened to be <= 0, no repeat happened.) The
4465 default is now 0s, meaning "never repeat". The "vacation" command was OK
4466 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4467 (I found it when inspecting the code).
4469 20. There is now an overall timeout for performing a callout verification. It
4470 defaults to 4 times the callout timeout, which applies to individual SMTP
4471 commands during the callout. The overall timeout applies when there is more
4472 than one host that can be tried. The timeout is checked before trying the
4473 next host. This prevents very long delays if there are a large number of
4474 hosts and all are timing out (e.g. when the network connections are timing
4475 out). The value of the overall timeout can be changed by specifying an
4476 additional sub-option for "callout", called "maxwait". For example:
4478 verify = sender/callout=5s,maxwait=20s
4480 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4481 to the end before writing, but this should make it even safer).
4483 22. Exim was forgetting that it had advertised PIPELINING for the second and
4484 subsequent messages on an SMTP connection. It was also not resetting its
4485 memory on STARTTLS and an internal HELO.
4487 23. When Exim logs an SMTP synchronization error within a session, it now
4488 records whether PIPELINING has been advertised or not.
4490 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4491 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4492 rather than long int.
4494 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4496 26. Added the never_mail option to autoreply.
4502 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4503 crash if the getsockname() call failed; this can happen if a connection is
4504 closed very soon after it is established. The problem was simply in the
4505 order in which certain operations were done, causing Exim to try to write
4506 to the SMTP stream before it had set up the file descriptor. The bug has
4507 been fixed by making things happen in the correct order.
4513 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4514 before the connection was closed, thus losing the rejection response.
4516 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4517 some early Solaris releases, but causes trouble in current releases where
4518 socklen_t is defined.
4520 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4523 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4526 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4527 that releasing the top of it at the end releases what was used for sub-
4528 expansions (unless the block got too big). However, discard this block if
4529 the first thing is a variable or header, so that we can use its block when
4530 it is dynamic (useful for very large $message_headers, for example).
4532 6. Lookups now cache *every* query, not just the most recent. A new, separate
4533 store pool is used for this. It can be recovered when all lookup caches are
4534 flushed. Lookups now release memory at the end of their result strings.
4535 This has involved some general refactoring of the lookup sources.
4537 7. Some code has been added to the store_xxx() functions to reduce the amount
4538 of flapping under certain conditions.
4540 8. log_incoming_interface used to affect only the <= reception log lines. Now
4541 it causes the local interface and port to be added to several more SMTP log
4542 lines, for example "SMTP connection from", and rejection lines.
4544 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4546 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4548 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4549 could be overwritten at the end of the current message (or the start of a
4550 new message if it was set in a HELO ACL). The value is now preserved for
4551 the duration of the SMTP connection.
4553 12. If a transport had a headers_rewrite setting, and a matching header line
4554 contained an unqualified address, that address was qualified, even if it
4555 did not match any rewriting rules. The underlying bug was that the values
4556 of the flags that permit the existence of unqualified sender and recipient
4557 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4558 non-local messages, and by -bnq for local messages) were not being
4559 preserved with the message after it was received.
4561 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4562 "next input=" as part of the text comprising the host identity instead of
4563 the correct text. The code was using the same buffer for two different
4564 strings. However, depending on which order the printing function evaluated
4565 its arguments, the bug did not always show up. Under Linux, for example, my
4566 test suite worked just fine.
4568 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4569 doesn't work with some older versions of Perl. It has been changed to "my",
4570 which in any case is probably the better facility to use.
4572 15. A really picky compiler found some instances of statements for creating
4573 error messages that either had too many or two few arguments for the format
4576 16. The size of the buffer for calls to the DNS resolver has been increased
4577 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4578 for addresses that have a lot of PTR records. This alleviates a problem; it
4579 does not fully solve it.
4581 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4582 buffer now truncates the list and logs the incident, which is the same
4583 action as happens when Exim is looking up a host name and its aliases.
4584 Previously in this situation something unpredictable would happen;
4585 sometimes it was "internal error: store_reset failed".
4587 18. If a server dropped the connection unexpectedly when an Exim client was
4588 using GnuTLS and trying to read a response, the client delivery process
4589 crashed while trying to generate an error log message.
4591 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4592 string, for example:
4594 warn message = H1: something\nH2: something
4596 the text was added as a single header line from Exim's point of view
4597 though it ended up OK in the delivered message. However, searching for the
4598 second and subsequent header lines using $h_h2: did not work. This has been
4599 fixed. Similarly, if a system filter added multiple headers in this way,
4600 the routers could not see them.
4602 20. Expanded the error message when iplsearch is called with an invalid key to
4603 suggest using net-iplsearch in a host list.
4605 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4606 ACLs are no longer actually imposed (and a message to that effect is
4609 22. If a "gecos" field in a passwd entry contained escaped characters, in
4610 particular, if it contained a \" sequence, Exim got it wrong when building
4611 a From: or a Sender: header from that name. A second bug also caused
4612 incorrect handling when an unquoted " was present following a character
4613 that needed quoting.
4615 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4616 was not being matched caselessly.
4618 24. Arranged for all hyphens in the exim.8 source to be escaped with
4621 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4622 properly. Recipient callout cache records were still being keyed to include
4623 the sender, even when use_sender was set false. This led to far more
4624 callouts that were necessary. The sender is no longer included in the key
4625 when use_sender is false.
4627 26. Added "control = submission" modifier to ACLs.
4629 27. Added the ${base62d: operator to decode base 62 numbers.
4631 28. dnsdb lookups can now access SRV records.
4633 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4634 the configuration file.
4636 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4637 (-v) mode. This makes the output for a verbose queue run more intelligible.
4639 31. Added a use_postmaster feature to recipient callouts.
4641 32. Added the $body_zerocount variable, containing the number of binary zero
4642 bytes in the message body.
4644 33. The time of last modification of the "new" subdirectory is now used as the
4645 "mailbox time last read" when there is a quota error for a maildir
4648 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4650 35. Added +ignore_unknown as a special item in host lists.
4652 36. Code for decoding IPv6 addresses in host lists is now included, even if
4653 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4654 address was recognized as an IP address, but was then not correctly decoded
4655 into binary, causing unexpected and incorrect effects when compared with
4662 1. Very minor rewording of debugging text in manualroute to say "list of
4663 hosts" instead of "hostlist".
4665 2. If verify=header_syntax was set, and a header line with an unqualified
4666 address (no domain) and a large number of spaces between the end of the
4667 name and the colon was received, the reception process suffered a buffer
4668 overflow, and (when I tested it) crashed. This was caused by some obsolete
4669 code that should have been removed. The fix is to remove it!
4671 3. When running in the test harness, delay a bit after writing a bounce
4672 message to get a bit more predictability in the log output.
4674 4. Added a call to search_tidyup() just before forking a reception process. In
4675 theory, someone could use a lookup in the expansion of smtp_accept_max_
4676 per_host which, without the tidyup, could leave open a database connection.
4678 5. Added the variables $recipient_data and $sender_data which get set from a
4679 lookup success in an ACL "recipients" or "senders" condition, or a router
4680 "senders" option, similar to $domain_data and $local_part_data.
4682 6. Moved the writing of debug_print from before to after the "senders" test
4685 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4686 problems for message scanning, either using a data ACL, or using
4687 local_scan() because the Received: header was not generated till after they
4688 were called (in order to set the time as the time of reception completion).
4689 I have revised the way this works. The header is now generated after the
4690 body is received, but before the ACL or local_scan() are called. After they
4691 are run, the timestamp in the header is updated.
4697 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4698 before starting a queue runner without re-exec. This happened only when
4699 deliver_drop_privilege was set or when the Exim user was set to root. The
4700 effect of the bug was that timeouts during subsequent deliveries caused
4701 crashes instead of being properly handled. The handler is now left at its
4702 default (and expected) setting.
4704 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4705 message, again when deliver_drop_privilege is set or Exim is run as root.
4706 The bug described in (1) was not present in this case, but the tidying up
4707 of the other signals was missing. I have made the two cases consistent.
4709 3. The ignore_target_hosts setting on a manualroute router was being ignored
4710 for hosts that were looked up using the /MX notation.
4712 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4715 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4716 operated on the sender address. After changing the $sender_address to <>
4717 for the sender address verify, Exim was re-instated it as the original
4718 (before rewriting) address, but remembering that it had rewritten it, so it
4719 wasn't rewriting it again. This bug also had the effect of breaking the
4720 sender address verification caching when the sender address was rewritten.
4722 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4723 This has been changed so that if the ip literal address matches
4724 ignore_target_hosts, the router declines.
4726 7. Added expansion conditions match_domain, match_address, and match_local_
4727 part (NOT match_host).
4729 8. The placeholder for the Received: header didn't have a length field set.
4731 9. Added code to Exim itself and to exim_lock to test for a specific race
4732 condition that could lead to file corruption when using MBX delivery. The
4733 issue is with the lockfile that is created in /tmp. If this file is removed
4734 after a process has opened it but before that process has acquired a lock,
4735 there is the potential for a second process to recreate the file and also
4736 acquire a lock. This could lead to two Exim processes writing to the file
4737 at the same time. The added code performs the same test as UW imapd; it
4738 checks after acquiring the lock that its file descriptor still refers to
4739 the same named file.
4741 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4742 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4745 11. Added the smtp_active_hostname option. If used, this will typically be made
4746 to depend on the incoming interface address. Because $interface_address is
4747 not set up until the daemon has forked a reception process, error responses
4748 that can happen earlier (such as "too many connections") no longer contain
4751 12. If an expansion in a condition on a "warn" statement fails because a lookup
4752 defers, the "warn" statement is abandoned, and the next ACL statement is
4753 processed. Previously this caused the whole ACL to be aborted.
4755 13. Added the iplsearch lookup type.
4757 14. Added ident_timeout as a log selector.
4759 15. Added tls_certificate_verified as a log selector.
4761 16. Added a global option tls_require_ciphers (compare the smtp transport
4762 option of the same name). This controls incoming TLS connections.
4764 17. I finally figured out how to make tls_require_ciphers do a similar thing
4765 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4766 before starting the TLS session.
4768 18. Tabs are now shown as \t in -bP output.
4770 19. If the log selector return_path_on_delivery was set, Exim crashed when
4771 bouncing a message because it had too many Received: header lines.
4773 20. If two routers both had headers_remove settings, and the first one included
4774 a superfluous trailing colon, the final name in the first list and the
4775 first name in the second list were incorrectly joined into one item (with a
4776 colon in the middle).
4782 1. Added -C and -D options to the exinext utility, mainly to make it easier
4783 to include in the automated testing, but these could be helpful when
4784 multiple configurations are in use.
4786 2. The exinext utility was not formatting the output nicely when there was
4787 an alternate port involved in the retry record key, nor when there was a
4788 message id as well (for retries that were specific to a specific message
4789 and a specific host). It was also confused by IPv6 addresses, because of
4790 the additional colons they contain. I have fixed the IPv4 problem, and
4791 patched it up to do a reasonable job for IPv6.
4793 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4794 delivery, the log line now contains "pipelined" if PIPELINING was used.
4796 4. An SMTP transport process used to panic and die if the bind() call to set
4797 an explicit outgoing interface failed. This has been changed; it is now
4798 treated in the same way as a connect() failure.
4800 5. A reference to $sender_host_name in the part of a conditional expansion
4801 that was being skipped was still causing a DNS lookup. This no longer
4804 6. The def: expansion condition was not recognizing references to header lines
4805 that used bh_ and bheader_.
4807 7. Added the _cache feature to named lists.
4809 8. The code for checking quota_filecount in the appendfile transport was
4810 allowing one more file than it should have been.
4812 9. For compatibility with Sendmail, the command line option
4820 and sets the incoming protocol and host name (for trusted callers). The
4821 host name and its colon can be omitted when only the protocol is to be set.
4822 Note the Exim already has two private options, -pd and -ps, that refer to
4823 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4824 "s", but I don't think that's a major issue.
4826 10. A number of refactoring changes to the code, none of which should affect
4829 (a) The number of logging options was getting close to filling up the
4830 32-bit word that was used as a bit map. I have split them into two classes:
4831 those that are passed in the argument to log_write(), and those that are
4832 only ever tested independently outside of that function. These are now in
4833 separate 32-bit words, so there is plenty of room for expansion again.
4834 There is no change in the user interface or the logging behaviour.
4836 (b) When building, for example, log lines, the code previously used a
4837 macro that called string_cat() twice, in order to add two strings. This is
4838 not really sufficiently general. Furthermore, there was one instance where
4839 it was actually wrong because one of the argument was used twice, and in
4840 one call a function was used. (As it happened, calling the function twice
4841 did not affect the overall behaviour.) The macro has been replaced by a
4842 function that can join an arbitrary number of extra strings onto a growing
4845 (c) The code for expansion conditions now uses a table and a binary chop
4846 instead of a serial search (which was left over from when there were very
4847 few conditions). Also, it now recognizes conditions like "pam" even when
4848 the relevant support is not compiled in: a suitably worded error message is
4849 given if an attempt is made to use such a condition.
4851 11. Added ${time_interval:xxxxx}.
4853 12. A bug was causing one of the ddress fields not to be passed back correctly
4854 from remote delivery subprocesses. The field in question was not being
4855 subsequently used, so this caused to problems in practice.
4857 13. Added new log selectors queue_time and deliver_time.
4859 14. Might have fixed a bug in maildirsizefile handling that threw up
4860 "unexpected character" debug warnings, and recalculated the data
4861 unnecessarily. In any case, I expanded the warning message to give more
4864 15. Added the message "Restricted characters in address" to the statements in
4865 the default ACL that block characters like @ and % in local parts.
4867 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4868 Three changes have been made:
4870 (a) There was a serious bug; a negative response to MAIL caused the whole
4871 recipient domain to be cached as invalid, thereby blocking all messages
4872 to all local parts at the same domain, from all senders. This bug has
4873 been fixed. The domain is no longer cached after a negative response to
4874 MAIL if the sender used is not empty.
4876 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4879 (c) A new callout option, "use_sender" has been added for people who want
4880 the modified behaviour.
4886 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4889 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4890 indeed breaks things for older releases.
4892 3. Added additional logging to the case where there is a problem reading data
4893 from a filter that is running in a subprocess using a pipe, in order to
4894 try to track down a specific problem.
4896 4. Testing facility fudge: when running in the test harness and attempting
4897 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4898 getting "No route to host". Convert this to a timeout.
4900 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4903 6. Some OS don't have socklen_t but use size_t instead. This affects the
4904 fifth argument of getsockopt() amongst other things. This is now
4905 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4906 can be set for individual OS. I have set it for SunOS5, OSF1, and
4907 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4908 some earlier ones do not.
4910 7. Change 4.30/15 was not doing the test caselessly.
4912 8. The standard form for an IPv6 address literal was being rejected by address
4913 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4914 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4915 this, as well as the form without the "IPv6" on the front (but only when
4916 address literals are enabled, of course).
4918 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4920 10. Exim crashed if a message with an empty sender address specified by -f
4921 encountered a router with an errors_to setting. This could be provoked only
4922 by a command such as
4926 where an empty string was supplied; "<>" did not hit this bug.
4928 11. Installed PCRE release 4.5.
4930 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4931 remained set. It is now erased.
4933 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4934 times from message ids (which are base 36 rather than the normal 62).
4936 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4937 were being counted as actual protocol errors, and logged if the log
4938 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4939 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4940 and DATA following a set of rejected RCPTs do not count as protocol errors.
4941 In other words, Exim assumes they were pipelined, though this may not
4942 actually be the case. Of course, in all cases the client gets an
4943 appropriate error code.
4945 15. If a lookup fails in an ACL condition, a message about the failure may
4946 be available; it is used if testing the ACL cannot continue, because most
4947 such messages specify what the cause of the deferral is. However, some
4948 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4949 that caused an old message to be retained and used if a later statement
4950 caused a defer, replacing the real cause of the deferral.
4952 16. If an IP address had so many PTR records that the DNS lookup buffer
4953 was not large enough to hold them, Exim could crash while trying to process
4954 the truncated data. It now detects and logs this case.
4956 17. Further to 4.21/58, another change has been made: if (and only if) the
4957 first line of a message (the first header line) ends with CRLF, a bare LF
4958 in a subsequent header line has a space inserted after it, so as not to
4959 terminate the header.
4961 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4962 unnecessarily, used atoi() instead of strtol(), and didn't check the
4963 termination when getting file sizes from file names by regex.
4965 19. Completely re-implemented the support for maildirsize files, in the light
4966 of a number of problems with the previous contributed implementation
4967 (4.30/29). In particular:
4969 . If the quota is zero, the maildirsize file is maintained, but no quota is
4972 . If the maildir directory does not exist, it is created before any attempt
4973 to write a maildirsize file.
4975 . The quota value in the file is just a cache; if the quota is changed in
4976 the transport, the new value overrides.
4978 . A regular expression is available for excluding directories from the
4981 20. The autoreply transport checks the characters in options that define the
4982 message's headers; it allows continued headers, but it was checking with
4983 isspace() after an embedded newline instead of explicitly looking for a
4986 21. If all the "regular" hosts to which an address was routed had passed their
4987 expiry times, and had not reached their retry times, the address was
4988 bounced, even if fallback hosts were defined. Now Exim should go on to try
4991 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4992 equivalent code in the SMTP transport. Some hosts send humungous responses
4993 to HELO/EHLO, more than 1024 it seems.
4995 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4996 gives compiler warnings in some environments. I've now done it "properly",
4999 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5000 (because of problems with the built-in one) was declared to return uschar *
5001 instead of char *, causing compiler failure.
5003 25. Fixed a file descriptor leak when processing alias/forward files.
5005 26. Fixed a minor format string issue in dbfn.c.
5007 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5009 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5010 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5013 29. When a sender address is verified, it is cached, to save repeating the test
5014 when there is more than one recipient in a message. However, when the
5015 verification involves a callout, it is possible for different callout
5016 options to be set for different recipients. It is too complicated to keep
5017 track of this in the cache, so now Exim always runs a verification when a
5018 callout is required, relying on the callout cache for the optimization.
5019 The overhead is duplication of the address routing, but this should not be
5022 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5023 to be verified with callout=postmaster, and the main callout worked but the
5024 postmaster check failed, the verification correctly failed. However, if a
5025 subsequent RCPT command asked for sender verification *without* the
5026 postmaster check, incorrect caching caused this verification also to fail,
5029 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5030 it was not caching the DNS options (qualify_single, search_parents) that
5031 were used when the lookup failed. A subsequent lookup with different
5032 options therefore always gave the same answer, though there were cases
5033 where it should not have. (Example: a "domains = !$mx_any" option on a
5034 dnslookup router: the "domains" option is always processed without any
5035 widening, but the router might have qualify_single set.) Now Exim uses the
5036 cached value only when the same options are set.
5038 32. Added John Jetmore's "exipick" utility to the distribution.
5040 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5041 than a timeout (e.g. a certificate is required, and is not provided), an
5042 Exim server now closes the connection immediately. Previously it waited for
5043 the client to close - but if the client is SSL, it seems that they each
5044 wait for each other, leading to a delay before one of them times out.
5046 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5047 maintained 0.8.x compatibility because I don't think many are using it, and
5048 it is clearly obsolete.
5050 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5053 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5054 client certificate was expired. A simple patch fixes this, though I don't
5055 understand the full logic of why the verify callback is called multiple
5058 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5059 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5060 which causes problems with some clients (such as the Certicom SSL Plus
5061 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5062 disables the coutermeasure allowing Eudora to connect."
5064 38. Exim was not checking that a write() to a log file succeeded. This could
5065 lead to Bad Things if a log got too big, in particular if it hit a file
5066 size limit. Exim now panics and dies if it cannot write to a log file, just
5067 as it does if it cannot open a log file.
5069 39. Modified OS/Makefile-Linux so that it now contains
5071 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5073 The two -D definitions ensure that Exim is compiled with large file
5074 support, which makes it possible to handle log files that are bigger than
5077 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5078 instance) a domain was checked against a named list that involved a lookup,
5079 causing $domain_data to be set, then another domain was checked against the
5080 same list, then the first domain was re-checked, the value of $domain_data
5081 after the final check could be wrong. In particular, if the second check
5082 failed, it could be set empty. This bug probably also applied to
5085 41. The strip_trailing_dot option was not being applied to the address given
5086 with the -f command-line option.
5088 42. The code for reading a message's header from the spool was incrementing
5089 $received_count, but never initializing it. This meant that the value was
5090 incorrect (doubled) while delivering a message in the same process in which
5091 it was received. In the most common configuration of Exim, this never
5092 happens - a fresh exec is done - but it can happen when
5093 deliver_drop_privilege is set.
5095 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5096 it now includes up to 150 characters of the unexpected data in the log
5099 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5100 and building data strings. The size of both of these buffers was 10 000
5101 bytes - far larger than anybody would *ever* want, thought I. Needless to
5102 say, somebody hit the limit. I have increased the maximum line length to
5103 20 000 and the maximum data length of concatenated lines to 100 000. I have
5104 also fixed two bugs, because there was no checking on these buffers. Tsk,
5105 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5106 buffer is too small.
5108 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5109 lsearch lookups. Now it does.
5111 46. When parsing a route_list item in a manualroute router, a fixed-length
5112 buffer was used for the list of hosts. I made this 1024 bytes long,
5113 thinking that nobody would ever have a list of hosts that long. Wrong.
5114 Somebody had a whole pile of complicated expansion conditions, and the
5115 string was silently truncated, leading to an expansion error. It turns out
5116 that it is easier to change to an unlimited length (owing to other changes
5117 that have happened since this code was originally written) than to build
5118 structure for giving a limitation error. The length of the item that
5119 expands into the list of hosts is now unlimited.
5121 47. The lsearch lookup could not handle data where the length of text line was
5122 more than 4095 characters. Such lines were truncated, leading to shortened
5123 data being returned. It should now handle lines of any length.
5125 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5126 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5129 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5130 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5131 no longer happens in comment lines. A list of replacements is now placed
5132 at the head of all of the source files, except those whose only change is
5133 to replace PERL_COMMAND in the very first #! line.
5135 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5136 messages on the queue, with a bottom-up merge sort, using code contributed
5137 by Michael Haardt. This should make operations like -bp somewhat faster on
5138 large queues. It won't affect queue runners, except when queue_run_in_order
5141 51. Installed eximstats 1.31 in the distribution.
5143 52. Added support for SRV lookups to the dnslookup router.
5145 53. If an ACL referred to $message_body or $message_body_end, the value was not
5146 reset for any messages that followed in the same SMTP session.
5148 54. The store-handling optimization for building very long strings was not
5149 differentiating between the different store pools. I don't think this
5150 actually made any difference in practice, but I've tidied it.
5152 55. While running the routers to verify a sender address, $sender_address
5153 was still set to the sender address. This is wrong, because when routing to
5154 send a bounce to the sender, it would be empty. Therefore, I have changed
5155 it so that, while verifying a sender address, $sender_address is set to <>.
5156 (There is no change to what happens when verifying a recipient address.)
5158 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5159 target A or AAAA records (if not already returned) without resetting the
5160 qualify_single or search_parents options of the DNS resolver. These are
5161 inappropriate in this case because the targets of MX and SRV records must
5162 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5163 target that, when qualified, matched something in the local domain. These
5164 two options are now turned off when doing these lookups.
5166 57. It seems that at least some releases of Reiserfs (which does not have the
5167 concept of a fixed number of inodes) returns zero and not -1 for the
5168 number of available inodes. This interacted badly with check_spool_inodes,
5169 which assumed that -1 was the "no such thing" setting. What I have done is
5170 to check that the total number of inodes is greater than zero before doing
5171 the test of how many are available.
5173 58. When a "warn" ACL statement has a log_message modifier, the message is
5174 remembered, and not repeated. This is to avoid a lot of repetition when a
5175 message has many recipients that cause the same warning to be written.
5176 Howewer, Exim was preserving the list of already written lines for an
5177 entire SMTP session, which doesn't seem right. The memory is now reset if a
5178 new message is started.
5180 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5181 debugging output unless log_rewrite was also set.
5183 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5184 of (int)(handle) when we know that handle contains (void *)(-1).
5186 61. The Exim daemon panic-logs an error return when it closes the incoming
5187 connection. However "connection reset by peer" seems to be common, and
5188 isn't really an error worthy of noting specially, so that particular error
5191 62. When Exim is trying to find all the local interfaces, it used to panic and
5192 die if the ioctl to get the interface flags failed. However, it seems that
5193 on at least one OS (Solaris 9) it is possible to have an interface that is
5194 included in the list of interfaces, but for which you get a failure error
5195 for this call. This happens when the interface is not "plumbed" into a
5196 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5197 failure of the "get flags" call assumes that the interface is down.
5199 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5200 makes life easier for people who are doing arithmetic on fields extracted
5201 from dates, where you often get leading zeros that should not be
5202 interpreted as octal.
5204 64. Added qualify_domain to the redirect router, to override the global
5207 65. If a pathologically long header line contained very many addresses (the
5208 report of this problem mentioned 10 000) and each of them was rewritten,
5209 Exim could use up a very large amount of memory. (It kept on making new
5210 copies of the header line as it rewrote, and never released the old ones.)
5211 At the expense of a bit more processing, the header rewriting function has
5212 been changed so that it no longer eats memory in this way.
5214 66. The generation of the Received: header has been moved from the time that a
5215 message starts to be received, to the time that it finishes. The timestamp
5216 in the Received: header should now be very close to that of the <= log
5217 line. There are two side-effects of this change:
5219 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5220 logged header lines no longer include the local Received: line, because
5221 it has not yet been created. The same applies to a copy of the message
5222 that is returned to a non-SMTP sender when a message is rejected.
5224 (b) When a filter file is tested using -bf, no additional Received: header
5225 is added to the test message. After some thought, I decided that this
5228 This change does not affect the value of $received_for. It is still set
5229 after address rewriting, but before local_scan() is called.
5231 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5233 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5234 gave an unhelpful panic error message, and a defer error. I have managed to
5235 change this behaviour so that it now rejects any supplied certificate,
5236 which seems right, as the list of acceptable certificates is empty.
5238 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5239 gave an unhelpful defer error. I have not managed to make this reject any
5240 supplied certificates, but the error message it gives is "no certificate
5241 supplied", which is not helpful.
5243 70. exigrep's output now also includes lines that are not associated with any
5244 message, but which match the given pattern. Implemented by a patch from
5245 Martin Sluka, which also tidied up the Perl a bit.
5247 71. Recipient callout verification, like sender verification, was using <> in
5248 the MAIL FROM command. This isn't really the right thing, since the actual
5249 sender may affect whether the remote host accepts the recipient or not. I
5250 have changed it to use the actual sender in the callout; this means that
5251 the cache record is now keyed on a recipient/sender pair, not just the
5252 recipient address. There doesn't seem to be a real danger of callout loops,
5253 since a callout by the remote host to check the sender would use <>.
5254 [SEE ABOVE: changed after hitting problems.]
5256 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5257 temporary errors. However, in the case of such a code being given after
5258 the end of a data transmission (i.e. after ".") Exim was failing to write
5259 a retry record for the message. (Yes, there was some broken host that was
5260 actually sending 8xx at this point.)
5262 73. An unknown lookup type in a host list could cause Exim to panic-die when
5263 the list was checked. (An example that provoked this was putting <; in the
5264 middle of a list instead of at the start.) If this happened during a DATA
5265 ACL check, a -D file could be left lying around. This kind of configuration
5266 error no longer causes Exim to die; instead it causes a defer errror. The
5267 incident is still logged to the main and panic logs.
5269 74. Buglet left over from Exim 3 conversion. The message "too many messages
5270 in one connection" was written to the rejectlog but not the mainlog, except
5271 when address rewriting (yes!) was being logged.
5273 75. Added write_rejectlog option.
5275 76. When a system filter was run not as root (that is, when system_filter_user
5276 was set), the values of the $n variables were not being returned to the
5277 main process; thus, they were not subsequently available in the $sn
5280 77. Added +return_path_on_delivery log selector.
5282 78. A connection timeout was being treated differently from recipients deferred
5283 when testing hosts_max_try with a message that was older than the host's
5284 retry timeout. (The host should not be counted, thus allowing all hosts to
5285 be tried at least once before bouncing.) This may have been the cause of an
5286 occasionally reported bug whereby a message would remain on the queue
5287 longer than the retry timeout, but would be bounced if a delivery was
5288 forced. I say "may" because I never totally pinned down the problem;
5289 setting up timeout/retry tests is difficult. See also the next item.
5291 79. The ultimate address timeout was not being applied to errors that involved
5292 a combination of host plus message (for example, a timeout on a MAIL
5293 command). When an address resolved to a number of possible hosts, and they
5294 were not all tried for each delivery (e.g. because of hosts_max_try), a
5295 message could remain on the queue longer than the retry timeout.
5297 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5300 81. Fixed an obscure SMTP outgoing bug which required at least the following
5301 conditions: (a) there was another message waiting for the same server;
5302 (b) the server returned 5xx to all RCPT commands in the first message so
5303 that the message was not completed; (c) the server dropped the connection
5304 or gave a negative response to the RSET that Exim sends to abort the
5305 transaction. The observed case was a dropped connection after DATA that had
5306 been sent in pipelining mode. That is, the server had advertised PIPELINING
5307 but was not implementing it correctly. The effect of the bug was incorrect
5308 behaviour, such as trying another host, and this could lead to a crash.
5314 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5315 and daemon.c were passed as pointers to ints; they should have been
5316 pointers to socklen_t variables (which are typically unsigned ints).
5318 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5321 3. Fixed a really odd bug that affected only the testing scheme; patching a
5322 certain fixed string in the binary changed the value of another string that
5323 happened to be identical to the end of the original first string.
5325 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5326 name", it returns that address as the IP address. On some operating
5327 systems (e.g. Solaris), it also passes back the IP address string as the
5328 "host name". However, on others (e.g. Linux), it passes back an empty
5329 string. Exim wasn't checking for this, and was changing the host name to an
5330 empty string, assuming it had been canonicized.
5332 5. Although rare, it is permitted to have more than one PTR record for a given
5333 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5334 all the names associated with an address, because they do in Solaris.
5335 However, it seems that they do not in Linux for data that comes from the
5336 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5337 I found this out when I moved to a new Linux workstation and tried to run
5338 the Exim test suite.
5340 To get round this problem I have changed the code so that it now does its
5341 own call to the DNS to look up PTR records when searching for a host name.
5342 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5343 addresses that are only in /etc/hosts are still found.
5345 This behaviour is, however, controlled by an option called host_lookup_
5346 order, which defaults to "bydns:byaddr". If people want to use the other
5347 order, or indeed, just use one or the other means of lookup, they can
5348 specify it in this variable.
5350 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5351 some operating systems, this comes back from gethostbyaddr() as an empty
5352 string, and this is what Exim used to test for. However, it seems that in
5353 other systems, "." is yielded. Exim now tests for this case too.
5355 7. The values of check_spool_space and check_log_space are now held internally
5356 as a number of kilobytes instead of an absolute number of bytes. If a
5357 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5358 kilobyte. This means that much larger values can be stored.
5360 8. Exim monitor: an attempt to get the action menu when not actually pointing
5361 at a message produces an empty menu entitled "No message selected". This
5362 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5363 no entries in it ("Shell widget menu has zero width and/or height"). So I
5364 have added a single, blank menu entry in this case.
5366 9. Added ${quote_local_part.
5368 10. MIME decoding is now applied to the contents of Subject: header lines when
5371 11. Now that a reference to $sender_host_address automatically causes a reverse
5372 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5373 host lookup before query-style lookups in lists that might use this
5374 variable. This has therefore been abolished, and the "net-" prefix is no
5375 longer necessary for query-style lookups.
5377 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5378 have been a typo for LFLAGS, so it has been changed.
5380 13. The install script calls Exim with "-C /dev/null" in order to find the
5381 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5382 to be output. Howeve, since Exim outputs its version number before the
5383 error, it didn't break the script. It just looked ugly. I fixed this by
5384 always allowing "-C /dev/null" if the caller is root.
5386 14. Ignore overlarge ACL variable number when reading spool file - insurance
5387 against a later release with more variables having written the file.
5389 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5390 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5391 this, as well as the form without the "IPv6" on the front.
5393 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5394 OS/Makefile-Darwin file.
5396 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5397 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5399 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5400 variable. However, some OS use ints for inodes. I've added cast to long int
5401 to get rid of the compiler warning.
5403 19. I had forgotten to lock out "/../" in configuration file names when
5404 ALT_CONFIG_PREFIX was set.
5406 20. Routers used for verification do not need to specify transports. However,
5407 if such a router generated a host list, and callout was configured, Exim
5408 crashed, because it could not find a port number from the (non-existent)
5409 transport. It now assumes port 25 in this circumstance.
5411 21. Added the -t option to exigrep.
5413 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5414 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5415 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5416 them is set, LOOKUP_LSEARCH is forced.
5418 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5420 24. Added sender and host information to the "rejected by local_scan()" log
5421 line; previously there was no indication of these.
5423 25. Added .include_if_exists.
5425 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5426 Linux. It turns out that not all file systems support this. Apparently some
5427 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5428 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5429 not supported on the file descriptor, is now ignored when Exim is trying to
5430 sync a directory. This applies only to Linux.
5432 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5434 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5435 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5438 29. Added support for maildirsize files from supplied patch (modified a bit).
5440 30. The use of :fail: followed by an empty string could lead Exim to respond to
5441 sender verification failures with (e.g.):
5443 550 Verification failed for <xxx>
5444 550 Sender verify failed
5446 where the first response line was missing the '-' that indicates it is not
5447 the final line of the response.
5449 31. The loop for finding the name of the user that called Exim had a hardwired
5450 limit of 10; it now uses the value of finduser_retries, which is used for
5451 all other user lookups.
5453 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5456 33. Exim was neglecting to zero errno before one call of strtol() when
5457 expanding a string and expecting an integer value. On some systems this
5458 resulted in spurious "integer overflow" errors. Also, it was casting the
5459 result into an int without checking.
5461 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5462 did not work. The code looks as if it has *never* worked, though it appears
5463 to have been documented since at least releast 1.62. I have made it work.
5465 35. The "timeout_DNS" error in retry rules, also documented since at least
5466 1.62, also never worked. As it isn't clear exactly what this means, and
5467 clearly it isn't a major issue, I have abolished the feature by treating it
5468 as "timeout", and writing a warning to the main and panic logs.
5470 36. The display of retry rules for -brt wasn't always showing the error code
5473 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5474 timeout_connect_A, timeout_connect_MX.
5476 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5477 to the empty sender.
5479 39. The daemon was not analysing the content of -oX till after it had closed
5480 stderr and disconnected from the controlling terminal. This meant that any
5481 syntax errors were only noted on the panic log, and the return code from
5482 the command was 0. By re-arranging the code a little, I've made the
5483 decoding happen first, so such errors now appear on stderr, and the return
5484 code is 1. However, the actual setting up of the sockets still happens in
5485 the disconnected process, so errors there are still only recorded on the
5488 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5489 connections (as happens on some IP stacks) was logged at start up time as
5490 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5491 it from "IPv6 and IPv4", which means that two separate sockets are being
5494 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5495 says whether AF_INET or AF_INET6 was passed as an argument.
5497 42. Exiwhat output was messed up when time zones were included in log
5500 43. Exiwhat now gives more information about the daemon's listening ports,
5501 and whether -tls-on-connect was used.
5503 44. The "port" option of the smtp transport is now expanded.
5505 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5506 silently ignored. Now an error message is written to the main and panic
5509 46. There's a new ACL modifier called "logwrite" which writes to a log file
5510 as soon as it is encountered.
5512 47. Added $local_user_uid and $local_user_gid at routing time.
5514 48. Exim crashed when trying to verify a sender address that was being
5517 49. Exim was recognizing only a space character after ".include". It now also
5518 recognizes a tab character.
5520 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5521 extracting the relevant information from the specification. The man page no
5522 longer contains scrambled data for the -d option, and I've added a section
5523 at the front about calling Exim under different names.
5525 51. Added "extra_headers" argument to the "mail" command in filter files.
5527 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5530 53. Installed eximstats 1.29.
5532 54. Added transport_filter_timeout as a generic transport option.
5534 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5535 Cc: header lines. This was required by RFC 822, but it not required by RFC
5538 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5539 incoming messages that did not have them. Now it does so only if the
5540 message originates locally, that is, if there is no associated remote host
5541 address. When Resent- header lines are present, this applies to the Resent-
5542 lines rather than the non-Resent- lines.
5544 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5545 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5547 58. Messages for configuration errors now include the name of the main
5548 configuration file - useful now that there may be more than one file in a
5549 list (.included file names were always shown).
5551 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5552 for those rare installations that do not start the daemon as root or run it
5553 setuid root. I've cut out the call to initgroups() if the daemon is not
5556 60. The Exim user and group can now be bound into the binary as text strings
5557 that are looked up at the start of Exim's processing.
5559 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5561 62. Added $mailstore_basename variable.
5563 63. Installed patch to sieve.c from Michael Haardt.
5565 64. When Exim failed to open the panic log after failing to open the main log,
5566 the original message it was trying to log was written to stderr and debug
5567 output, but if they were not available (the usual case in production), it
5568 was lost. Now it is written to syslog before the two lines that record the
5569 failures to open the logs.
5571 65. Users' Exim filters run in subprocesses under the user's uid. It is
5572 possible for a "deliver" command or an alias in a "personal" command to
5573 provoke an address rewrite. If logging of address rewriting is configured,
5574 this fails because the process is not running as root or exim. There may be
5575 a better way of dealing with this, but for the moment (because 4.30 needs
5576 to be released), I have disabled address rewrite logging when running a
5577 filter in a non-root, non-exim process.
5583 1. The buildconfig auxiliary program wasn't quoting the value set for
5584 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5585 not defined. This bug was masked in 4.22 by the effect that was fixed in
5588 2. Some messages that were rejected after a message id was allocated were
5589 shown as "incomplete" by exigrep. It no longer does this for messages that
5590 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5592 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5593 have allow_domain_literals set, the ID did not get logged in the <= line.
5594 Domain literals are now always recognized in Message-ID: header lines.
5596 4. The first argument for a ${extract expansion item is the key name or field
5597 number. Leading and trailing spaces in this item were not being ignored,
5598 causing some misleading effects.
5600 5. When deliver_drop_privilege was set, single queue runner processes started
5601 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5602 same command in the process it spins off) were not dropping privilege.
5604 6. When the daemon running as "exim" started a queue runner, it always
5605 re-executed Exim in the spun-off process. This is a waste of effort when
5606 deliver_drop_privilege is set. The new process now just calls the
5607 queue-runner function directly.
5613 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5616 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5617 ignored. Though the use of -oP was forcing the writing of a pid file, it
5618 was always written to the default place.
5620 3. If the message "no IP address found for host xxxx" is generated during
5621 incoming verification, it is now followed by identification of the incoming
5622 connection (so you can more easily find what provoked it).
5624 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5626 5. Added some features to "harden" Exim a bit more against certain attacks:
5628 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5629 be put in Local/Makefile. This is like the never_users runtime option,
5630 but it cannot be overridden. The default setting is "root".
5632 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5633 prefix string with which any file named in a -C command line option
5636 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5637 is retained for -C and -D only if the caller of Exim is root. Without
5638 it, the exim user may also use -C and -D and retain privilege.
5640 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5641 command line option is disabled.
5643 6. Macro names set by the -D option must start with an upper case letter, just
5644 like macro names defined in the configuration file.
5646 7. Added "dereference=" facility to LDAP.
5648 8. Two instances of the typo "uknown" in the source files are fixed.
5650 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5651 the Configure-Makefile script screwed up while processing it.
5653 10. Incorporated PCRE 4.4.
5655 11. The SMTP synchronization check was not operating right at the start of an
5656 SMTP session. For example, it could not catch a HELO sent before the client
5657 waited for the greeting. There is now a check for outstanding input at the
5658 point when the greeting is written. Because of the duplex, asynchronous
5659 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5660 way, but not yet received, when the check is performed.
5662 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5663 on TCP/IP sockets, because this apparently causes some broken clients to
5666 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5667 unchanged) from the Cygwin maintainer.
5669 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5670 was supported instead of testing for mbx. Effectively a typo.
5672 15. The spa authenticator server code was not checking that the input it
5673 received was valid base64.
5675 16. The debug output line for the "set" modifier in ACLs was not showing the
5676 name of the variable that was being set.
5678 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5680 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5681 lookup on its own. Something else was needed to trigger the lookup. For
5682 example, a match in host_lookup or the need for a host name in a host list.
5683 Now, if $sender_host_name is referenced and the host name has not yet been
5684 looked up, a lookup is performed. If the lookup fails, the variable remains
5685 empty, and $host_lookup_failed is set to "1".
5687 19. Added "eqi" as a case-independent comparison operator.
5689 20. The saslauthd authentication condition could segfault if neither service
5690 nor realm was specified.
5692 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5693 error message that was logged was misleading, and incoming SMTP
5694 connections were dropped. The message is now more accurate, and temporary
5695 errors are given to SMTP connections.
5697 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5698 (except RSET) with a 421 error, until QUIT is received. However, it was
5699 failing to send a response to QUIT.
5701 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5702 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5703 ACL is now run after the helo_try_verify_hosts code.
5705 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5706 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5707 case-independent, so other case variants are also recognized). Apparently
5708 some systems use these upper case variants.
5710 25. If more than two messages were waiting for the same host, and a transport
5711 filter was specified for the transport, Exim sent two messages over the
5712 same TCP/IP connection, and then failed with "socket operation on non-
5713 socket" when it tried to send the third.
5715 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5717 27. The extern definition of crypt16() in expand.c was not being excluded when
5718 the OS had its own crypt16() function.
5720 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5721 as a preferred synonym for return_size_limit, both as an option and as an
5724 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5726 30. Changed the default configuration ACL to relax the local part checking rule
5727 for addresses that are not in any local domains. For these addresses,
5728 slashes and pipe symbols are allowed within local parts, but the sequence
5729 /../ is explicitly forbidden.
5731 31. SPA server authentication was not clearing the challenge buffer before
5734 32. log_message in a "warn" ACL statement was writing to the reject log as
5735 well as to the main log, which contradicts the documentation and doesn't
5736 seem right (because no rejection is happening). So I have stopped it.
5738 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5739 However, I am unable to do any testing of this.
5741 34. Fixed an infelicity in the appendfile transport. When checking directories
5742 for a mailbox, to see if any needed to be created, it was accidentally
5743 using path names with one or more superfluous leading slashes; tracing
5744 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5746 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5747 contents are added to the log line that is written for every discarded
5748 recipient. (Previously a log_message setting was ignored.)
5750 36. The ${quote: operator now quotes the string if it is empty.
5752 37. The install script runs exim in order to find its version number. If for
5753 some reason other than non-existence or emptiness, which it checks, it
5754 could not run './exim', it was installing it with an empty version number,
5755 i.e. as "exim-". This error state is now caught, and the installation is
5758 38. An argument was missing from the function that creates an error message
5759 when Exim fails to connect to the socket for saslauthd authentication.
5760 This could cause Exim to crash, or give a corrupted message.
5762 39. Added isip, isip4, and isip6 to ${if conditions.
5764 40. The ACL variables $acl_xx are now saved with the message, and can be
5765 accessed later in routers, transports, and filters.
5767 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5768 strings in the file are not string-expanded.
5770 42. If a MAIL command specified a SIZE value that was too large to fit into an
5771 int variable, the check against message_size_limit failed. Such values are
5772 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5773 day this will have to be increased, but I don't think I want to be around
5774 when emails are that large.
5781 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5782 iconv() is not standard in FreeBSD.
5784 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5785 IPv6 enabled. The observed symptom was a segmentation fault on return from
5786 the function os_common_find_running_interfaces() in src/os.c.
5788 3. In the check_special_case() function in daemon.c I had used "errno" as an
5789 argument name, which causes warnings on some systems. This was basically a
5790 typo, since it was named "eno" in the comments!
5792 4. The code that waits for the clock to tick (at a resolution of some fraction
5793 of a second) so as to ensure message-id uniqueness was always waiting for
5794 at least one whole tick, when it could have waited for less. [This is
5795 almost certainly not relevant at current processor speeds, where it is
5796 unlikely to ever wait at all. But we try to future-proof.]
5798 5. The function that sleeps for a time interval that includes fractions of a
5799 second contained a race. It did not block SIGALRM between setting the
5800 timer, and suspending (a couple of lines later). If the interval was short
5801 and the sigsuspend() was delayed until after it had expired, the suspension
5802 never ended. On busy systems this could lead to processes getting stuck for
5805 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5806 process, before it forks any delivery processes. The open lookup caching
5807 mechanism meant that the open file or database connection was passed into
5808 the delivery process. The problem was that delivery processes always tidy
5809 up cached lookup data. This could cause a problem for the next delivery
5810 process started by the queue runner, because the external queue runner
5811 process does not know about the closure. So the next delivery process
5812 still has data in the lookup cache. In the case of a file lookup, there was
5813 no problem because closing a file descriptor in a subprocess doesn't affect
5814 the parent. However, if the lookup was caching a connection to a database,
5815 the connection was closed, and the second delivery process was likely to
5816 see errors such as "PGSQL: query failed: server closed the connection
5817 unexpectedly". The problem has been fixed by closing all cached lookups
5818 in a queue runner before running a delivery process.
5820 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5821 seem to have the "const" qualifier which it has on other OS. I've
5824 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5825 *of the same type* (client or server) with the same public name that an
5826 error should be diagnosed.
5828 9. When Exim looked up a host name for an IP address, but failed to find the
5829 original IP address when looking up the host name (a safety check), it
5830 output the message "<ip address> does not match any IP for NULL", which was
5831 confusing, to say the least. The bug was that the host name should have
5832 appeared instead of "NULL".
5834 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5835 user that is built into the binary, and the -C or -D options is used, root
5836 privilege is dropped before the configuration file is read. In addition,
5837 logging is switched to stderr instead of the normal log files. If the
5838 configuration then re-defines the Exim user, the unprivileged environment
5839 is probably not what is expected, so Exim logs a panic warning message (but
5842 However, if deliver_drop_privilege is set, the unprivileged state may well
5843 be exactly what is intended, so the warning has been cut out in that case,
5844 and Exim is allowed to try to write to its normal log files.
5850 1. smtp_return_error_details was not giving details for temporary sender
5851 or receiver verification errors.
5853 2. Diagnose a configuration error if two authenticators have the same public
5856 3. Exim used not to create the message log file for a message until the first
5857 delivery attempt. This could be confusing when incoming messages were held
5858 for policy or load reasons. The message log file is now created at the time
5859 the message is received, and an initial "Received" line is written to it.
5861 4. The automatically generated man page for command line options had a minor
5862 bug that caused no ill effects; however, a more serious problem was that
5863 the procedure for building the man page automatically didn't always
5864 operate. Consequently, release 4.20 contains an out-of-date version. This
5865 shouldn't happen again.
5867 5. When building Exim with embedded Perl support, the script that builds the
5868 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5869 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5871 6. The freeze_tell option was not being used for messages that were frozen on
5872 arrival, either by an ACL or by local_scan().
5874 7. Added the smtp_incomplete_transaction log selector.
5876 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5877 was accepting AUTH without a new EHLO.
5879 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5880 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5883 10. Logging of TCP/IP connections (when configured) now happens in the main
5884 daemon process instead of the child process, so that the TCP/IP connection
5885 count is more accurate (but it can never be perfect).
5887 11. The use of "drop" in a nested ACL was not being handled correctly in the
5888 outer ACL. Now, if condition failure induced by the nested "drop" causes
5889 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5890 or "require"), the connection is dropped.
5892 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5893 that yield "discard" can now be used with an "accept" or a "discard" verb,
5894 but an error is generated for any others (because I can't see a useful way
5895 to define what should happen).
5897 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5898 are now processed in the same way as lines in the Exim configuration file.
5899 In particular, continuation lines are supported.
5901 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5903 15. Added -ti meaning -t -i.
5905 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5906 domains, and warn by logging if others are found.
5908 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5909 returned from GIFCONF when Exim is trying to find the list of interfaces on
5910 a host. The code in os.c has been modified to copy each ifreq to an aligned
5911 structure in all cases.
5913 Also, in some cases, the returned ifreq's were being copied to a 'struct
5914 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5915 means the last couple of bytes of an IPv6 address could be chopped if the
5916 ifreq contained only a normal sockaddr (14 bytes storage).
5918 18. Named domain lists were not supported in the hosts_treat_as_local option.
5919 An entry such as +xxxx was not recognized, and was treated as a literal
5922 19. Ensure that header lines added by a DATA ACL are included in the reject log
5923 if the ACL subsequently rejects the message.
5925 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5926 MD5 (which is deprecated).
5928 21. When testing a filter file using -bf, Exim was writing a message when it
5929 took the sender from a "From " line in the message, but it was not doing so
5930 when it took $return_path from a Return-Path: header line. It now does.
5932 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5933 with a valid header line field name (a series of printing characters
5934 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5936 23. Changed "disc" in the source to "disk" to conform to the documentation and
5937 the book and for uniformity.
5939 24. Ignore Sendmail's -Ooption=value command line item.
5941 25. When execve() failed while trying to run a command in a pipe transport,
5942 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5943 could be confused with a return value of 69 from the command itself. This
5944 has been changed to 127, the value the shell returns if it is asked to run
5945 a non-existent command. The wording for the related log line suggests a
5946 non-existent command as the problem.
5948 26. If received_header_text expands to an empty string, do not add a Received:
5949 header line to the message. (Well, it adds a token one on the spool, but
5950 marks it "old" so that it doesn't get used or transmitted.)
5952 27. Installed eximstats 1.28 (addition of -nt option).
5954 28. There was no check for failure on the call to getsockname() in the daemon
5955 code. This can fail if there is a shortage of resources on the system, with
5956 ENOMEM, for example. A temporary error is now given on failure.
5958 29. Contrary to the C standard, it seems that in some environments, the
5959 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5960 program. Exim now does this explicitly; it affects the formatting of
5961 timestamps using strftime().
5963 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5964 complaints. I've now initialized all the variables, to avoid this.
5966 32. Header lines added by a system filter were not being "seen" during
5967 transport-time rewrites.
5969 33. The info_callback() function passed to OpenSSL is set up with type void
5970 (*)(SSL *, int, int), as described somewhere. However, when calling the
5971 function (actually a macro) that sets it up, the type void(*)() is
5972 expected. I've put in a cast to prevent warnings from picky compilers.
5974 34. If a DNS black list lookup found a CNAME record, but there were no A
5975 records associated with the domain it pointed at, Exim crashed.
5977 35. If a DNS black list lookup returned more than one A record, Exim ignored
5978 all but the first. It now scans all returned addresses if a particular IP
5979 value is being sought. In this situation, the contents of the
5980 $dnslist_value variable are a list of all the addresses, separated by a
5983 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5984 to accept a host name and all its aliases if the forward lookup for any of
5985 them yielded the IP address of the incoming connection. Now it accepts only
5986 those names whose forward lookup yields the correct IP address. Any other
5987 names are discarded. This closes a loophole whereby a rogue DNS
5988 administrator could create reverse DNS records to break through a
5989 wildcarded host restriction in an ACL.
5991 37. If a user filter or a system filter that ran in a subprocess used any of
5992 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5993 the wrong values were passed to the pipe command ($thisaddress had the
5994 value of $0, $0 had the value of $1, etc). This bug was introduced by
5995 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5997 38. Improved the line breaking for long SMTP error messages from ACLs.
5998 Previously, if there was no break point between 40 and 75 characters, Exim
5999 left the rest of the message alone. Two changes have been made: (a) I've
6000 reduced the minimum length to 35 characters; (b) if it can't find a break
6001 point between 35 and 75 characters, it looks ahead and uses the first one
6002 that it finds. This may give the occasional overlong line, but at least the
6003 remaining text gets split now.
6005 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6006 file descriptors might be low, and that setting 1000 would always raise it.
6007 It turns out that in some environments, the limit is already over 1000 and
6008 that lowering it causes trouble. So now Exim takes care not to decrease it.
6010 40. When delivering a message, the value of $return_path is set to $sender_
6011 address at the start of routing (routers may change the value). By an
6012 oversight, this default was not being set up when an address was tested by
6013 -bt or -bv, which affected the outcome if any router or filter referred to
6016 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6017 log only when "message" or "log_message" are set. However, if one of the
6018 conditions was an address verification, or a call to a nested ACL, the
6019 messages generated by the underlying test were being passed through. This
6020 no longer happens. The underlying message is available in $acl_verify_
6021 message for both "message" and "log_message" expansions, so it can be
6022 passed through if needed.
6024 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6025 new expansion $bh_ to give the encoded byte string without charset
6026 translation. Translation happens only if iconv() is available; HAVE_ICONV
6027 indicates this at build time. HEADERS_CHARSET gives the charset to
6028 translate to; headers_charset can change it in the configuration, and
6029 "headers charset" can change it in an individual filter file.
6031 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6032 that creates RFC 2047 encoded "words" labels them as that charset instead
6033 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6034 expansion operator; (ii) when Exim creates a From: line for a local
6035 message; (iii) when a header line is rewritten to include a "phrase" part.
6037 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6038 buggy, causing it to skip the first lines of messages whose message ID
6039 ended in 'D'. This would not have bitten before Exim release 4.14, because
6040 message IDs were unlikely to end in 'D' before then. The effect was to have
6041 incorrect size information for certain domains.
6043 45. #include "config.h" was missing at the start of the crypt16.c module. This
6044 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6047 46. If there was a timeout during a "random" callout check, Exim treated it as
6048 a failure of the random address, and carried on sending RSET and the real
6049 address. If the delay was just some slowness somewhere, the response to the
6050 original RCPT would be taken as a response to RSET and so on, causing
6051 mayhem of various kinds.
6053 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6054 when I implemented it. It didn't allow for the fact that some option values
6055 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6056 the right test for positive values.
6058 48. Domain names in DNS records are case-independent. Exim always looks them up
6059 in lower case. Some resolvers return domain names in exactly the case they
6060 appear in the zone file, that is, they may contain uppercase letters. Not
6061 all resolvers do this - some return always lower case. Exim was treating a
6062 change of case by a resolver as a change of domain, similar to a widening
6063 of a domain abbreviation. This triggered its re-routing code and so it was
6064 trying to route what was effectively the same domain again. This normally
6065 caused routing to fail (because the router wouldn't handle the domain
6066 twice). Now Exim checks for this case specially, and just changes the
6067 casing of the domain that it ultimately uses when it transmits the message
6070 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6073 50. If a filter generated a file delivery with a non-absolute name (possible if
6074 no home directory exists for the router), the forbid_file option was not
6077 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6078 the existing equality matching.
6080 52. Exim was using ints instead of ino_t variables in some places where it was
6081 dealing with inode numbers.
6083 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6084 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6085 called TMPDIR, and if it finds it is different, it changes its value.
6087 54. The smtp_printf() function is now made available to local_scan() so
6088 additional output lines can be written before returning. There is also an
6089 smtp_fflush() function to enable the detection of a dropped connection.
6090 The variables smtp_input and smtp_batched_input are exported to
6093 55. Changed the default runtime configuration: the message "Unknown user"
6094 has been removed from the ACL, and instead placed on the localuser router,
6095 using the cannot_route_message feature. This means that any verification
6096 failures that generate their own messages won't get overridden. Similarly,
6097 the "Unrouteable address" message that was in the ACL for unverifiable
6098 relay addresses has also been removed.
6100 56. Added hosts_avoid_esmtp to the smtp transport.
6102 57. The exicyclog script was not checking for the esoteric option
6103 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6104 will work only if exicyclog is run under the appropriate euid.
6106 58. Following a discussion on the list, the rules by which Exim recognises line
6107 endings on incoming messages have been changed. The -dropcr and drop_cr
6108 options are now no-ops, retained only for backwards compatibility. The
6109 following line terminators are recognized: LF CRLF CR. However, special
6110 processing applies to CR:
6112 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6113 nor a local message in the state where . is a terminator.
6115 (ii) If a bare CR is encountered in a header line, an extra space is added
6116 after the line terminator so as not to end the header. The reasoning
6117 behind this is that bare CRs in header lines are most likely either
6118 to be mistakes, or people trying to play silly games.
6120 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6121 was being incorrectly given as 18 bytes larger than it should have been.
6122 This is a VOB (very old bug).
6124 60. This may never have affected anything current, but just in case it has:
6125 When the local host is found other than at the start of a list of hosts,
6126 the local host, those with the same MX, and any that follow, are discarded.
6127 When the list in question was part of a longer list of hosts, the following
6128 hosts (not currently being processed) were also being discarded. This no
6129 longer happens. I'm not sure if this situation could ever has previously
6132 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6135 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6136 header. This is something that is recommended in a new Internet Draft, and
6137 is something that is documented as being done by Sendmail. There are two
6138 possible values. For messages generated by the autoreply transport, Exim
6141 Auto-Submitted: auto-replied
6143 whereas for all other generated messages (e.g. bounces) it adds
6145 Auto-Submitted: auto-generated
6147 63. The "personal" condition in filters now includes a test for the
6148 Auto-Submitted: header. If it contains the string "auto-" the message it
6149 not considered personal.
6151 64. Added rcpt_include_affixes as a generic transport option.
6153 65. Added queue_only_override (default true).
6155 66. Added the syslog_duplication option.
6157 67. If what should have been the first header line of a message consisted of
6158 a space followed by a colon, Exim was mis-interpreting it as a header line.
6159 It isn't of course - it is syntactically invalid and should therefore be
6160 treated as the start of the message body. The misbehaviour could have
6161 caused a number of strange effects, including loss of data in subsequent
6162 header lines, and spool format errors.
6164 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6165 client host had authenticated. This control can now be exercised by an ACL
6166 for more flexibility.
6168 69. By default, callouts do not happen when testing with -bh. There is now a
6169 variant, -bhc, which does actually run the callout code, including
6170 consulting and updating the callout cache.
6172 70. Added support for saslauthd authentication, courtesy of Alexander
6175 71. If statvfs() failed on the spool or log directories while checking their
6176 size for availability, Exim confusingly gave the error "space shortage".
6177 Furthermore, in debugging mode it crashed with a floating point exception.
6178 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6179 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6180 serious problem, Exim now writes to the main and panic logs when this
6181 happens, with details of the failure. It then refuses to accept the
6182 incoming message, giving the message "spool directory problem" or "log
6183 directory problem" with a 421 code for SMTP messages.
6185 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6186 0, 1, and 2 exist, because some OS complain for execs without them (see
6187 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6188 descriptors. However, the code omitted to check that the open succeeded,
6189 causing mysterious errors if for some reason the permissions on /dev/null
6190 got screwed. Now Exim writes a message to the main and panic logs, and
6191 bombs out if it can't open /dev/null.
6193 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6194 interact so that it is all more flexible. It is supposed to remain
6195 backwards compatible. Also added extra_local_interfaces.
6197 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6198 to bomb out with an assertion failure - to the client this appears as a
6199 connection drop. This problem occurs in the part of the code that was taken
6200 from the Samba project. Fortunately, the assertion is in a very simple
6201 function, so I have fixed this by reproducing the function inline in the
6202 one place where it is called, and arranging for authentication to fail
6203 instead of killing the process with assert().
6205 75. The SPA client code was not working when the server requested OEM rather
6206 than Unicode encoding.
6208 76. Added code to make require_files with a specific uid setting more usable in
6209 the case where statting the file as root fails - usually a non-root-mounted
6210 NFS file system. When this happens and the failure is EACCES, Exim now
6211 forks a subprocess and does the per-uid checking as the relevant uid.
6213 77. Added process_log_path.
6215 78. If log_file_path was not explicitly set, a setting of check_log_space or
6216 check_log_inodes was ignored.
6218 79. If a space check for the spool or log partitions fails, the incident is now
6219 logged. Of course, in the latter case the data may get lost...
6221 80. Added the %p formatting code to string_format() so that it can be used to
6222 print addresses in debug_print(). Adjusted all the address printing in the
6223 debugging in store.c to use %p rather than %d.
6225 81. There was a concern that a line of code in smtp_in.c could overflow a
6226 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6227 initially expressed, the concern was not well-founded, because trailing
6228 spaces are removed early. However, if the trailing spaces were followed by
6229 a NULL, they did not get removed, so the overflow was possible. Two fixes
6232 (a) I re-wrote the offending code in a cleaner fashion.
6233 (b) If an incoming SMTP command contains a NULL character, it is rejected
6236 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6237 runs initgroups(), so that if the Exim user is in any additional groups,
6238 they will be used during message reception.
6244 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob