1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
8 and 3.4.4 - once the server is enabled to respond to an OCSP request
9 it does even when not requested, resulting in a stapling non-aware
10 client dropping the TLS connection.
12 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
13 support variable-length bit vectors. No functional change.
18 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
21 JH/02 The smtp transport option "multi_domain" is now expanded.
23 JH/03 The smtp transport now requests PRDR by default, if the server offers
26 JH/04 Certificate name checking on server certificates, when exim is a client,
27 is now done by default. The transport option tls_verify_cert_hostnames
28 can be used to disable this per-host. The build option
29 EXPERIMENTAL_CERTNAMES is withdrawn.
31 JH/05 The value of the tls_verify_certificates smtp transport and main options
32 default to the word "system" to access the system default CA bundle.
33 For GnuTLS, only version 3.0.20 or later.
35 JH/06 Verification of the server certificate for a TLS connection is now tried
36 (but not required) by default. The verification status is now logged by
37 default, for both outbound TLS and client-certificate supplying inbound
40 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
43 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
44 Status Notification (bounce) messages are now MIME format per RFC 3464.
45 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
46 under the control of the dsn_advertise_hosts option, and routers may
47 have a dsn_lasthop option.
49 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
50 default, modifiable by a malware= option. The list separator for
51 the options can now be changed in the usual way. Bug 68.
53 JH/10 The smtp_receive_timeout main option is now expanded before use.
55 JH/11 The incoming_interface log option now also enables logging of the
56 local interface on delivery outgoing connections.
58 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
59 if the interface and destination host and port all match.
61 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
64 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
65 Patch from Andrew Lewis.
67 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
68 now supports optional time-restrictions, weighting, and priority
69 modifiers per server. Patch originally by <rommer@active.by>.
71 JH/16 The spamd_address main option now supports a mixed list of local
72 and remote servers. Remote servers can be IPv6 addresses, and
75 JH/17 Bug 68: The spamd_address main option now supports an optional
76 timeout value per server.
78 JH/18 Bug 1581: Router and transport options headers_add/remove can
79 now have the list separator specified.
81 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
84 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
87 JH/21 Support for the A6 type of dns record is withdrawn.
89 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
90 rather than the verbs used.
92 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
93 from 255 to 1024 chars.
95 JH/24 Verification callouts now attempt to use TLS by default.
97 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
98 are generic router options now. The defaults didn't change.
100 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
101 Original patch from Alexander Shikoff, worked over by JH.
103 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
104 files if ZCAT_COMMAND is not executable.
106 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
108 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
110 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
111 Normally benign, it bites when the pair was led to by a CNAME;
112 modern usage is to not canoicalize the domain to a CNAME target
113 (and we were inconsistent anyway for A-only vs AAAA+A).
115 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
117 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
118 when evaluating $sender_host_dnssec.
120 JH/31 Check the HELO verification lookup for DNSSEC, adding new
121 $sender_helo_dnssec variable.
123 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
125 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
127 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
129 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
130 documented as working, but never had. Support all but $spam_report.
132 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
133 added for tls authenticator.
138 TL/01 When running the test suite, the README says that variables such as
139 no_msglog_check are global and can be placed anywhere in a specific
140 test's script, however it was observed that placement needed to be near
141 the beginning for it to behave that way. Changed the runtest perl
142 script to read through the entire script once to detect and set these
143 variables, reset to the beginning of the script, and then run through
144 the script parsing/test process like normal.
146 TL/02 The BSD's have an arc4random API. One of the functions to induce
147 adding randomness was arc4random_stir(), but it has been removed in
148 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
149 function when detected.
151 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
152 cause callback expansion.
154 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
155 syntax errors in an expansion can be treated as a string instead of
156 logging or causing an error, due to the internal use of bool_lax
157 instead of bool when processing it.
159 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
160 server certificates when making smtp deliveries.
162 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
164 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
166 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
168 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
169 Merged patch from Sebastian Wiedenroth.
171 JH/05 Fix results-pipe from transport process. Several recipients, combined
172 with certificate use, exposed issues where response data items split
173 over buffer boundaries were not parsed properly. This eventually
174 resulted in duplicates being sent. This issue only became common enough
175 to notice due to the introduction of conection certificate information,
176 the item size being so much larger. Found and fixed by Wolfgang Breyha.
178 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
179 size buffer was used, resulting in syntax errors when an expansion
182 JH/07 Add support for directories of certificates when compiled with a GnuTLS
183 version 3.3.6 or later.
185 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
186 is EXPERIMENTAL_EVENT, the main-configuration and transport options
187 both become "event_action", the variables become $event_name, $event_data
188 and $event_defer_errno. There is a new variable $verify_mode, usable in
189 routers, transports and related events. The tls:cert event is now also
190 raised for inbound connections, if the main configuration event_action
193 TL/06 In test suite, disable OCSP for old versions of openssl which contained
194 early OCSP support, but no stapling (appears to be less than 1.0.0).
196 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
197 server certificate names available under the smtp transport option
198 "tls_verify_cert_hostname" now do not permit multi-component wildcard
201 JH/10 Time-related extraction expansions from certificates now use the main
202 option "timezone" setting for output formatting, and are consistent
203 between OpenSSL and GnuTLS compilations. Bug 1541.
205 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
206 encoded parameter in the incoming message. Bug 1558.
208 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
209 include certificate info, eximon was claiming there were spoolfile
212 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
214 JH/14 Log delivery-related information more consistently, using the sequence
215 "H=<name> [<ip>]" wherever possible.
217 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
218 are problematic for Debian distribution, omit them from the release
221 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
223 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
226 JH/17 Fix a null-indirection in certextract expansions when a nondefault
227 output list separator was used.
232 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
233 checkers that were complaining about end of non-void function with no
236 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
237 This was a regression intruduced in 4.83 by another bugfix.
239 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
241 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
242 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
248 TF/01 Correctly close the server side of TLS when forking for delivery.
250 When a message was received over SMTP with TLS, Exim failed to clear up
251 the incoming connection properly after forking off the child process to
252 deliver the message. In some situations the subsequent outgoing
253 delivery connection happened to have the same fd number as the incoming
254 connection previously had. Exim would try to use TLS and fail, logging
255 a "Bad file descriptor" error.
257 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
258 utilities have not been installed.
260 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
261 temporary space as the ACL may create new global variables.
263 TL/01 LDAP support uses per connection or global context settings, depending
264 upon the detected version of the libraries at build time.
266 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
267 to extract and use the src ip:port in logging and expansions as if it
268 were a direct connection from the outside internet. PPv2 support was
269 updated based on HAProxy spec change in May 2014.
271 JH/02 Add ${listextract {number}{list}{success}{fail}}.
273 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
274 Properly escape header and check for NULL return.
276 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
279 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
281 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
282 characters in header names, implemented as a verify condition.
283 Contributed by Michael Fischer v. Mollard.
285 TL/05 Rename SPF condition results err_perm and err_temp to standardized
286 results permerror and temperror. Previous values are deprecated but
287 still accepted. In a future release, err_perm and err_temp will be
288 completely removed, which will be a backward incompatibility if the
289 ACL tests for either of these two old results. Patch contributed by
290 user bes-internal on the mailing list.
292 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
294 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
295 selectors, in both main and reject logs.
297 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
300 JH/07 Add malware type "sock" for talking to simple daemon.
302 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
304 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
305 routers/transports under cutthrough routing.
307 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
308 numbers. Touch up "bool" conditional to keep the same definition.
310 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
312 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
314 JH/12 Expand items in router/transport headers_add or headers_remove lists
315 individually rather than the list as a whole. Bug 1452.
317 Required for reasonable handling of multiple headers_ options when
318 they may be empty; requires that headers_remove items with embedded
319 colons must have them doubled (or the list-separator changed).
321 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
322 view the policy declared in the DMARC record. Currently, $dmarc_status
323 is a combined value of both the record presence and the result of the
326 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
328 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
329 dnslookup router and the smtp transport (applying to the forward
332 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
333 of ldap servers used for a specific lookup. Patch provided by Heiko
336 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
337 New variable $lookup_dnssec_authenticated for observability.
339 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
340 Patch submitted by Lars Timman.
342 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
344 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
345 Requires trusted mode and valid format message id, aborts otherwise.
346 Patch contributed by Heiko Schlichting.
348 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
349 certextract with support for various fields. Bug 1358.
351 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
352 is requested by default, modifiable by smtp transport option
355 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
356 operate on certificate variables to give certificate fingerprints
357 Also new ${sha256:cert_variable}.
359 JH/23 The PRDR feature is moved from being Experimental into the mainline.
361 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
362 Christian Aistleitner.
364 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
366 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
367 file. Patch from Wolfgang Breyha.
369 JH/25 Expand the coverage of the delivery $host and $host_address to
370 client authenticators run in verify callout. Bug 1476.
372 JH/26 Port service names are now accepted for tls_on_connect_ports, to
373 align with daemon_smtp_ports. Bug 72.
375 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
376 support and error reporting did not work properly.
378 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
379 and is readable. Patch from Andrew Colin Kissa.
381 TL/14 Enhance documentation of ${run expansion and how it parses the
382 commandline after expansion, particularly in the case when an
383 unquoted variable expansion results in an empty value.
385 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
387 PP/02 Fix internal collision of T_APL on systems which support RFC3123
388 by renaming away from it. Addresses GH issue 15, reported by
391 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
393 TL/15 SECURITY: prevent double expansion in math comparison functions
394 (can expand unsanitized data). Not remotely exploitable.
401 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
403 PP/02 Make -n do something, by making it not do something.
404 When combined with -bP, the name of an option is not output.
406 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
409 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
410 $sender_host_name and config options to manage this, and basic check
413 PP/05 DSCP support for outbound connections and control modifier for inbound.
415 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
416 (Only plugin which currently uses this is kerberos4, which nobody should
417 be using, but we should make it available and other future plugins might
418 conceivably use it, even though it would break NAT; stuff *should* be
419 using channel bindings instead).
421 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
422 name; added for Sendmail compatibility; requires admin caller.
423 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
424 just ignore it); requires trusted caller.
425 Also parse but ignore: -Ac -Am -X<logfile>
428 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
430 TL/02 Add +smtp_confirmation as a default logging option.
432 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
433 Patch by Magnus Holmgren from 2007-02-20.
435 TL/04 Bugzilla 1281 - Spec typo.
436 Bugzilla 1283 - Spec typo.
437 Bugzilla 1290 - Spec grammar fixes.
439 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
441 TL/06 Add Experimental DMARC support using libopendmarc libraries.
443 TL/07 Fix an out of order global option causing a segfault. Reported to dev
444 mailing list by by Dmitry Isaikin.
446 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
448 JH/02 Support "G" suffix to numbers in ${if comparisons.
450 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
452 NM/01 Bugzilla 1197 - Spec typo
453 Bugzilla 1196 - Spec examples corrections
455 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
457 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
458 gnutls_enable_pkcs11, but renamed to more accurately indicate its
461 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
462 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
464 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
465 "acl {{name}{arg}...}", and optional args on acl condition
468 JH/05 Permit multiple router/transport headers_add/remove lines.
470 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
472 JH/07 Avoid using a waiting database for a single-message-only transport.
473 Performance patch from Paul Fisher. Bugzilla 1262.
475 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
478 JH/09 Add $headers_added variable, with content from use of ACL modifier
479 add_header (but not yet added to the message). Bugzilla 199.
481 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
482 Pulled from Bugzilla 817 by Wolfgang Breyha.
484 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
486 (nb: this is the same fix as in Exim 4.80.1)
488 JH/11 Add A= logging on delivery lines, and a client_set_id option on
491 JH/12 Add optional authenticated_sender logging to A= and a log_selector
494 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
496 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
497 advertise SMTP AUTH mechanism to us, instead of a generic
498 protocol violation error. Also, make Exim more robust to bad
499 data from the Dovecot auth socket.
501 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
503 When a queue runner is handling a message, Exim first routes the
504 recipient addresses, during which it prunes them based on the retry
505 hints database. After that it attempts to deliver the message to
506 any remaining recipients. It then updates the hints database using
509 So if a recipient address works intermittently, it can get repeatedly
510 deferred at routing time. The retry hints record remains fresh so the
511 address never reaches the final cutoff time.
513 This is a fairly common occurrence when a user is bumping up against
514 their storage quota. Exim had some logic in its local delivery code
515 to deal with this. However it did not apply to per-recipient defers
516 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
518 This change adds a proper retry rule check during routing so that the
519 final cutoff time is checked against the message's age. We only do
520 this check if there is an address retry record and there is not a
521 domain retry record; this implies that previous attempts to handle
522 the address had the retry_use_local_parts option turned on. We use
523 this as an approximation for the destination being like a local
524 delivery, as in LMTP.
526 I suspect this new check makes the old local delivery cutoff check
527 redundant, but I have not verified this so I left the code in place.
529 TF/02 Correct gecos expansion when From: is a prefix of the username.
531 Test 0254 submits a message to Exim with the header
535 When I ran the test suite under the user fanf2, Exim expanded
536 the header to contain my full name, whereas it should have added
537 a Resent-Sender: header. It erroneously treats any prefix of the
538 username as equal to the username.
540 This change corrects that bug.
542 GF/01 DCC debug and logging tidyup
543 Error conditions log to paniclog rather than rejectlog.
544 Debug lines prefixed by "DCC: " to remove any ambiguity.
546 TF/03 Avoid unnecessary rebuilds of lookup-related code.
548 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
549 Bug spotted by Jeremy Harris; was flawed since initial commit.
550 Would have resulted in OCSP responses post-SNI triggering an Exim
551 NULL dereference and crash.
553 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
555 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
556 Bug detection, analysis and fix by Samuel Thibault.
557 Bugzilla 1331, Debian bug #698092.
559 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
561 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
562 Server implementation by Todd Lyons, client by JH.
563 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
564 config variable "prdr_enable" controls whether the server
565 advertises the facility. If the client requests PRDR a new
566 acl_data_smtp_prdr ACL is called once for each recipient, after
567 the body content is received and before the acl_smtp_data ACL.
568 The client is controlled by bolth of: a hosts_try_prdr option
569 on the smtp transport, and the server advertisement.
570 Default client logging of deliveries and rejections involving
571 PRDR are flagged with the string "PRDR".
573 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
574 fclose(). Diagnosis by Todd Lyons.
576 PP/17 Update configure.default to handle IPv6 localhost better.
577 Patch by Alain Williams (plus minor tweaks).
580 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
581 This is now consistent with GnuTLS, and is now documented: the
582 previous undocumented portable approach to treating the option as
583 unset was to force an expansion failure. That still works, and
584 an empty string is now equivalent.
586 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
587 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
588 not performing validation itself.
590 PP/20 Added force_command boolean option to pipe transport.
591 Patch from Nick Koston, of cPanel Inc.
593 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
596 TF/04 Added udpsend ACL modifer and hexquote expansion operator
598 PP/21 Fix eximon continuous updating with timestamped log-files.
599 Broken in a format-string cleanup in 4.80, missed when I repaired the
600 other false fix of the same issue.
601 Report and fix from Heiko Schlichting.
604 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
605 Report from Prashanth Katuri.
607 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
608 It's SecureTransport, so affects any MacOS clients which use the
609 system-integrated TLS libraries, including email clients.
611 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
612 using a MIME ACL for non-SMTP local injection.
613 Report and assistance in diagnosis by Warren Baker.
615 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
617 JH/16 Fix comparisons for 64b. Bugzilla 1385.
619 TL/09 Add expansion variable $authenticated_fail_id to keep track of
620 last id that failed so it may be referenced in subsequent ACL's.
622 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
625 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
626 ldap library initialization, allowing self-signed CA's to be
627 used. Also properly sets require_cert option later in code by
628 using NULL (global ldap config) instead of ldap handle (per
629 session). Bug diagnosis and testing by alxgomz.
631 TL/12 Enhanced documentation in the ratelimit.pl script provided in
632 the src/util/ subdirectory.
634 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
635 renamed to Transport Post Delivery Action by Jeremy Harris, as
638 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
639 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
640 redis_servers = needs to be configured which will be used by the redis
641 lookup. Patch from Warren Baker, of The Packet Hub.
643 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
645 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
646 hostname or reverse DNS when processing a host list. Used suggestions
647 from multiple comments on this bug.
649 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
651 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
652 Missed a few lines, added it to make the runtest require no keyboard
655 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
656 contains upper case chars. Make router use caseful_local_part.
658 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
659 support when GnuTLS has been built with p11-kit.
665 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
667 This, or similar/improved, will also be change PP/11 of 4.82.
673 PP/01 Handle short writes when writing local log-files.
674 In practice, only affects FreeBSD (8 onwards).
675 Bugzilla 1053, with thanks to Dmitry Isaikin.
677 NM/01 Bugzilla 949 - Documentation tweak
679 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
682 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
684 PP/02 Implemented gsasl authenticator.
686 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
688 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
689 `pkg-config foo` for cflags/libs.
691 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
692 with rest of GSASL and with heimdal_gssapi.
694 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
695 `pkg-config foo` for cflags/libs for the TLS implementation.
697 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
698 properties get this fed in as external SSF. A number of robustness
699 and debugging improvements to the cyrus_sasl authenticator.
701 PP/08 cyrus_sasl server now expands the server_realm option.
703 PP/09 Bugzilla 1214 - Log authentication information in reject log.
704 Patch by Jeremy Harris.
706 PP/10 Added dbmjz lookup type.
708 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
710 PP/12 MAIL args handles TAB as well as SP, for better interop with
711 non-compliant senders.
712 Analysis and variant patch by Todd Lyons.
714 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
715 Bug report from Lars Müller <lars@samba.org> (via SUSE),
716 Patch from Dirk Mueller <dmueller@suse.com>
718 PP/13 tls_peerdn now print-escaped for spool files.
719 Observed some $tls_peerdn in wild which contained \n, which resulted
720 in spool file corruption.
722 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
723 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
724 or write after TLS renegotiation, which otherwise led to messages
727 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
728 as a tracking header (ie: a signed header comes before the signature).
729 Patch from Wolfgang Breyha.
731 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
732 comma-sep list; embedded commas doubled.
734 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
736 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
738 Report and patch from Dmitry Banschikov.
740 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
741 Removed SSL_clear() after SSL_new() which led to protocol negotiation
742 failures. We appear to now support TLS1.1+ with Exim.
744 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
745 lets Exim select keys and certificates based upon TLS SNI from client.
746 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
747 before an outbound SMTP session. New log_selector, +tls_sni.
749 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
750 NULL dereference. Report and patch from Alun Jones.
752 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
753 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
754 Not seeing resolver debug output on NetBSD, but suspect this is a
755 resolver implementation change.
757 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
758 Left warnings. Added "eximon gdb" invocation mode.
760 PP/21 Defaulting "accept_8bitmime" to true, not false.
762 PP/22 Added -bw for inetd wait mode support.
764 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
765 locate the relevant includes and libraries. Made this the default.
767 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
768 Bugzilla 1246, report and most of solution from Tomasz Kusy.
770 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
771 This may cause build issues on older platforms.
773 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
774 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
775 gnutls_require_mac & gnutls_require_protocols (no longer supported).
776 Added SNI support via GnuTLS too.
777 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
779 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
781 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
782 Applied second patch from Janne, implementing suggestion to default
783 multiple-strings-in-record handling to match SPF spec.
785 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
787 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
788 read-only, out of scope).
789 Patch from Wolfgang Breyha, report from Stuart Northfield.
791 PP/29 Fix three issues highlighted by clang analyser static analysis.
792 Only crash-plausible issue would require the Cambridge-specific
793 iplookup router and a misconfiguration.
794 Report from Marcin Mirosław.
796 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
798 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
799 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
800 As part of this, removing so much warning spew let me fix some minor
801 real issues in debug logging.
803 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
804 assignment on my part. Fixed.
806 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
807 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
808 Janne Snabb (who went above and beyond: thank you).
810 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
811 string otherwise requires a connection and a bunch more work and it's
812 relatively easy to get wrong. Should also expose TLS library linkage
815 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
816 64-bit ${eval} (JH/03).
818 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
819 GNU libc to support some of the 64-bit stuff, should not lead to
820 conflicts. Defined before os.h is pulled in, so if a given platform
821 needs to override this, it can.
823 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
824 protection layer was required, which is not implemented.
825 Bugzilla 1254, patch from Wolfgang Breyha.
827 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
828 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
829 tls_dhparam take prime identifiers. Also unbreak combination of
830 OpenSSL+DH_params+TLSSNI.
832 PP/39 Disable SSLv2 by default in OpenSSL support.
838 PP/01 Solaris build fix for Oracle's LDAP libraries.
839 Bugzilla 1109, patch from Stephen Usher.
841 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
843 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
846 TF/02 Fix a couple more cases where we did not log the error message
847 when unlink() failed. See also change 4.74-TF/03.
849 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
850 lock up or crash if it happened to be inside a call to libc when it
851 got a SIGUSR1 from exiwhat.
853 The SIGUSR1 handler appends the current process status to the process
854 log which is later printed by exiwhat. It used to use the general
855 purpose logging code to do this, but several functions it calls are
856 not safe for signals.
858 The new output code in the SIGUSR1 handler is specific to the process
859 log, and simple enough that it's easy to inspect for signal safety.
860 Removing some special cases also simplifies the general logging code.
861 Removing the spurious timestamps from the process log simplifies
864 TF/04 Improved ratelimit ACL condition.
866 The /noupdate option has been deprecated in favour of /readonly which
867 has clearer semantics. The /leaky, /strict, and /readonly update modes
868 are mutually exclusive. The update mode is no longer included in the
869 database key; it just determines when the database is updated. (This
870 means that when you upgrde Exim will forget old rate measurements.)
872 Exim now checks that the per_* options are used with an update mode that
873 makes sense for the current ACL. For example, when Exim is processing a
874 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
875 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
876 must specify per_mail/readonly. If you omit the update mode it defaults to
877 /leaky where that makes sense (as before) or /readonly where required.
879 The /noupdate option is now undocumented but still supported for
880 backwards compatibility. It is equivalent to /readonly except that in
881 ACLs where /readonly is required you may specify /leaky/noupdate or
882 /strict/noupdate which are treated the same as /readonly.
884 A useful new feature is the /count= option. This is a generalization
885 of the per_byte option, so that you can measure the throughput of other
886 aggregate values. For example, the per_byte option is now equivalent
887 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
889 The per_rcpt option has been generalized using the /count= mechanism
890 (though it's more complicated than the per_byte equivalence). When it is
891 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
892 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
893 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
894 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
895 Note that using per_rcpt with a non-readonly update mode in more than
896 one ACL will cause the recipients to be double-counted. (The per_mail
897 and per_byte options don't have this problem.)
899 The handling of very low rates has changed slightly. If the computed rate
900 is less than the event's count (usually one) then this event is the first
901 after a long gap. In this case the rate is set to the same as this event's
902 count, so that the first message of a spam run is counted properly.
904 The major new feature is a mechanism for counting the rate of unique
905 events. The new per_addr option counts the number of different
906 recipients that someone has sent messages to in the last time period. It
907 behaves like per_rcpt if all the recipient addresses are different, but
908 duplicate recipient addresses do not increase the measured rate. Like
909 the /count= option this is a general mechanism, so the per_addr option
910 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
911 example, measure the rate that a client uses different sender addresses
912 with the options per_mail/unique=$sender_address. There are further
913 details in the main documentation.
915 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
917 TF/06 Removed a few PCRE remnants.
919 TF/07 Automatically extract Exim's version number from tags in the git
920 repository when doing development or release builds.
922 PP/02 Raise smtp_cmd_buffer_size to 16kB.
923 Bugzilla 879. Patch from Paul Fisher.
925 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
926 Heavily based on revision 40f9a89a from Simon Arlott's tree.
929 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
931 PP/05 Variable $av_failed, true if the AV scanner deferred.
932 Bugzilla 1078. Patch from John Horne.
934 PP/06 Stop make process more reliably on build failure.
935 Bugzilla 1087. Patch from Heiko Schlittermann.
937 PP/07 Make maildir_use_size_file an _expandable_ boolean.
938 Bugzilla 1089. Patch from Heiko Schlittermann.
940 PP/08 Handle ${run} returning more data than OS pipe buffer size.
941 Bugzilla 1131. Patch from Holger Weiß.
943 PP/09 Handle IPv6 addresses with SPF.
944 Bugzilla 860. Patch from Wolfgang Breyha.
946 PP/10 GnuTLS: support TLS 1.2 & 1.1.
948 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
951 PP/11 match_* no longer expand right-hand-side by default.
952 New compile-time build option, EXPAND_LISTMATCH_RHS.
953 New expansion conditions, "inlist", "inlisti".
955 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
957 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
959 PP/14 fix log_write() format string regression from TF/03.
960 Bugzilla 1152. Patch from Dmitry Isaikin.
966 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
968 PP/02 Harmonised TLS library version reporting; only show if debugging.
969 Layout now matches that introduced for other libraries in 4.74 PP/03.
971 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
973 PP/04 New "dns_use_edns0" global option.
975 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
978 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
979 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
981 TK/01 Updated PolarSSL code to 0.14.2.
982 Bugzilla 1097. Patch from Andreas Metzler.
984 PP/07 Catch divide-by-zero in ${eval:...}.
987 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
990 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
991 format-string attack -- SECURITY: remote arbitrary code execution.
993 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
994 time unintentionally subject to list matching rules, letting the header
995 cause arbitrary Exim lookups (of items which can occur in lists, *not*
996 arbitrary string expansion). This allowed for information disclosure.
998 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
999 INT_MIN/-1 -- value coerced to INT_MAX.
1005 NM/01 Workround for PCRE version dependency in version reporting
1008 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1009 This fixes portability to compilers other than gcc, notably
1010 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1012 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1013 makefiles for portability to HP-UX and POSIX correctness.
1015 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1016 Also via indented variable definition in the Makefile.
1017 (Debugging by Oliver Heesakkers).
1019 PP/02 Restore caching of spamd results with expanded spamd_address.
1020 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1022 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1023 Improves build reliability. Fix from: Frank Elsner
1025 NM/02 Fix wide character breakage in the rfc2047 coding
1026 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1028 NM/03 Allow underscore in dnslist lookups
1029 Fixes bug 1026. Patch from Graeme Fowler
1031 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1032 Code patches from Adam Ciarcinski of NetBSD.
1034 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1037 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1038 is logged, to avoid truncation. Patch from John Horne.
1040 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1041 Patch from Jakob Hirsch.
1043 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1044 SQL string expansion failure details.
1045 Patch from Andrey Oktyabrski.
1047 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1048 Patch from Simon Arlott.
1050 PP/09 New lookups functionality failed to compile on old gcc which rejects
1051 extern declarations in function scope.
1052 Patch from Oliver Fleischmann
1054 PP/10 Use sig_atomic_t for flags set from signal handlers.
1055 Check getgroups() return and improve debugging.
1056 Fixed developed for diagnosis in bug 927 (which turned out to be
1059 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1060 Patch from Mark Zealey.
1062 PP/12 Bugzilla 1056: Improved spamd server selection.
1063 Patch from Mark Zealey.
1065 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1066 Based on patch from Heiko Schlittermann.
1068 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1069 Patch from Uwe Doering, sign-off by Michael Haardt.
1071 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1072 variable declaration deep within a block. Bug and patch from
1075 PP/15 lookups-Makefile IRIX compatibilty coercion.
1077 PP/16 Make DISABLE_DKIM build knob functional.
1079 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1080 Patch by Simon Arlott
1082 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1083 variable argument macros. Our copy now differs from upstream.
1089 TF/01 Failure to get a lock on a hints database can have serious
1090 consequences so log it to the panic log.
1092 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1093 controlled using the smtp_confirmation log selector.
1095 TF/03 Include the error message when we fail to unlink a spool file.
1097 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1098 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1099 for maintaining out-of-tree patches for some time.
1101 PP/01 Bugzilla 139: Documentation and portability issues.
1102 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1103 Handle per-OS dynamic-module compilation flags.
1105 PP/02 Let /dev/null have normal permissions.
1106 The 4.73 fixes were a little too stringent and complained about the
1107 permissions on /dev/null. Exempt it from some checks.
1108 Reported by Andreas M. Kirchwitz.
1110 PP/03 Report version information for many libraries, including
1111 Exim version information for dynamically loaded libraries. Created
1112 version.h, now support a version extension string for distributors
1113 who patch heavily. Dynamic module ABI change.
1115 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1116 privilege escalation vulnerability whereby the Exim run-time user
1117 can cause root to append content of the attacker's choosing to
1120 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1123 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1124 If dropping privileges for untrusted macros, we disabled normal logging
1125 on the basis that it would fail; for the Exim run-time user, this is not
1126 the case, and it resulted in successful deliveries going unlogged.
1127 Fixed. Reported by Andreas Metzler.
1133 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1134 only prepend for the Resent-* case. Fixes regression introduced in
1135 Exim 4.70 by NM/22 for Bugzilla 607.
1137 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1138 increasing numbers of administrators be bitten by this.
1140 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1142 PP/03 Bugzilla 994: added openssl_options main configuration option.
1144 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1146 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1148 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1150 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1151 without a peer certificate, leading to a segfault because of an
1152 assumption that peers always have certificates. Be a little more
1153 paranoid. Problem reported by Martin Tscholak.
1155 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1156 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1157 NB: ClamAV planning to remove STREAM in "middle of 2010".
1158 CL also introduces -bmalware, various -d+acl logging additions and
1159 more caution in buffer sizes.
1161 PP/09 Implemented reverse_ip expansion operator.
1163 PP/10 Bugzilla 937: provide a "debug" ACL control.
1165 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1167 PP/12 Bugzilla 973: Implement --version.
1169 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1171 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1173 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1175 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1176 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1177 ignore trailing whitespace.
1179 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1181 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1184 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1185 Notification from Dr Andrew Aitchison.
1187 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1188 ExtendedDetectionInfo response format.
1189 Notification from John Horne.
1191 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1194 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1195 XSL and documented dependency on system catalogs, with examples of how
1198 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1201 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1202 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1203 configuration file which is writeable by the Exim user or group.
1205 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1206 of configuration files to cover files specified with the -C option if
1207 they are going to be used with root privileges, not just the default
1210 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1211 option (effectively making it always true).
1213 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1214 files to be used while preserving root privileges.
1216 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1217 that rogue child processes cannot use them.
1219 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1220 run-time user, instead of root.
1222 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1223 Exim run-time user without dropping privileges.
1225 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1226 result string, instead of calling string_vformat() twice with the same
1229 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1230 for other users. Others should always drop root privileges if they use
1231 -C on the command line, even for a whitelisted configure file.
1233 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1235 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1241 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1242 $data_path, and $header_path variables; fixed documentation bugs and
1245 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1246 exipick to access non-standard spools, including the "frozen" queue
1249 NM/01 Bugzilla 965: Support mysql stored procedures.
1250 Patch from Alain Williams
1252 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1254 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1255 Patch from Andreas Metzler
1257 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1258 Patch from Kirill Miazine
1260 NM/05 Bugzilla 671: Added umask to procmail example.
1262 JJ/03 installed exipick 20100323.0, fixing doc bug
1264 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1265 directory. Notification and patch from Dan Rosenberg.
1267 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1269 TK/02 Improve log output when DKIM signing operation fails.
1271 MH/01 Treat the transport option dkim_domain as a colon separated
1272 list, not as a single string, and sign the message with each element,
1273 omitting multiple occurences of the same signer.
1275 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1276 Bugzilla 985, 986. Patch by Simon Arlott
1278 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1279 Patch by Simon Arlott
1281 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1282 MBX locking. Notification from Dan Rosenberg.
1288 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1290 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1292 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1294 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1296 NM/04 Bugzilla 915: Flag broken perl installation during build.
1302 TK/01 Added patch by Johannes Berg that expands the main option
1303 "spamd_address" if it starts with a dollar sign.
1305 TK/02 Write list of recipients to X-Envelope-Sender header when building
1306 the mbox-format spool file for content scanning (suggested by Jakob
1309 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1310 (http://www.dcc-servers.net/) support via dccifd. Activated by
1311 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1313 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1314 by Mark Daniel Reidel <mr@df.eu>.
1316 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1317 When building exim an external PCRE library is now needed -
1318 PCRE is a system library on the majority of modern systems.
1319 See entry on PCRE_LIBS in EDITME file.
1321 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1322 conversation. Added nologin parameter to request.
1323 Patch contributed by Kirill Miazine.
1325 TF/01 Do not log submission mode rewrites if they do not change the address.
1327 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1329 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1330 log files in place. Contributed by Roberto Lima.
1332 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1334 TF/03 Bugzilla 615: When checking the local_parts router precondition
1335 after a local_part_suffix or local_part_prefix option, Exim now
1336 does not use the address's named list lookup cache, since this
1337 contains cached lookups for the whole local part.
1339 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1340 Robert Millan. Documentation is in experimental-spec.txt.
1342 TF/04 Bugzilla 668: Fix parallel build (make -j).
1344 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1346 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1347 Patch provided by Jan Srzednicki.
1349 TF/05 Leading white space used to be stripped from $spam_report which
1350 wrecked the formatting. Now it is preserved.
1352 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1353 that they are available at delivery time.
1355 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1357 TF/08 TLS error reporting now respects the incoming_interface and
1358 incoming_port log selectors.
1360 TF/09 Produce a more useful error message if an SMTP transport's hosts
1361 setting expands to an empty string.
1363 NM/06 Bugzilla 744: EXPN did not work under TLS.
1364 Patch provided by Phil Pennock.
1366 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1367 Patch provided by Richard Godbee.
1369 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1370 acl_smtp_notquit, added index entry.
1372 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1373 Patch provided by Eugene Bujak.
1375 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1376 accept(). Patch provided by Maxim Dounin.
1378 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1379 Patch provided by Phil Pennock.
1381 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1383 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1384 Patch provided by Brad "anomie" Jorsch.
1386 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1387 Patch provided by Dean Brooks.
1389 TK/05 Add native DKIM support (does not depend on external libraries).
1391 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1392 Patch provided by Graeme Fowler.
1394 NM/16 Bugzilla 851: Documentation example syntax fix.
1396 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1398 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1401 NM/19 Bugzilla 745: TLS version reporting.
1402 Patch provided by Phil Pennock.
1404 NM/20 Bugzilla 167: bool: condition support.
1405 Patch provided by Phil Pennock.
1407 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1408 clients. Patch provided by Phil Pennock.
1410 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1411 Patch provided by Brad "anomie" Jorsch.
1413 NM/23 Bugzilla 687: Fix misparses in eximstats.
1414 Patch provided by Heiko Schlittermann.
1416 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1417 Patch provided by Heiko Schlittermann.
1419 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1420 plus update to original patch.
1422 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1424 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1425 Patch provided by David Brownlee.
1427 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1429 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1431 NM/30 Bugzilla 888: TLS documentation bugfixes.
1433 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1435 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1436 Unlike the original bugzilla I have changed all shell scripts in src tree.
1438 NM/33 Bugzilla 898: Transport filter timeout fix.
1439 Patch by Todd Rinaldo.
1441 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1442 Patch by Serge Demonchaux.
1444 NM/35 Bugzilla 39: Base64 decode bug fixes.
1445 Patch by Jakob Hirsch.
1447 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1449 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1451 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1453 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1459 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1460 ALT-N's libdkim that I have put here:
1461 http://duncanthrax.net/exim-experimental/
1463 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1464 were called 'true' and it seems that C99 defines that as a reserved
1465 keyword to be used with 'bool' variable types. That means you could
1466 not include C99-style headers which use bools without triggering
1467 build errors in sieve.c.
1469 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1470 as mailq or other aliases. Changed the --help handling significantly
1471 to do whats expected. exim_usage() emits usage/help information.
1473 SC/01 Added the -bylocaldomain option to eximstats.
1475 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1477 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1479 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1485 PH/01 Another patch from the Sieve maintainer.
1487 PH/02 When an IPv6 address is converted to a string for single-key lookup
1488 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1489 dots are used instead of colons so that keys in lsearch files need not
1490 contain colons. This was done some time before quoting was made available
1491 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1492 (notated using the quote facility) so as to distinguish them from IPv4
1493 keys. This meant that lookups for IP addresses in host lists did not work
1494 for iplsearch lookups.
1496 This has been fixed by arranging for IPv6 addresses to be expressed with
1497 colons if the lookup type is iplsearch. This is not incompatible, because
1498 previously such lookups could never work.
1500 The situation is now rather anomolous, since one *can* have colons in
1501 ordinary lsearch keys. However, making the change in all cases is
1502 incompatible and would probably break a number of configurations.
1504 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1507 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1508 conversion specification without a maximum field width, thereby enabling
1509 a rogue spamd server to cause a buffer overflow. While nobody in their
1510 right mind would setup Exim to query an untrusted spamd server, an
1511 attacker that gains access to a server running spamd could potentially
1512 exploit this vulnerability to run arbitrary code as the Exim user.
1514 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1515 $primary_hostname instead of what libspf2 thinks the hosts name is.
1517 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1518 a directory entry by the name of the lookup key. Previously, if a
1519 symlink pointed to a non-existing file or a file in a directory that
1520 Exim lacked permissions to read, a lookup for a key matching that
1521 symlink would fail. Now it is enough that a matching directory entry
1522 exists, symlink or not. (Bugzilla 503.)
1524 PH/03 The body_linecount and body_zerocount variables are now exported in the
1527 PH/04 Added the $dnslist_matched variable.
1529 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1530 This means they are set thereafter only if the connection becomes
1533 PH/06 Added the client_condition to authenticators so that some can be skipped
1534 by clients under certain conditions.
1536 PH/07 The error message for a badly-placed control=no_multiline_responses left
1537 "_responses" off the end of the name.
1539 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1541 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1542 (without spaces) instead of just copying the configuration text.
1544 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1546 PH/11 Added $max_received_linelength.
1548 PH/12 Added +ignore_defer and +include_defer to host lists.
1550 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1551 way in which PCRE > 7.0 is built.
1553 PH/14 Implemented queue_only_load_latch.
1555 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1556 MAIL command. The effect was to mangle the value on 64-bit systems.
1558 PH/16 Another patch from the Sieve maintainer.
1560 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1562 PH/18 If a system quota error occurred while trying to create the file for
1563 a maildir delivery, the message "Mailbox is full" was not appended to the
1564 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1565 only to a quota excession during the actual writing of the file.
1567 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1568 characters?) which causes problems in log lines. The DN values are now
1569 passed through string_printing() before being added to log lines.
1571 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1572 and InterBase are left for another time.)
1574 PH/21 Added message_body_newlines option.
1576 PH/22 Guard against possible overflow in moan_check_errorcopy().
1578 PH/23 POSIX allows open() to be a macro; guard against that.
1580 PH/24 If the recipient of an error message contained an @ in the local part
1581 (suitably quoted, of course), incorrect values were put in $domain and
1582 $local_part during the evaluation of errors_copy.
1588 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1589 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1592 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1593 issue a MAIL command.
1595 PH/02 In an ACL statement such as
1597 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1599 if a client was not listed at all, or was listed with a value other than
1600 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1601 the condition was not true (as it should be), so access was not denied.
1602 The bug was that the ! inversion was incorrectly passed on to the second
1603 item. This has been fixed.
1605 PH/03 Added additional dnslists conditions == and =& which are different from
1606 = and & when the dns lookup returns more than one IP address.
1608 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1609 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1611 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1612 FSYNC, which compiles an option called disable_fsync that allows for
1613 bypassing fsync(). The documentation is heavily laced with warnings.
1615 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1617 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1618 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1619 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1620 including adding "make clean"; (3) Added -fPIC when compiling the test
1621 dynamically loaded module, to get rid of a warning.
1623 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1624 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1625 = 0s. The bug is otherwise harmless.
1627 PH/07 There was a bug in the dovecot authenticator such that the value of
1628 $auth1 could be overwritten, and so not correctly preserved, after a
1629 successful authentication. This usually meant that the value preserved by
1630 the server_setid option was incorrect.
1632 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1634 PH/09 Installed PCRE release 7.0.
1636 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1637 run for batched SMTP input. It is now run at the start of every message
1638 in the batch. While fixing this I discovered that the process information
1639 (output by running exiwhat) was not always getting set for -bs and -bS
1640 input. This is fixed, and it now also says "batched" for BSMTP.
1642 PH/11 Added control=no_pipelining.
1644 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1645 patch, slightly modified), and move the expansion of helo_data till after
1646 the connection is made in the smtp transport (so it can use these
1649 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1651 PH/14 Added log_selector = +pid.
1653 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1655 PH/16 Add ${if forany and ${if forall.
1657 PH/17 Added dsn_from option to vary the From: line in DSNs.
1659 PH/18 Flush SMTP output before performing a callout, unless control =
1660 no_callout_flush is set.
1662 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1663 was true (the default) a successful delivery failed to delete the retry
1664 item, thus causing premature timeout of the address. The bug is now
1667 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1669 PH/21 Long custom messages for fakedefer and fakereject are now split up
1670 into multiline reponses in the same way that messages for "deny" and
1671 other ACL rejections are.
1673 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1674 with slight modification.
1676 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1677 draft, changing the syntax and factoring some duplicate code".
1679 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1680 for deliveries of the second and subsequent messages over the same SMTP
1683 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1684 ${reduce, with only minor "tidies".
1686 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1688 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1689 expansion side effects.
1691 PH/27 When a message times out after an over-quota error from an Exim-imposed
1692 quota, the bounce message says "mailbox is full". This message was not
1693 being given when it was a system quota that was exceeded. It now should
1696 MH/03 Made $recipients available in local_scan(). local_scan() already has
1697 better access to the recipient list through recipients_list[], but
1698 $recipients can be useful in postmaster-provided expansion strings.
1700 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1701 in the case of a MAIL command with additional options following the
1702 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1703 were accidentally chopped off.
1705 PH/29 SMTP synchronization checks are implemented when a command is read -
1706 there is a check that no more input is waiting when there shouldn't be
1707 any. However, for some commands, a delay in an ACL can mean that it is
1708 some time before the response is written. In this time, more input might
1709 arrive, invalidly. So now there are extra checks after an ACL has run for
1710 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1711 pipelining has not been advertised.
1713 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1715 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1716 returned in the SMTP response when smtp_return_error_details was false.
1717 This has been fixed.
1719 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1720 instead of the C I/O that was originally supplied, because problems were
1721 reported on Solaris.
1723 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1724 Exim which did not show up earlier: it was assuming that a call to
1725 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1726 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1727 was a macro that became an assignment, so it seemed to work. This has
1728 changed to a proper function call with a void return, hence the compile
1729 error. Exim's code has been fixed.
1731 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1734 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1736 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1738 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1739 because it turns out to be too big or there is a timeout) there is no
1740 "Completed" line in the log. When some messages of this type were
1741 selected by exigrep, they were listed as "not completed". Others were
1742 picked up by some special patterns. I have improved the selection
1743 criteria to be more general.
1745 PH/38 The host_find_failed option in the manualroute router can now be set
1746 to "ignore", to completely ignore a host whose IP address cannot be
1747 found. If all hosts are ignored, the behaviour is controlled by the new
1748 host_all_ignored option.
1750 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1751 homing or because of multiple MX records with /mx) generated more than
1752 one IP address, and the following item turned out to be the local host,
1753 all the secondary addresses of the first item were incorrectly removed
1754 from the list, along with the local host and any following hosts (which
1755 is what is supposed to happen).
1757 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1758 whoever called Exim into the -H file. In the case of the daemon it was
1759 behaving confusingly. When first started, it used values for whoever
1760 started the daemon, but after a SIGHUP it used the Exim user (because it
1761 calls itself on a restart). I have changed the code so that it now always
1764 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1765 message are rejected with the same error (e.g. no authentication or bad
1766 sender address), and a DATA command is nevertheless sent (as can happen
1767 with PIPELINING or a stupid MUA), the error message that was given to the
1768 RCPT commands is included in the rejection of the DATA command. This is
1769 intended to be helpful for MUAs that show only the final error to their
1772 PH/42 Another patch from the Sieve maintainer.
1774 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1775 Eximstats - Fixed some broken HTML links and added missing column headers
1777 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1778 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1780 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1782 PH/43 Yet another patch from the Sieve maintainer.
1784 PH/44 I found a way to check for a TCP/IP connection going away before sending
1785 the response to the final '.' that terminates a message, but only in the
1786 case where the client has not sent further data following the '.'
1787 (unfortunately, this is allowed). However, in many cases there won't be
1788 any further data because there won't be any more messages to send. A call
1789 to select() can be used: if it shows that the input is "ready", there is
1790 either input waiting, or the socket has been closed. An attempt to read
1791 the next input character can distinguish the two cases. Previously, Exim
1792 would have sent an OK response which the client would never have see.
1793 This could lead to message repetition. This fix should cure that, at
1794 least in a lot of common cases.
1796 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1797 advertised in response to EHLO.
1803 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1804 fixed by 4.65/MH/01 (is this a record?) are fixed:
1806 (i) An empty string was always treated as zero by the numeric comparison
1807 operators. This behaviour has been restored.
1809 (ii) It is documented that the numeric comparison operators always treat
1810 their arguments as decimal numbers. This was broken in that numbers
1811 starting with 0 were being interpreted as octal.
1813 While fixing these problems I realized that there was another issue that
1814 hadn't been noticed. Values of message_size_limit (both the global option
1815 and the transport option) were treated as octal if they started with 0.
1816 The documentation was vague. These values are now always treated as
1817 decimal, and I will make that clear in the documentation.
1823 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1824 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1827 MH/01 Don't check that the operands of numeric comparison operators are
1828 integers when their expansion is in "skipping" mode (fixes bug
1829 introduced by 4.64-PH/07).
1831 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1832 child addresses, Exim now panics and dies. Previously, because the count
1833 is held in a short int, deliveries were likely to be lost. As such a
1834 large number of recipients for a single message is ridiculous
1835 (performance will be very, very poor), I have chosen to impose a limit
1836 rather than extend the field.
1842 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1843 leftover -K file (the existence of which was triggered by #402).
1844 While we were at it, introduced process PID as part of the -K
1845 filename. This should rule out race conditions when creating
1848 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1849 processing considerably. Previous code took too long for large mails,
1850 triggering a timeout which in turn triggers #401.
1852 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1853 in the DK code in transports.c. sendfile() is not really portable,
1854 hence the _LINUX specificness.
1856 TF/01 In the add_headers option to the mail command in an Exim filter,
1857 there was a bug that Exim would claim a syntax error in any
1858 header after the first one which had an odd number of characters
1861 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1862 callout verification, Exim cached a "reject" for the entire domain. This
1863 is correct for most verifications, but it is not correct for a recipient
1864 verification with use_sender or use_postmaster set, because in that case
1865 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1866 case of MAIL FROM:<> rejection from other early rejections (e.g.
1867 rejection of HELO). When verifying a recipient using a non-null MAIL
1868 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1869 Whatever the result of the callout, the value of the domain cache is
1870 left unchanged (for any other kind of callout, getting as far as trying
1871 RCPT means that the domain itself is ok).
1873 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1876 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1877 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1878 session, the daemon ignores ECONNECT errors and logs others; it now
1879 ignores EPIPE as well.
1881 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1882 (quoted-printable decoding).
1884 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1885 later the small subsequent patch to fix an introduced bug.
1887 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1889 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1891 PH/08 An error is now given if message_size_limit is specified negative.
1893 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1894 to be given (somewhat) arbitrary names.
1896 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1899 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1900 miscellaneous code fixes
1902 PH/10 Added the log_reject_target ACL modifier to specify where to log
1905 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1906 hostname. This is wrong, because it relates to the incoming message (and
1907 probably the interface on which it is arriving) and not to the outgoing
1908 callout (which could be using a different interface). This has been
1909 changed to use the value of the helo_data option from the smtp transport
1910 instead - this is what is used when a message is actually being sent. If
1911 there is no remote transport (possible with a router that sets up host
1912 addresses), $smtp_active_hostname is used.
1914 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1915 tweaks were necessary in order to get it to work (see also 21 below):
1916 (a) The code assumed that strncpy() returns a negative number on buffer
1917 overflow, which isn't the case. Replaced with Exim's string_format()
1919 (b) There were several signed/unsigned issues. I just did the minimum
1920 hacking in of casts. There is scope for a larger refactoring.
1921 (c) The code used strcasecmp() which is not a standard C function.
1922 Replaced with Exim's strcmpic() function.
1923 (d) The code set only $1; it now sets $auth1 as well.
1924 (e) A simple test gave the error "authentication client didn't specify
1925 service in request". It would seem that Dovecot has changed its
1926 interface. Fortunately there's a specification; I followed it and
1927 changed what the client sends and it appears to be working now.
1929 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1932 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1933 address A is aliased to B and C, where B exists and C does not. Without
1934 -v the output is "A verified" because verification stops after a
1935 successful redirection if more than one address is generated. However,
1936 with -v the child addresses are also verified. Exim was outputting "A
1937 failed to verify" and then showing the successful verification for C,
1938 with its parentage. It now outputs "B failed to verify", showing B's
1939 parentage before showing the successful verification of C.
1941 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1942 look up a TXT record in a specific list after matching in a combined
1945 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1946 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1947 they consult the DNS. I had assumed they would set it the way they
1948 wanted; and indeed my experiments on Linux seem to show that in some
1949 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1950 To be on the safe side, however, I have now made the interface to
1951 host_find_byname() similar to host_find_bydns(), with an argument
1952 containing the DNS resolver options. The host_find_byname() function now
1953 sets these options at its start, just as host_find_bydns() does. The smtp
1954 transport options dns_qualify_single and dns_search_parents are passed to
1955 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1956 of host_find_byname() use the default settings of RES_DEFNAMES
1957 (qualify_single) but not RES_DNSRCH (search_parents).
1959 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1960 spool_read_header() do less string testing, by means of a preliminary
1961 switch on the second character of optional "-foo" lines. (This is
1962 overdue, caused by the large number of possibilities that now exist.
1963 Originally there were few.) While I was there, I also converted the
1964 str(n)cmp tests so they don't re-test the leading "-" and the first
1965 character, in the hope this might squeeze out yet more improvement.
1967 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1968 flag allowing group syntax was set by the header_syntax check but not
1969 turned off, possible causing trouble later; (2) The flag was not being
1970 set at all for the header_verify test, causing "group"-style headers to
1971 be rejected. I have now set it in this case, and also caused header_
1972 verify to ignore an empty address taken from a group. While doing this, I
1973 came across some other cases where the code for allowing group syntax
1974 while scanning a header line wasn't quite right (mostly, not resetting
1975 the flag correctly in the right place). These bugs could have caused
1976 trouble for malformed header lines. I hope it is now all correct.
1978 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1979 with the "reply" argument non-NULL. The code, however (which originally
1980 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1981 but it didn't always do it. This confused somebody who was copying the
1982 code for some other use. I have removed all the tests.
1984 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1985 feature that was used to support insecure browsers during the U.S. crypto
1986 embargo. It requires special client support, and Exim is probably the
1987 only MTA that supported it -- and would never use it because real RSA is
1988 always available. This code has been removed, because it had the bad
1989 effect of slowing Exim down by computing (never used) parameters for the
1990 RSA_EXPORT functionality.
1992 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1993 authenticator to fail if there's a tab character in the incoming data
1994 (there should never be unless someone is messing about, as it's supposed
1995 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1996 if the connection is using TLS or if the remote IP is the same as the
1997 local IP, and the "valid-client-cert option" if a client certificate has
2000 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2001 authenticators. This can be used for authorization after authentication
2002 succeeds. (In the case of plaintext, it servers for both authentication
2005 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2006 if any retry times were supplied.
2008 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2009 connection, where there is no client IP address to check. In this
2010 situation, the verify now always succeeds.
2012 PH/25 Applied John Jetmore's -Mset patch.
2014 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2016 PH/27 In a string expansion for a processed (not raw) header when multiple
2017 headers of the same name were present, leading whitespace was being
2018 removed from all of them, but trailing whitespace was being removed only
2019 from the last one. Now trailing whitespace is removed from each header
2020 before concatenation. Completely empty headers in a concatenation (as
2021 before) are ignored.
2023 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2024 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2026 PH/29 [Removed. This was a change that I later backed out, and forgot to
2027 correct the ChangeLog entry (that I had efficiently created) before
2028 committing the later change.]
2030 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2031 address errors (4xx response to RCPT) over the same connection as other
2032 messages routed to the same hosts. Such deliveries are always "forced",
2033 so retry times are not inspected. This resulted in far too many retries
2034 for the affected addresses. The effect occurred only when there were more
2035 hosts than the hosts_max_try setting in the smtp transport when it had
2036 the 4xx errors. Those hosts that it had tried were not added to the list
2037 of hosts for which the message was waiting, so if all were tried, there
2038 was no problem. Two fixes have been applied:
2040 (i) If there are any address or message errors in an SMTP delivery, none
2041 of the hosts (tried or untried) are now added to the list of hosts
2042 for which the message is waiting, so the message should not be a
2043 candidate for sending over the same connection that was used for a
2044 successful delivery of some other message. This seems entirely
2045 reasonable: after all the message is NOT "waiting for some host".
2046 This is so "obvious" that I'm not sure why it wasn't done
2047 previously. Hope I haven't missed anything, but it can't do any
2048 harm, as the worst effect is to miss an optimization.
2050 (ii) If, despite (i), such a delivery is accidentally attempted, the
2051 routing retry time is respected, so at least it doesn't keep
2052 hammering the server.
2054 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2055 in ${readsocket because some servers need this prod.
2057 PH/32 Added some extra debug output when updating a wait-xxx database.
2059 PH/33 The hint "could be header name not terminated by colon", which has been
2060 given for certain expansion errors for a long time, was not being given
2061 for the ${if def:h_colon_omitted{... case.
2063 PH/34 The spec says: "With one important exception, whenever a domain list is
2064 being scanned, $domain contains the subject domain." There was at least
2065 one case where this was not true.
2067 PH/35 The error "getsockname() failed: connection reset by peer" was being
2068 written to the panic log as well as the main log, but it isn't really
2069 panic-worthy as it just means the connection died rather early on. I have
2070 removed the panic log writing for the ECONNRESET error when getsockname()
2073 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2074 runs only) independently of the message's sender address. This meant
2075 that, if the 4xx error was in fact related to the sender, a different
2076 message to the same recipient with a different sender could confuse
2077 things. In particualar, this can happen when sending to a greylisting
2078 server, but other circumstances could also provoke similar problems.
2079 I have changed the default so that the retry time for these errors is now
2080 based a combination of the sender and recipient addresses. This change
2081 can be overridden by setting address_retry_include_sender=false in the
2084 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2085 remote server are returned as part of bounce messages. This was not
2086 happening for LMTP over a pipe (the lmtp transport), but now it is the
2087 same for both kinds of LMTP.
2089 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2090 in header lines that were in fact CNAMEs. This is no longer the case.
2092 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2093 and queue runs started by the daemon processed all messages. This has
2094 been fixed so that -R and -S can now usefully be given with -q<time>.
2096 PH/40 Import PCRE release 6.7 (fixes some bugs).
2098 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2100 PH/42 Give an error if -q is specified more than once.
2102 PH/43 Renamed the variables $interface_address and $interface_port as
2103 $received_ip_address and $received_port, to make it clear that these
2104 values apply to message reception, and not to the outgoing interface when
2105 a message is delivered. (The old names remain recognized, of course.)
2107 PH/44 There was no timeout on the connect() call when using a Unix domain
2108 socket in the ${readsocket expansion. There now is.
2110 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2111 be meaningful with "accept".
2113 SC/01 Eximstats V1.43
2114 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2116 SC/02 Eximstats V1.44
2117 Use a glob alias rather than an array ref in the generated
2118 parser. This improves both readability and performance.
2120 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2121 Collect SpamAssassin and rejection statistics.
2122 Don't display local sender or destination tables unless
2123 there is data to show.
2124 Added average volumes into the top table text output.
2126 SC/04 Eximstats V1.46
2127 Collect data on the number of addresses (recipients)
2128 as well as the number of messages.
2130 SC/05 Eximstats V1.47
2131 Added 'Message too big' to the list of mail rejection
2132 reasons (thanks to Marco Gaiarin).
2134 SC/06 Eximstats V1.48
2135 Mainlog lines which have GMT offsets and are too short to
2136 have a flag are now skipped.
2138 SC/07 Eximstats V1.49 (Alain Williams)
2139 Added the -emptyok flag.
2141 SC/08 Eximstats V1.50
2142 Fixes for obtaining the IP address from reject messages.
2144 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2145 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2146 whitesspace changes from 4.64-PH/27
2148 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2151 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2152 are found, allow negative numbers in numeric criteria)
2154 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2156 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2157 to match changes made in 4.64-PH/43
2159 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2161 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2162 "domain" is actually an IP address, and force a failure. This locks out
2163 those revolvers/nameservers that support "A-for-A" lookups, in
2164 contravention of the specifications.
2166 PH/48 When a host name was looked up from an IP address, and the subsequent
2167 forward lookup of the name timed out, the host name was left in
2168 $sender_host_name, contrary to the specification.
2170 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2171 restricted to single-key lookups, Exim was not diagnosing an error if
2172 * or *@ was used with a query-style lookup.
2174 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2176 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2177 long ago, but noone interested enough thought of it. Let's just say that
2178 the "1.1" means that there are some new functions that weren't there at
2179 some point in the past.
2181 PH/51 Error processing for expansion failure of helo_data from an smtp
2182 transport during callout processing was broken.
2184 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2185 tested/used via the -bh/-bhc/-bs options.
2187 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2188 bug, fixed in subsequent PCRE releases).
2190 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2191 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2193 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2199 SC/01 Use a glob alias rather than an array ref in eximstats generated
2200 parser. This improves both readability and performance.
2202 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2203 Don't display local sender or destination tables in eximstats unless
2204 there is data to show.
2205 Added average volumes into the eximstats top table text output.
2207 SC/03 Collect data on the number of addresses (recipients) as well
2208 as the number of messages in eximstats.
2210 TF/01 Correct an error in the documentation for the redirect router. Exim
2211 does (usually) call initgroups() when daemonizing.
2213 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2214 with consistent privilege compared to when running as a daemon.
2216 TF/03 Note in the spec that $authenticated_id is not set for local
2217 submissions from trusted users.
2219 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2220 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2222 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2223 by adding some example configuration directives to the default
2224 configuration file. A little bit of work is required to uncomment the
2225 directives and define how usernames and passwords are checked, but
2226 there is now a framework to start from.
2228 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2229 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2230 without this. I don't know how relevant this is to other LDAP libraries.
2232 PH/02 Add the verb name to the "unknown ACL verb" error.
2234 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2236 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2238 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2239 directory not expanded when it should be if an expanded home directory
2240 was set for the address (which is overridden by the transport).
2242 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2245 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2246 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2247 because it is too late at that time, and has no effect.
2249 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2250 security issue with \' (bugzilla #107). I could not use the
2251 PQescapeStringConn() function, because it needs a PGconn value as one of
2254 PH/08 When testing addresses using -bt, indicate those final addresses that
2255 are duplicates that would not cause an additional delivery. At least one
2256 person was confused, thinking that -bt output corresponded to deliveries.
2257 (Suppressing duplicates isn't a good idea as you lose the information
2258 about possibly different redirections that led to the duplicates.)
2260 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2261 systems where poll() doesn't work, in particular OS X.
2263 PH/10 Added more information to debugging output for retry time not reached.
2265 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2266 operations in malware.c.
2268 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2271 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2272 syslog_duplication set false, log lines that would normally be written
2273 both the the main log and to the reject log were not written to syslog at
2276 PH/14 In the default configuration, change the use of "message" in ACL warn
2277 statements to "add_header".
2279 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2280 not followed by a command (e.g. "seen endif").
2282 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2283 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2286 PH/17 Added extra conditions to the default value of delay_warning_condition
2290 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2291 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2292 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2295 The Auto-Submitted: and various List- headers are standardised, whereas I
2296 don't think Precedence: ever was.
2298 PH/18 Refactored debugging code in route_finduser() to show more information,
2299 in particular, the error code if getpwnam() issues one.
2301 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2302 This is apparently needed in addition to the PH/07 change above to avoid
2303 any possible encoding problems.
2305 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2306 but not after initializing Perl.
2308 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2309 output them only if debugging. By default they are written stderr,
2310 apparently, which is not desirable.
2312 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2315 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2318 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2320 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2321 authenticated or an ident call has been made. Suppress the default
2322 values for $authenticated_id and $authenticated_sender (but permit -oMai
2323 and -oMas) when testing with -bh.
2325 PH/24 Re-jigged the order of the tests in the default configuration so that the
2326 tests for valid domains and recipients precede the DNS black list and CSA
2327 tests, on the grounds that those ones are more expensive.
2329 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2330 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2331 This bug exists in every version of Exim that I still have, right back to
2334 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2335 However, an attempt to turn on case-sensitivity in a regex key by
2336 including (?-i) didn't work because the subject string was already
2337 lowercased, and the effects were non-intuitive. It turns out that a
2338 one-line patch can be used to allow (?-i) to work as expected.
2344 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2345 other effects) broke the use of negated acl sub-conditions.
2347 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2350 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2351 "Deny" causes Exim to reject the incoming connection with a 554 error.
2352 Unfortunately, if there is a major crisis, such as a disk failure,
2353 tcp-wrappers gives "deny", whereas what one would like would be some
2354 kind of temporary error. A kludge has been added to help with this.
2355 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2356 554 error is used if errno is still zero or contains ENOENT (which occurs
2357 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2360 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2362 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2363 errors. Otherwise a message that provokes a temporary error (when other
2364 messages do not) can cause a whole host to time out.
2366 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2367 addresses were routed directly to files or pipes from a redirect router.
2368 File deliveries just didn't batch; pipe deliveries might have suffered
2371 PH/06 A failure to get a lock for a hints database would erroneously always say
2372 "Failed to get write lock", even when it was really a read lock.
2374 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2375 of 0600. This has been changed to use the value of the lockfile_mode
2376 option (which defaults to 0600).
2378 PH/08 Applied small patch from the Sieve maintainer.
2380 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2381 folder from quota calculations, a direct delivery into this folder messed
2382 up the contents of the maildirsize file. This was because the regex was
2383 used only to exclude .Trash (or whatever) when the size of the mailbox
2384 was calculated. There was no check that a delivery was happening into an
2385 excluded directory. This bug has been fixed by ignoring all quota
2386 processing for deliveries into excluded directories.
2388 PH/10 Added the maildirfolder_create_regex option to appendfile.
2394 PH/01 The code for finding all the local interface addresses on a FreeBSD
2395 system running IPv6 was broken. This may well have applied to all BSD
2396 systems, as well as to others that have similar system calls. The broken
2397 code found IPv4 interfaces correctly, but gave incorrect values for the
2398 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2399 that it would not match correctly against @[] and not recognize the IPv6
2402 PH/02 The ipliteral router was not recognizing addresses of the form user@
2403 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2405 PH/03 Added disable_ipv6.
2407 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2408 decoded form, because it is most often used to construct To: headers
2409 lines in autoreplies, and the decoded form may well be syntactically
2410 invalid. However, $reply_address has leading white space removed, and all
2411 newlines turned into spaces so that the autoreply transport does not
2414 PH/05 If group was specified without a user on a router, and no group or user
2415 was specified on a transport, the group from the router was ignored.
2417 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2418 for visible compile-time settings that can be used to change these
2419 numbers, for those that want even more. Backwards compatibility with old
2420 spool files has been maintained. However, going back to a previous Exim
2421 release will lost any variables that are in spool files.
2423 PH/07 Two small changes when running in the test harness: increase delay when
2424 passing a TCP/IP connection to a new process, in case the original
2425 process has to generate a bounce, and remove special handling of
2426 127.0.0.2 (sic), which is no longer necessary.
2428 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2429 be the same on different OS.
2431 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2434 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2435 whether --show-vars was specified or not
2437 JJ/02 exipick: Added support for new ACL variable spool format introduced
2440 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2441 syntactically invalid From: or Reply-to: line, and a filter used this to
2442 generate an autoreply, and therefore failed to obtain an address for the
2443 autoreply, Exim could try to deliver to a non-existent relative file
2444 name, causing unrelated and misleading errors. What now happens is that
2445 it logs this as a hard delivery error, but does not attempt to create a
2448 PH/11 The exinext utility has a -C option for testing purposes, but although
2449 the given file was scanned by exinext itself; it wasn't being passed on
2450 when Exim was called.
2452 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2453 an end-of-file indication when reading a command response.
2455 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2456 compiled. In many other places in Exim, IPv6 addresses are always
2457 recognized, so I have changed this. It also means that IPv4 domain
2458 literals of the form [IPV4:n.n.n.n] are now always recognized.
2460 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2461 used if the router is not running as root, for example, when verifying at
2462 ACL time, or when using -bh. The debugging output from this situation was
2463 non-existent - all you got was a failure to exec. I have made two
2466 (a) Failures to set uid/gid, the current directory, or a process leader
2467 in a subprocess such as that created by queryprogram now generate
2468 suitable debugging ouput when -d is set.
2470 (b) The queryprogram router detects when it is not running as root,
2471 outputs suitable debugging information if -d is set, and then runs
2472 the subprocess without attempting to change uid/gid.
2474 PH/15 Minor change to Makefile for building test_host (undocumented testing
2477 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2478 additional section of a DNS packet that returns MX or SRV records.
2479 Instead, it always explicitly searches for A/AAAA records. This avoids
2480 major problems that occur when a DNS server includes only records of one
2481 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2482 fixed another bug: if SRV records were looked up and the corresponding
2483 address records were *not* found in the additional section, the port
2484 values from the SRV records were lost.
2486 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2487 using the correct key (the original address) when searching the retry
2488 rules in order to find which one to use for generating the retry hint.
2490 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2491 adding the default one. Similarly, if it contains a Reply-To: header, the
2492 errors_reply_to option, if set, is not used.
2494 PH/19 When calculating a retry time, Exim used to measure the "time since
2495 failure" by looking at the "first failed" field in the retry record. Now
2496 it does not use this if it is later than than the arrival time of the
2497 message. Instead it uses the arrival time. This makes for better
2498 behaviour in cases where some deliveries succeed, thus re-setting the
2499 "first failed" field. An example is a quota failure for a huge message
2500 when small messages continue to be delivered. Without this change, the
2501 "time since failure" will always be short, possible causing more frequent
2502 delivery attempts for the huge message than are intended.
2503 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2505 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2506 $1, $2, $3) because the numerical variables can be reset during some
2507 expansion items (e.g. "match"), thereby losing the authentication data.
2509 PH/21 Make -bV show the size of off_t variables so that the test suite can
2510 decide whether to run tests for quotas > 2G.
2512 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2513 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2514 filecount value is greater than 2G or if a quota value is greater than 2G
2515 on a system where the size of off_t is not greater than 4, a panic error
2518 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2519 never match. The debug and -bh output now contains an explicit error
2520 message indicating a malformed IPv4 address or mask.
2522 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2523 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2524 PH/23 above applies.
2526 PH/25 Do not write to syslog when running in the test harness. The only
2527 occasion when this arises is a failure to open the main or panic logs
2528 (for which there is an explicit test).
2530 PH/26 Added the /no_tell option to "control=freeze".
2532 PH/27 If a host name lookup failed very early in a connection, for example, if
2533 the IP address matched host_lookup and the reverse lookup yielded a name
2534 that did not have a forward lookup, an error message of the form "no IP
2535 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2536 could be logged. Now it outputs the IP address instead of "NULL".
2538 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2539 allows the sender and the authenticated sender to be set when
2540 submitting a message from within Exim. Since child_open_exim() is
2541 documented for local_scan(), the new function should be too.
2543 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2544 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2545 results in an empty string is now treated as unset.
2547 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2549 PH/31 Added sender_verify_fail logging option.
2551 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2552 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2553 tidied the source and removed it altogether.
2555 PH/33 When a queue run was abandoned because the load average was too high, a
2556 log line was always written; now it is written only if the queue_run log
2557 selector is set. In addition, the log line for abandonment now contains
2558 information about the queue run such as the pid. This is always present
2559 in "start" and "stop" lines but was omitted from the "abandon" line.
2561 PH/34 Omit spaces between a header name and the colon in the error message that
2562 is given when verify = headers_syntax fails (if there are lots of them,
2563 the message gets confusing).
2565 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2566 names, as there are now some PTR records that contain slashes. This check
2567 is only to protect against broken name servers that fall over on strange
2568 characters, so the fact that it applies to all lookups doesn't matter.
2570 PH/36 Now that the new test suite is complete, we can remove some of the
2571 special code in Exim that was needed for the old test suite. For example,
2572 sorting DNS records because real resolvers return them in an arbitrary
2573 order. The new test suite's fake resolver always returns records in the
2576 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2577 bounces) except when queue_only is set, to avoid logging races between
2578 the different processes.
2580 PH/38 Panic-die if .include specifies a non-absolute path.
2582 PH/39 A tweak to the "H" retry rule from its user.
2584 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2585 a label. They prevented compilation on older perls.
2587 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2588 a warning to be raised on newish perls.
2590 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2591 on queue. Changes to match documented behaviour of showing count of
2592 messages matching specified criteria.
2594 PH/40 Changed the default ident timeout from 30s to 5s.
2596 PH/41 Added support for the use of login_cap features, on those BSD systems
2597 that have them, for controlling the resources used by pipe deliveries.
2599 PH/42 The content-scanning code uses fopen() to create files in which to put
2600 message data. Previously it was not paying any attention to the mode of
2601 the files. Exim runs with umask(0) because the rest of the code creates
2602 files with open(), and sets the required mode explicitly. Thus, these
2603 files were ending up world-writeable. This was not a big issue, because,
2604 being within the spool directory, they were not world-accessible. I have
2605 created a function called modefopen, which takes an additional mode
2606 argument. It sets umask(777), creates the file, chmods it to the required
2607 mode, then resets the umask. All the relevant calls to fopen() in the
2608 content scanning code have been changed to use this function.
2610 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2611 to 24 hours. This avoids potential overflow problems when processing G
2612 and H retry rules. I suspect nobody ever tinkers with this value.
2614 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2616 PH/45 When the plaintext authenticator is running as a client, the server's
2617 challenges are checked to ensure they are valid base64 strings. By
2618 default, the authentication attempt is cancelled if an invalid string is
2619 received. Setting client_ignore_invalid_base64 true ignores these errors.
2620 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2621 they are received. Thus, the responses can be made to depend on the
2622 challenges. If an invalid string is ignored, an empty string is placed in
2625 PH/46 Messages that are created by the autoreply transport now contains a
2626 References: header, in accordance with RFCs 2822 and 3834.
2628 PH/47 Added authenticated_sender_force to the smtp transport.
2630 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2632 PH/49 Installed latest patch from the Sieve maintainer.
2634 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2635 was also set, the appendfile transport was unnecessarily scanning a
2636 directory of message files (e.g. for maildir delivery) to find the count
2637 of files (along with the size), even though it did not need this
2638 information. It now does the scan only if it needs to find either the
2639 size of the count of files.
2641 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2643 PH/52 Two bugs concerned with error handling when the smtp transport is
2646 (i) Exim was not creating retry information for temporary errors given
2647 for individual recipients after the DATA command when the smtp transport
2648 was used in LMTP mode. This meant that they could be retried too
2649 frequently, and not timed out correctly.
2651 (ii) Exim was setting the flag that allows error details to be returned
2652 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2653 recipients that were returned after the DATA command.
2655 PH/53 This is related to PH/52, but is more general: for any failing address,
2656 when detailed error information was permitted to be returned to the
2657 sender, but the error was temporary, then after the final timeout, only
2658 "retry timeout exceeded" was returned. Now it returns the full error as
2659 well as "retry timeout exceeded".
2661 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2662 do this, and (what is worse) MTAs that accept it.
2664 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2665 will now be deprecated.
2667 PH/56 New os.c-cygwin from the Cygwin maintainer.
2669 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2670 formats (previously only available in exim formats via -bpr, -bpru,
2671 and -bpra. Now also available in native and exiqgrep formats)
2673 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2674 with very large, slow to parse queues
2676 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2678 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2680 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2681 responses to authentication challenges, though it was showing the
2682 challenges; (ii) I've removed the CR characters from the debug output for
2685 PH/58 Allow for the insertion of a newline as well as a space when a string
2686 is turned into more than one encoded-word during RFC 2047 encoding. The
2687 Sieve code now uses this.
2689 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2690 data_4xx, lost_connection, tls_required.
2692 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2693 message was being sent as an SMTP response.
2695 PH/61 Add -l and -k options to exicyclog.
2697 PH/62 When verifying, if an address was redirected to one new address, so that
2698 verification continued, and the new address failed or deferred after
2699 having set something in $address_data, the value of $address_data was not
2700 passed back to the ACL. This was different to the case when no
2701 redirection occurred. The value is now passed back in both cases.
2703 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2704 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2705 use login_cap.h, so on its own it isn't the distinguishing feature. The
2706 new name refers directly to the setclassresources() function.
2708 PH/65 Added configuration files for NetBSD3.
2710 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2712 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2713 is preferred over IPv4.
2715 PH/68 The bounce_return_message and bounce_return_body options were not being
2716 honoured for bounces generated during the reception of non-SMTP messages.
2717 In particular, this applied to messages rejected by the ACL. This bug has
2718 been fixed. However, if bounce_return_message is true and bounce_return_
2719 body is false, the headers that are returned for a non-SMTP message
2720 include only those that have been read before the error was detected.
2721 (In the case of an ACL rejection, they have all been read.)
2723 PH/69 The HTML version of the specification is now built in a directory called
2724 spec_html instead of spec.html, because the latter looks like a path with
2725 a MIME-type, and this confuses some software.
2727 PH/70 Catch two compiler warnings in sieve.c.
2729 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2730 function verify_get_ident() calls ip_connect() to connect a socket, but
2731 if the "connect()" function timed out, ip_connect() used to close the
2732 socket. However, verify_get_ident() also closes the socket later, and in
2733 between Exim writes to the log, which may get opened at this point. When
2734 the socket was closed in ip_connect(), the log could get the same file
2735 descriptor number as the socket. This naturally causes chaos. The fix is
2736 not to close the socket in ip_connect(); the socket should be closed by
2737 the function that creates it. There was only one place in the code where
2738 this was missing, in the iplookup router, which I don't think anybody now
2739 uses, but I've fixed it anyway.
2741 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2742 well as to direct DNS lookups. Otherwise the handling of names in host
2743 lists is inconsistent and therefore confusing.
2749 PH/01 Two changes to the default runtime configuration:
2751 (1) Move the checks for relay_from_hosts and authenticated clients from
2752 after to before the (commented out) DNS black list checks.
2754 (2) Add control=submission to the relay_from_hosts and authenticated
2755 clients checks, on the grounds that messages accepted by these
2756 statements are most likely to be submissions.
2758 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2760 (1) Generate an error if the third argument for the ${prvs expansion is
2763 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2766 (3) Reset the variables that are obtained from the first argument of
2767 ${prvscheck and used in the second argument before leaving the code,
2768 because their memory is reclaimed, so using them afterwards may do
2771 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2772 one (it's much easier than Tom thought :-).
2774 (5) Because of (4), we can now allow for the use of $prvscheck_result
2775 inside the third argument.
2777 PH/03 For some reason, the default setting of PATH when running a command from
2778 a pipe transport was just "/usr/bin". I have changed it to
2781 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2782 anything to be listed in the output from -bV.
2784 PH/05 When a filter generated an autoreply, the entire To: header line was
2785 quoted in the delivery log line, like this:
2787 => >A.N.Other <ano@some.domain> <original@ddress> ...
2789 This has been changed so that it extracts the operative address. There
2790 may be more than one such address. If so, they are comma-separated, like
2793 => >ano@some.domain,ona@other.domain <original@ddress> ...
2795 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2796 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2797 not being looked up in the rDNS to get a host name, Exim was showing the
2798 IP address twice in Received: lines, even though the IP addresses were
2799 identical. For example:
2801 Received: from [1.2.3.4] (helo=[1.2.3.4])
2803 However, if the real host name was known, it was omitting the HELO data
2804 if it matched the actual IP address. This has been tidied up so that it
2805 doesn't show the same IP address twice.
2807 PH/07 When both +timestamp and +memory debugging was on, the value given by
2808 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2809 called by the debug printing, thereby overwriting the timestamp buffer.
2810 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2812 PH/08 When the original message was included in an autoreply transport, it
2813 always said "this is a copy of the message, including all the headers",
2814 even if body_only or headers_only was set. It now gives an appropriate
2817 PH/09 Applied a patch from the Sieve maintainer which:
2819 o fixes some comments
2820 o adds the (disabled) notify extension core
2821 o adds some debug output for the result of if/elsif tests
2822 o points to the current vacation draft in the documentation
2823 and documents the missing references header update
2827 o fixes a bug in processing the envelope test (when testing
2828 multiple envelope elements, the last element determinted the
2831 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2832 Electronic Mail") by including:
2834 Auto-submitted: auto-generated
2836 in the messages that it generates (bounce messages and others, such as
2837 warnings). In the case of bounce messages for non-SMTP mesages, there was
2838 also a typo: it was using "Auto_submitted" (underscore instead of
2839 hyphen). Since every message generated by Exim is necessarily in response
2840 to another message, thes have all been changed to:
2842 Auto-Submitted: auto-replied
2844 in accordance with these statements in the RFC:
2846 The auto-replied keyword:
2848 - SHOULD be used on messages sent in direct response to another
2849 message by an automatic process,
2851 - MUST NOT be used on manually-generated messages,
2853 - MAY be used on Delivery Status Notifications (DSNs) and Message
2854 Disposition Notifications (MDNs),
2856 - MUST NOT be used on messages generated by automatic or periodic
2857 processes, except for messages which are automatic responses to
2860 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2861 to the default Received: header definition.
2863 PH/12 Added log selector acl_warn_skipped (default on).
2865 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2866 variables because (a) they are in the wrong storage pool and (b) even if
2867 they were copied, it wouldn't work properly because of the caching.
2869 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2870 checking when decoding. Apparently there are clients that generate
2871 overlong encoded strings. Why am I not surprised?
2873 PH/15 If the first argument of "${if match_address" was not empty, but did not
2874 contain an "@" character, Exim crashed. Now it writes a panic log message
2875 and treats the condition as false.
2877 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2879 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2880 extension "envelope-auth". The code is finished and in agreement with
2881 other implementations, but there is no documentation so far and in fact,
2882 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2883 not changing the active code.
2885 Print executed "if" and "elsif" statements when debugging is used. This
2886 helps a great deal to understand what a filter does.
2888 Document more things not specified clearly in RFC3028. I had all this
2889 sorted out, when out of a sudden new issues came to my mind. Oops."
2891 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2894 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2895 canonical form (as documented). However, after a host name lookup from
2896 the IP address, check_host() was doing a simple string comparison with
2897 addresses acquired from the DNS when checking that the found name did
2898 have the original IP as one of its addresses. Since any found IPv6
2899 addresses are likely to be in abbreviated form, the comparison could
2900 fail. Luckily, there already exists a function for doing the comparison
2901 by converting both addresses to binary, so now that is used instead of
2902 the text comparison.
2904 PH/20 There was another similar case to PH/19, when a complete host name was
2905 given in a host list; looking up its IP address could give an abbreviated
2906 form, whereas the current host's name might or might not be abbreviated.
2907 The same fix has been applied.
2913 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2914 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2917 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2918 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2920 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2922 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2923 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2924 Also, for "Auto-Submitted", treat anything other than "no" as
2925 non-personal, in accordance with RFC 3834. (Previously it treated
2926 anything starting "auto-" as non-personal.)
2928 TF/01 The control=submission/name=... option had a problem with syntax
2929 errors if the name included a slash character. The /name= option
2930 now slurps the rest of the string, so it can include any characters
2931 but it must come last in the list of options (after /sender_retain
2934 PH/05 Some modifications to the interface to the fake nameserver for the new
2942 TK/01 Added the "success_on_redirect" address verification option. See
2943 NewStuff for rationale and an example.
2945 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2947 PH/02 Patch to exigrep to allow it to work on syslog lines.
2949 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2950 fread() to skip over the body file's header line, because in Cygwin the
2951 header line is locked and is inaccessible.
2953 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2954 co-exist for some time) to make it clear that it is the Exim ID that is
2955 referenced, not the Message-ID: header line.
2957 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2958 string_format() function, because snprintf() does not exist on all
2961 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2962 ${stat: expansion item.
2964 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2965 protocol synchronization error", to keep the pedants happy.
2967 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2968 well as for IRIX systems, when gcc is being used. See the host.c source
2971 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2973 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2976 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2977 transport and to the smtp transport in LMTP mode.
2979 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2981 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2982 run a filter in a subprocess. This could lead to confusion in subsequent
2983 lookups in the parent process. There should also be a search_tidyup() at
2984 the end of the subprocess.
2986 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2987 only if the host matched helo_try_verify_hosts, which caused the
2988 verification to occur when the EHLO/HELO command was issued. The ACL just
2989 tested the remembered result. Now, if a previous verification attempt has
2990 not happened, "verify = helo" does it there and then.
2992 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2994 TK/03 Fix log output including CR from clamd.
2996 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2997 exist provoked a memory error which could cause a segfault.
2999 PH/15 Installed PCRE 6.2
3001 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3003 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3004 of the problem. Specifically, suggested +O2 rather than +O1 for the
3007 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3009 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3010 record, the port was not correctly logged when the outgoing_port log
3011 selector was set (it logged the transort's default port).
3013 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3014 fallback_hosts, and "hosts" in the smtp transport.
3016 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3017 host errors such as "Connection refused".
3019 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3020 authentication with radiusclient 0.4.9:
3022 - Error returned from rc_read_config was caught wrongly
3023 - Username/password not passed on to radius server due to wrong length.
3025 The presumption is that some radiusclient API changes for 4.51/PH/17
3026 were not taken care of correctly. The code is still untested by me (my
3027 Linux distribution still has 0.3.2 of radiusclient), but it was
3028 contributed by a Radius user.
3030 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3031 expanding the "port" option of the smtp transport.
3033 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3034 while reading a MIME header. Thanks to Tom Hughes for a patch.
3036 PH/24 Include config.h inside local_scan.h so that configuration settings are
3039 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3040 that in an ACL for RCPT (for example), you can examine exactly what was
3043 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3044 commands, but it was not correctly comparing the address with the actual
3045 client host address. Thus, it would show the EHLO address in Received:
3046 header lines when this was not necessary.
3048 PH/27 Added the % operator to ${eval:}.
3050 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3051 it should be ignoring failures (because with -C, for example, it has lost
3052 privilege). It wasn't ignoring creation failures other than "already
3055 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3058 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3059 because an input line was too long, either on its own, or by virtue of
3060 too many continuations, the temporary file was not being removed, and the
3061 return code was incorrect.
3063 PH/31 Missing "BOOL" in function definition in filtertest.c.
3065 PH/32 Applied Sieve patches from the maintainer.
3067 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3069 PH/33 Added "verify = not_blind".
3071 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3072 Local/Makefile (with some defaults set). These are used in built scripts
3073 such as exicyclog, but they have never been used in the exim_install
3074 script (though there are many overriding facilities there). I have
3075 arranged that the exim_install script now takes note of these two
3078 PH/35 Installed configuration files for Dragonfly.
3080 PH/36 When a locally submitted message by a trusted user did not contain a
3081 From: header, and the sender address was obtained from -f or from an SMTP
3082 MAIL command, and the trusted user did not use -F to supply a sender
3083 name, $originator_name was incorrectly used when constructing a From:
3084 header. Furthermore, $originator_name was used for submission mode
3085 messages from external hosts without From: headers in a similar way,
3086 which is clearly wrong.
3088 PH/37 Added control=suppress_local_fixups.
3090 PH/38 When log_selector = +received_sender was set, and the addition of the
3091 sender made the log line's construction buffer exactly full, or one byte
3092 less than full, an overflow happened when the terminating "\n" was
3095 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3096 when the result of a list match is failure because a DNS lookup failed.
3098 PH/40 RM_COMMAND is now used in the building process.
3100 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3101 the "build-* directories that it finds.
3103 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3104 address in a domain literal was a prefix of an interface address.
3106 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3107 when verifying a sender address, unless rewrite_headers is false.
3109 PH/44 Wrote a long comment about why errors_to addresses are verified as
3110 recipients, not senders.
3112 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3113 the ratelimit ACL was added.
3115 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3117 PH/46 Added extra information about PostgreSQL errors to the error string.
3119 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3120 avoiding the need to install special zones in a real server. This is
3121 backwards compatible; if it can't find the fake resolver, it drops back.
3122 Thus, both old and new test suites can be run.
3124 TF/02 Added util/ratelimit.pl
3126 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3127 clock is set back in time.
3129 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3130 Candler <B.Candler@pobox.com>.
3132 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3133 OK for addresses that are the result of redirections.
3135 PH/48 A number of further additions for the benefit of the new test suite,
3136 including a fake gethostbyname() that interfaces to the fake DNS resolver
3139 TF/06 The fix for widen_domains has also been applied to qualify_single and
3140 search_parents which are the other dnslookup options that can cause
3143 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3146 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3148 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3149 service" if it encountered an MX record with an empty target hostname.
3150 The message is now "an MX or SRV record indicated no SMTP service".
3152 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3153 if the DNS of the sending site is misconfigured. This is quite a
3154 common situation. This change restores the behaviour of treating a
3155 helo verification defer as a failure.
3157 PH/51 If self=fail was set on a router, the bounce message did not include the
3158 actual error message.
3164 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3166 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3167 command itself ended in error, the underlying message about the transport
3168 filter timeout was being overwritten with the pipe command error. Now the
3169 underlying error message should be appended to the second error message.
3171 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3173 PH/02 Reduce the amount of output that "make" produces by default. Full output
3174 can still be requested.
3176 PH/03 The warning log line about a condition test deferring for a "warn" verb
3177 was being output only once per connection, rather than after each
3178 occurrence (because it was using the same function as for successful
3179 "warn" verbs). This seems wrong, so I have changed it.
3181 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3182 it should not have, which might have caused a crash in the right
3183 circumstances, but probably never did.
3185 PH/04 Installed a modified version of Tony Finch's patch to make submission
3186 mode fix the return path as well as the Sender: header line, and to
3187 add a /name= option so that you can make the user's friendly name appear
3190 TF/03 Added the control = fakedefer ACL modifier.
3192 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3193 Mark Lowes for thorough testing.
3195 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3197 TK/03 Merged latest SRS patch from Miles Wilton.
3199 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3200 with the definition in sysexits.h (which is #included earlier).
3201 Fortunately, Exim does not actually use EX_OK. The code used to try to
3202 preserve the sysexits.h value, by assumimg that macro definitions were
3203 scanned for macro replacements. I have been disabused of this notion,
3204 so now the code just undefines EX_OK before #including unistd.h.
3206 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3207 in the smtp transport. When a block could not be written in a single
3208 write() function, the timeout was being re-applied to each part-write.
3209 This seems wrong - if the receiver was accepting one byte at a time it
3210 would take for ever. The timeout is now adjusted when this happens. It
3211 doesn't have to be particularly precise.
3213 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3214 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3216 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3217 without a domain if the check to <postmaster@domain> fails.
3219 SC/01 Eximstats: added -xls and the ability to specify output files
3220 (patch written by Frank Heydlauf).
3222 SC/02 Eximstats: use FileHandles for outputing results.
3224 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3226 SC/04 Eximstats: fixed display of large numbers with -nvr option
3228 SC/05 Eximstats: fixed merging of reports with empty tables.
3230 SC/06 Eximstats: added the -include_original_destination flag
3232 SC/07 Eximstats: removed tabs and trailing whitespace.
3234 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3236 TK/06 MBOX spool code: Add real "From " MBOX separator line
3237 so the .eml file is really in mbox format (even though
3238 most programs do not really care). Patch from Alex Miller.
3240 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3241 The latter is generated from $received_to and is only set if the
3242 message has one envelope recipient. SA can use these headers,
3243 obviously out-of-the-box. Patch from Alex Miller.
3245 PH/08 The ${def test on a variable was returning false if the variable's
3246 value was "0", contrary to what the specification has always said!
3247 The result should be true unless the variable is empty.
3249 PH/09 The syntax error of a character other than { following "${if
3250 def:variable_name" (after optional whitespace) was not being diagnosed.
3251 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3252 accidental colon was present, for example, could give incorrect results.
3254 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3255 result is used (not including appendfile, where other changes are about
3258 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3259 This involved changing a lot of size variables from int to off_t. It
3260 should work with maildirs and everything.
3262 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3263 spamd dying while we are connected to it.
3265 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3268 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3269 function for BDB 4.3.
3271 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3273 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3274 the sources for PCRE have been re-arranged and more files are now
3277 PH/15 The code I had for printing potentially long long variables in PH/11
3278 above was not the best (it lost precision). The length of off_t variables
3279 is now inspected at build time, and an appropriate printing format (%ld
3280 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3281 to be "long long int" or "long int". This is needed for the internal
3282 formatting function string_vformat().
3284 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3285 the configuration file to be ":syslog", then the script "guesses" where
3286 the logs files are, rather than using the compiled in default. In our
3287 case the guess is not the same as the compiled default, so the script
3288 suddenly stopped working when I started to use syslog. The patch checks
3289 to see if log_file_path is "". If so, it attempts to read it from exim
3290 with no configuration file to get the compiled in version, before it
3291 falls back to the previous guessing code."
3293 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3294 implementing BATV in an Exim configuration. See NewStuff for the gory
3297 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3298 Makefile that are specific to HP-UX.
3300 PH/18 If the "use_postmaster" option was set for a recipient callout together
3301 with the "random" option, the postmaster address was used as the MAIL
3302 FROM address for the random test, but not for the subsequent recipient
3303 test. It is now used for both.
3305 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3306 patch removes a few documentation additions to RFC 3028, because the
3307 latest draft now contains them. It adds the new en;ascii-case comparator
3308 and a new error check for 8bit text in MIME parts. Comparator and
3309 require names are now matched exactly. I enabled the subaddress
3310 extension, but it is not well tested yet (read: it works for me)."
3312 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3313 rework some of the code of TK/09 above to avoid the hardwired use of
3314 "%lld" and "long long". Replaced the call to snprintf() with a call to
3317 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3318 records point to non-existent hosts", "retry timeout exceeded", and
3319 "retry time not reached for any host after a long failure period".
3321 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3322 experimental DomainKeys support:
3324 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3325 (2) On an error such as an illegally used "control", the wrong name for
3326 the control was given.
3328 These problems did NOT occur unless DomainKeys support was compiled.
3330 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3332 PH/24 Added ${if match_ip condition.
3334 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3335 databases so that it will be absolutely obvious if a crash occurs in the
3336 DB library. This is a regular occurrence (often caused by mis-matched
3339 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3340 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3341 picked up on a user's system that detects such things. There doesn't seem
3342 to be a gcc warning option for this - only an attribute that has to be
3343 put on the function's prototype. It seems that in Fedora Core 4 they have
3344 set this on a number of new functions. No doubt there will be more in due
3347 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3348 specify a transport. However, if an address that was verified by such a
3349 router was the subject of a callout, Exim crashed because it tried to
3350 read the rcpt_include_affixes from the non-existent transport. Now it
3351 just assumes that the setting of that option is false. This bug was
3352 introduced by 4.51/PH/31.
3354 PH/28 Changed -d+all to exclude +memory, because that information is very
3355 rarely of interest, but it makes the output a lot bigger. People tend to
3356 do -d+all out of habit.
3358 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3359 code in os-type was giving problems when libc.so lives in lib64, like on
3362 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3363 aren't the modern standard, and it seems that some systems' include files
3364 don't always have them. Exim was already checking for some of the newer
3365 ones like T_AAAA, and defining it itself. I've added checks for all the
3366 record types that Exim uses.
3368 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3369 not automatically generating a new one, as it is supposed to. This
3370 prevented TLS from working. If the file did exist, but contained invalid
3371 data, a new version was generated, as expected. It was only the case of a
3372 non-existent file that was broken.
3374 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3375 with a change in libdomainkeys > 0.64.
3377 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3378 from DNS. If the selector record carries the flag, it now has
3379 precedence over the domain-wide flag.
3381 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3383 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3384 the use of a port name that isn't defined in /etc/services) occurred, the
3385 message was deferred as in a normal delivery, and thus remained on the
3386 spool, instead of being failed because of the mua_wrapper setting. This
3387 is now fixed, and I tidied up some of the mua_wrapper messages at the
3390 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3391 the messages in a hash of arrays rather than using individual hashes.
3392 This is a bit cleaner and results in dramatic memory savings, albeit
3393 at a slight CPU cost.
3395 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3396 as requested by Marc Sherman.
3398 SC/10 Eximstats: added histograms for user specified patterns as requested
3401 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3403 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3404 fopen() in the content-scanning modules that did not already have it.
3410 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3411 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3413 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3415 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3417 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3418 to test Sieve filters that use "vacation".
3420 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3421 that changes the way the GnuTLS parameters are stored in the cache file.
3422 The new format can be generated externally. For backward compatibility,
3423 if the data in the cache doesn't make sense, Exim assumes it has read an
3424 old-format file, and it generates new data and writes a new file. This
3425 means that you can't go back to an older release without removing the
3428 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3429 work if there are any delivery delays because "one_time" forces the
3430 parent to be marked "delivered", so its unseen clone is never tried
3431 again. For this reason, Exim now forbids the simultaneous setting of
3434 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3435 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3436 entry if you want to know the details. The fix, however, neglected to
3437 consider the case when local delivery batching is involved. The test for
3438 "previously delivered" was not happening when checking to see if an
3439 address could be batched with a previous (undelivered) one; under
3440 certain circumstances this could lead to multiple deliveries to the same
3443 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3444 in its include files, and this causes problems building Exim.
3446 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3447 header_syntax) but Exim was just ignoring anything given after a slash.
3448 In particular, this caused confusion with an attempt to use "verify =
3449 reverse_host_lookup/defer_ok". An error is now given when options are
3450 supplied for verify items that do not have them. (Maybe reverse_host_
3451 lookup should have a defer_ok option, but that's a different point.)
3453 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3454 defined by RFC 821) to 2048, because there were problems with some AUTH
3455 commands, and RFC 1869 says the size should be increased for extended
3456 SMTP commands that take arguments.
3458 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3461 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3462 "unknown" error; now it says that the functionality isn't in the binary.
3464 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3465 an address' error message when a string expansion fails (syntax or
3466 whatever). Otherwise the password may appear in the log. Following change
3467 PH/42 below, there is no longer a chance of it appearing in a bounce
3470 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3472 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3473 of its addresses was ever tried. (Bugzilla bug #2.)
3475 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3476 the result incorrectly in the debug output. (It correctly added a newline
3477 to what was transported.)
3479 TF/01 Added $received_time.
3481 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3482 commented out examples of how to interface to a virus scanner and to
3483 SpamAssassin. Also added commented examples of av_scanner and
3484 spamd_address settings.
3486 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3487 and controls are allowed in which ACLs. There were a couple of minor
3488 errors. Some of the entries in the conditions table (which is a table of
3489 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3490 negation of where the condition IS allowed.
3492 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3494 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3495 header file does not have a version number, so I've had to invent a new
3496 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3497 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3498 radiusclient), but it was contributed by a Radius user.
3500 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3501 files or directories, for OpenSSL.
3503 PH/19 When an Exim process that is running as root has to create an Exim log
3504 file, it does so in a subprocess that runs as exim:exim so as to get the
3505 ownership right at creation (otherwise, other Exim processes might see
3506 the file with the wrong ownership). There was no test for failure of this
3507 fork() call, which would lead to the process getting stuck as it waited
3508 for a non-existent subprocess. Forks do occasionally fail when resources
3509 run out. I reviewed all the other calls to fork(); they all seem to check
3512 PH/20 When checking for unexpected SMTP input at connect time (before writing
3513 the banner), Exim was not dealing correctly with a non-positive return
3514 from the read() function. If the client had disconnected by this time,
3515 the result was a log entry for a synchronization error with an empty
3516 string after "input=" when read() returned zero. If read() returned -1
3517 (an event I could not check), uninitialized data bytes were printed.
3518 There were reports of junk text (parts of files, etc) appearing after
3521 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3523 PH/22 Added support for macro redefinition, and (re)definition in between
3524 driver and ACL definitions.
3526 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3527 forgetting to use the resulting value; it was using the unexpanded value.
3529 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3530 hadn't been configured. The fix is from Juergen Kreileder, who
3531 understands it better than I do:
3533 "Here's what I see happening with three configured cyrus_sasl
3534 authenticators configured (plain, login, cram-md5):
3536 On startup auth_cyrus_sasl_init() gets called for each of these.
3537 This means three calls to sasl_listmech() without a specified mech_list.
3538 => SASL tests which mechs of all available mechs actually work
3539 => three warnings about OTP not working
3540 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3542 With the patch, sasl_listmech() also gets called three times. But now
3543 SASL's mech_list option is set to the server_mech specified in the the
3544 authenticator. Or in other words, the answer from sasl_listmech()
3545 gets limited to just the mech you're testing for (which is different
3547 => the return list contains just 'plain' or 'login', 'cram-md5' or
3548 nothing depending on the value of ob->server_mech.
3550 I've just tested the patch: Authentication still works fine,
3551 unavailable mechs specified in the exim configuration are still
3552 caught, and the auth.log warnings about OTP are gone."
3554 PH/25 When debugging is enabled, the contents of the command line are added
3555 to the debugging output, even when log_selector=+arguments is not
3558 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3559 answer is "GNU", and only if the return is "GNU/something" is the answer
3562 PH/27 $acl_verify_message is now set immediately after the failure of a
3563 verification in an ACL, and so is available in subsequent modifiers. In
3564 particular, the message can be preserved by coding like this:
3566 warn !verify = sender
3567 set acl_m0 = $acl_verify_message
3569 Previously, $acl_verify_message was set only while expanding "message"
3570 and "log_message" when a very denied access.
3572 PH/28 Modified OS/os.c-Linux with
3574 -#ifndef OS_LOAD_AVERAGE
3575 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3577 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3578 nomenclature these days.)
3580 PH/29 Installed patch from the Sieve maintainer that adds the options
3581 sieve_useraddress and sieve_subaddress to the redirect router.
3583 PH/30 In these circumstances:
3584 . Two addresses routed to the same list of hosts;
3585 . First host does not offer TLS;
3586 . First host accepts first address;
3587 . First host gives temporary error to second address;
3588 . Second host offers TLS and a TLS session is established;
3589 . Second host accepts second address.
3590 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3591 and peerdn, if requested) that were in fact used only for the second
3594 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3595 attention to any local part prefix or suffix that was matched by the
3596 router that accepted the address. It now behaves in the same way as it
3597 does for delivery: the affixes are removed from the local part unless
3598 rcpt_include_affixes is set on the transport.
3600 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3601 timeout during the DATA phase of an incoming message.
3603 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3604 applied a patch sanctioned by the Sieve maintainer.
3606 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3607 the uid or gid is negative. A case of a negative gid caused this to be
3608 noticed. The fix allows for either to be negative.
3610 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3611 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3614 PH/36 The change PH/12 above was broken. Fixed it.
3616 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3617 the grounds that routing the same address twice would always produce the
3618 same answer. This might have been true once, but it is certainly no
3619 longer true now. Routing a child address may depend on the previous
3620 routing that produced that child. Some complicated redirection strategies
3621 went wrong when messages had multiple recipients, and made Exim's
3622 behaviour dependent on the order in which the addresses were given.
3624 I have moved the duplicate checking until after the routing is complete.
3625 Exim scans the addresses that are assigned to local and remote
3626 transports, and removes any duplicates. This means that more work will be
3627 done, as duplicates will always all be routed, but duplicates are
3628 presumably rare, so I don't expect this is of any significance.
3630 For deliveries to pipes, files, and autoreplies, the duplicate checking
3631 still happens during the routing process, since they are not going to be
3634 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3635 It corrects a timeout issue with spamd. This is Ian's comment: "The
3636 background is that sometimes spamd either never reads data from a
3637 connection it has accepted, or it never writes response data. The exiscan
3638 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3639 blindly assumes that writes won't block so it may never time out."
3641 PH/39 Allow G after quota size as well as K and M.
3643 PH/40 The value set for $authenticated_id in an authenticator may not contain
3644 binary zeroes or newlines because the value is written to log lines and
3645 to spool files. There was no check on this. Now the value is run through
3646 the string_printing() function so that such characters are converted to
3647 printable escape sequences.
3649 PH/41 $message_linecount is a new variable that contains the total number of
3650 lines in the message. Compare $body_linecount, which is the count for the
3653 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3654 bounce and delay warning messages, except in certain special cases, which
3657 (a) An SMTP error message from a remote host;
3658 (b) A message specified in a :fail: redirection;
3659 (c) A message specified in a "fail" command in a system filter;
3660 (d) A message specified in a FAIL return from the queryprogram router;
3661 (e) A message specified by the cannot_route_message router option.
3663 In these cases only, Exim does include the error details in bounce and
3664 warning messages. There are also a few cases where bland messages such
3665 as "unrouteable address" or "local delivery error" are given.
3667 PH/43 $value is now also set for the "else" part of a ${run expansion.
3669 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3670 being worked on, but at least Exim now implements the latest version to
3673 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3674 process to complete was treated as a delivery failure, a timeout while
3675 writing the message to the pipe was logged, but erroneously treated as a
3676 successful delivery. Such timeouts include transport filter timeouts. For
3677 consistency with the overall process timeout, these timeouts are now
3678 treated as errors, giving rise to delivery failures by default. However,
3679 there is now a new Boolean option for the pipe transport called
3680 timeout_defer, which, if set TRUE, converts the failures into defers for
3681 both kinds of timeout. A transport filter timeout is now identified in
3684 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3685 systems where "make" and "gmake" are different, calling "gmake" at top
3686 level broke things. I've arranged for the value of $(MAKE) to be passed
3687 from the Makefile to this script so that it can call the same version of
3691 A note about Exim versions 4.44 and 4.50
3692 ----------------------------------------
3694 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3695 changes of various kinds. As a consequence, a big documentation update was
3696 needed. This delayed the release for rather longer than seemed good, especially
3697 in the light of a couple of (minor) security issues. Therefore, the changes
3698 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3699 release. So 4.44 and 4.50 are in effect two different branches that both start
3702 I have left the 4.50 change log unchanged; it contains all the changes since
3703 4.43. The change log for 4.44 is below; many of its items are identical to
3704 those for 4.50. This seems to be the most sensible way to preserve the
3705 historical information.
3711 1. Minor wording change to the doc/README.SIEVE file.
3713 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3714 computation of the current number of files was incorrect.
3716 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3717 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3720 4. Give more explanation in the error message when the command for a transport
3721 filter fails to execute.
3723 5. There are several places where Exim runs a non-Exim command in a
3724 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3725 was being done only for the command run by the queryprogram router. It is
3726 now done for all such subprocesses. The other cases are: ${run, transport
3727 filters, and the commands run by the lmtp and pipe transports.
3729 6. Added CONFIGURE_GROUP build-time option.
3731 7. Some older OS have a limit of 256 on the maximum number of file
3732 descriptors. Exim was using setrlimit() to set 1000 as a large value
3733 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3734 systems. I've change it so that if it can't get 1000, it tries for 256.
3736 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3737 was an oversight, and furthermore, ever since the addition of extra
3738 controls (e.g. 4.43/32), the checks on when to allow different forms of
3739 "control" were broken. There should now be diagnostics for all cases when a
3740 control that does not make sense is encountered.
3742 9. Added the /retain_sender option to "control=submission".
3744 10. $recipients is now available in the predata ACL (oversight).
3746 11. Tidy the search cache before the fork to do a delivery from a message
3747 received from the command line. Otherwise the child will trigger a lookup
3748 failure and thereby defer the delivery if it tries to use (for example) a
3749 cached ldap connection that the parent has called unbind on.
3751 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3752 of $address_data from the recipient verification was clobbered by the
3753 sender verification.
3755 13. The value of address_data from a sender verification is now available in
3756 $sender_address_data in subsequent conditions in the ACL statement.
3758 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3760 15. Added a new option "connect=<time>" to callout options, to set a different
3763 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3764 was its contents. (It was OK if the option was not defined at all.)
3766 17. A "Completed" log line is now written for messages that are removed from
3767 the spool by the -Mrm option.
3769 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3770 information about exactly what failed.
3772 19. Added -dd to debug only the daemon process.
3774 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3775 handles timeouts, both on the server side and network timeouts. Renamed the
3776 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3778 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3779 It is now set to "smtps".
3781 22. $host_address is now set to the target address during the checking of
3782 ignore_target_hosts.
3784 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3785 being passed; this would have caused $sender_host_name to have been used if
3786 matching the list had actually called for a host name (not very likely,
3787 since this list is usually IP addresses). A host name is now passed as
3790 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3791 code that specifies a non-restarting handler (typically sigaction() in
3792 modern systems) in an attempt to fix a rare and obscure crash bug.
3794 25. Narrowed the window for a race in the daemon that could cause it to ignore
3795 SIGCHLD signals. This is not a major problem, because they are used only to
3796 wake it up if nothing else does.
3798 26. A malformed maildirsize file could cause Exim to calculate negative values
3799 for the mailbox size or file count. Odd effects could occur as a result.
3800 The maildirsize information is now recalculated if the size or filecount
3803 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3804 support for a long time. Removed HAVE_SYS_VFS_H.
3806 28. Installed the latest version of exipick from John Jetmore.
3808 29. In an address list, if the pattern was not a regular expression, an empty
3809 subject address (from a bounce message) matched only if the pattern was an
3810 empty string. Non-empty patterns were not even tested. This was the wrong
3811 because it is perfectly reasonable to use an empty address as part of a
3812 database query. An empty address is now tested by patterns that are
3813 lookups. However, all the other forms of pattern expect the subject to
3814 contain a local part and a domain, and therefore, for them, an empty
3815 address still always fails if the pattern is not itself empty.
3817 30. Exim went into a mad DNS loop when attempting to do a callout where the
3818 host was specified on an smtp transport, and looking it up yielded more
3819 than one IP address.
3821 31. Re-factored the code for checking spool and log partition space into a
3822 function that finds that data and another that does the check. The former
3823 is then used to implement four new variables: $spool_space, $log_space,
3824 $spool_inodes, and $log_inodes.
3826 32. The RFC2047 encoding function was originally intended for short strings
3827 such as real names; it was not keeping to the 75-character limit for
3828 encoded words that the RFC imposes. It now respects the limit, and
3829 generates multiple encoded words if necessary. To be on the safe side, I
3830 have increased the buffer size for the ${rfc2047: expansion operator from
3833 33. It is now permitted to omit both strings after an "if" condition; if the
3834 condition is true, the result is "true". As before, when the second string
3835 is omitted, a false condition yields an empty string. This makes it less
3836 cumbersome to write custom ACL and router conditions.
3838 34. Failure to deliver a bounce message always caused it to be frozen, even if
3839 there was an errors_to setting on the router. The errors_to setting is now
3842 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3843 canonical form (fully expanded) before being placed in
3844 $sender_host_address.
3846 36. The table in the code that translates DNS record types into text (T_A to
3847 "A" for instance) was missing entries for NS and CNAME. It is just possible
3848 that this could have caused confusion if both these types were looked up
3849 for the same domain, because the text type is used as part of Exim's
3850 per-process caching. But the chance of anyone hitting this buglet seems
3853 37. The dnsdb lookup has been extended in a number of ways.
3855 (1) There is a new type, "zns", which walks up the domain tree until it
3856 finds some nameserver records. It should be used with care.
3858 (2) There is a new type, "mxh", which is like "mx" except that it returns
3859 just the host names, not the priorities.
3861 (3) It is now possible to give a list of domains (or IP addresses) to be
3862 looked up. The behaviour when one of the lookups defers can be
3863 controlled by a keyword.
3865 (4) It is now possible to specify the separator character for use when
3866 multiple records are returned.
3868 38. The dnslists ACL condition has been extended: it is now possible to supply
3869 a list of IP addresses and/or domains to be looked up in a particular DNS
3872 39. Added log_selector=+queue_time_overall.
3874 40. When running the queue in the test harness, wait just a tad after forking a
3875 delivery process, to get repeatability of debugging output.
3877 41. Include certificate and key file names in error message when GnuTLS fails
3878 to set them up, because the GnuTLS error message doesn't include the name
3879 of the failing file when there is a problem reading it.
3881 42. Allow both -bf and -bF in the same test run.
3883 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3885 44. The "Exiscan patch" is now merged into the mainline Exim source.
3887 45. Sometimes the final signoff response after QUIT could fail to get
3888 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3889 < 0 before doing a fflush(). This bug looks as though it goes back to the
3890 introduction of TLS in release 3.20, but "sometimes" must have been rare
3891 because the tests only now provoked it.
3893 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3894 (this can affect the format of dates).
3896 47. exim_tidydb, when checking for the continued existence of a message for
3897 which it has found a message-specific retry record, was not finding
3898 messages that were in split spool directories. Consequently, it was
3899 deleting retry records that should have stayed in existence.
3901 48. Steve fixed some bugs in eximstats.
3903 49. The SPA authentication driver was not abandoning authentication and moving
3904 on to the next authenticator when an expansion was forced to fail,
3905 contradicting the general specification for all authenticators. Instead it
3906 was generating a temporary error. It now behaves as specified.
3908 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3909 (the order specifies the preference for clients). The order is now AES256,
3910 AES128, 3DES, ARCFOUR128.
3912 51. Small patch to Sieve code - explicitly set From: when generating an
3915 52. Exim crashed if a remote delivery caused a very long error message to be
3916 recorded - for instance if somebody sent an entire SpamAssassin report back
3917 as a large number of 550 error lines. This bug was coincidentally fixed by
3918 increasing the size of one of Exim's internal buffers (big_buffer) that
3919 happened as part of the Exiscan merge. However, to be on the safe side, I
3920 have made the code more robust (and fixed the comments that describe what
3923 53. Now that there can be additional text after "Completed" in log lines (if
3924 the queue_time_overall log selector is set), a one-byte patch to exigrep
3925 was needed to allow it to recognize "Completed" as not the last thing in
3928 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3929 patch that reportedly fixes this has been added. I am not expert enough to
3930 create a test for it. This is what the patch creator wrote:
3932 "I found a little strange behaviour of ldap code when working with
3933 Windows 2003 AD Domain, where users was placed in more than one
3934 Organization Units. When I tried to give exim partial DN, the exit code
3935 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3936 But simultaneously result of request was absolutely normal ldap result,
3937 so I produce this patch..."
3939 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3940 REFERENCE, so I have modified the code to exclude the patch when that macro
3943 55. Some experimental protocols are using DNS PTR records for new purposes. The
3944 keys for these records are domain names, not reversed IP addresses. The
3945 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3946 leaves it alone. Component reversal etc. now happens only for IP addresses.
3949 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3951 57. Double the size of the debug message buffer (to 2048) so that more of very
3952 long debug lines gets shown.
3954 58. The exicyclog utility now does better if the number of log files to keep
3955 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3957 59. Two changes related to the smtp_active_hostname option:
3959 (1) $smtp_active_hostname is now available as a variable.
3960 (2) The default for smtp_banner uses $smtp_active_hostname instead
3961 of $primary_hostname.
3963 60. The host_aton() function is supposed to be passed a string that is known
3964 to be a valid IP address. However, in the case of IPv6 addresses, it was
3965 not checking this. This is a hostage to fortune. Exim now panics and dies
3966 if the condition is not met. A case was found where this could be provoked
3967 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3968 components; fortuitously, this particular loophole had already been fixed
3969 by change 4.50/55 above.
3971 If there are any other similar loopholes, the new check in host_aton()
3972 itself should stop them being exploited. The report I received stated that
3973 data on the command line could provoke the exploit when Exim was running as
3974 exim, but did not say which command line option was involved. All I could
3975 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3976 running as the user.
3979 61. There was a buffer overflow vulnerability in the SPA authentication code
3980 (which came originally from the Samba project). I have added a test to the
3981 spa_base64_to_bits() function which I hope fixes it.
3984 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3985 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3987 63. The daemon start-up calls getloadavg() while still root for those OS that
3988 need the first call to be done as root, but it missed one case: when
3989 deliver_queue_load_max is set with deliver_drop_privilege. This is
3990 necessary for the benefit of the queue runner, because there is no re-exec
3991 when deliver_drop_privilege is set.
3993 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3994 This has been fixed.
3996 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3997 was in use, was not putting the data itself into the right store pool;
3998 consequently, it could be overwritten for a subsequent message in the same
3999 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4002 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4004 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4005 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4006 the function were treating the return as a boolean value, which happened to
4007 work because 0=false and not-0=true, but is not correct code.
4009 68. The host_aton() function was not handling scoped IPv6 addresses (those
4010 with, for example, "%eth0" on the end) correctly.
4012 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4013 negated items (that is, ~something) in unsigned ints. Some compilers
4014 apparently mutter when there is no cast.
4016 70. If an address verification called from an ACL failed, and did not produce a
4017 user-specific message (i.e. there was only a "system" message), nothing was
4018 put in $acl_verify_message. In this situation, it now puts the system
4021 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4022 session; change 4.31/43 added the unwanted input to the log line - except
4023 that it did not do this in the start of session case. It now does.
4025 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4026 This is wrong and can cause the other end to generate a synchronization
4027 error if it is another Exim or anything else that does the synchronization
4028 check. A QUIT command is no longer sent after a timeout.
4030 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4031 during host lookups.
4033 74. The defer_ok option of callout verification was not working if it was used
4034 when verifying addresses in header lines, that is, for this case:
4036 verify = header_sender/callout=defer_ok
4038 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4039 those file descriptors could be used for SMTP connections. If anything
4040 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4041 could be sent to the SMTP client, causing chaos. The daemon now opens
4042 stdin, stdout, and stderr to /dev/null when it puts itself into the
4045 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4046 log by default. The user can override this with suitable Perl magic.
4048 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4049 the log message when discard triggers, was not working for the DATA ACL or
4050 for the non-SMTP ACL.
4052 78. Error message wording change in sieve.c.
4054 79. If smtp_accept_max_per_host was set, the number of connections could be
4055 restricted to fewer than expected, because the daemon was trying to set up
4056 a new connection before checking whether the processes handling previous
4057 connections had finished. The check for completed processes is now done
4058 earlier. On busy systems, this bug wouldn't be noticed because something
4059 else would have woken the daemon, and it would have reaped the completed
4062 80. If a message was submitted locally by a user whose login name contained one
4063 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4064 It caused a spool format error. I have fixed the spool reading code. A
4065 related problem was that the "from" clause in the Received: line became
4066 illegal because of the space(s). It is now covered by ${quote_local_part.
4068 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4071 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4073 83. Patch from Sieve maintainer to fix unterminated string problem in
4074 "vacation" handling.
4076 84. Some minor changes to the Linux configuration files to help with other
4077 OS variants using glibc.
4079 85. One more patch for Sieve to update vacation handling to latest spec.
4082 ----------------------------------------------------
4083 See the note above about the 4.44 and 4.50 releases.
4084 ----------------------------------------------------
4090 1. Change 4.43/35 introduced a bug that caused file counts to be
4091 incorrectly computed when quota_filecount was set in an appendfile
4094 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4095 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4098 3. Give more explanation in the error message when the command for a transport
4099 filter fails to execute.
4101 4. There are several places where Exim runs a non-Exim command in a
4102 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4103 was being done only for the command run by the queryprogram router. It is
4104 now done for all such subprocesses. The other cases are: ${run, transport
4105 filters, and the commands run by the lmtp and pipe transports.
4107 5. Some older OS have a limit of 256 on the maximum number of file
4108 descriptors. Exim was using setrlimit() to set 1000 as a large value
4109 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4110 systems. I've change it so that if it can't get 1000, it tries for 256.
4112 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4113 was an oversight, and furthermore, ever since the addition of extra
4114 controls (e.g. 4.43/32), the checks on when to allow different forms of
4115 "control" were broken. There should now be diagnostics for all cases when a
4116 control that does not make sense is encountered.
4118 7. $recipients is now available in the predata ACL (oversight).
4120 8. Tidy the search cache before the fork to do a delivery from a message
4121 received from the command line. Otherwise the child will trigger a lookup
4122 failure and thereby defer the delivery if it tries to use (for example) a
4123 cached ldap connection that the parent has called unbind on.
4125 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4126 of $address_data from the recipient verification was clobbered by the
4127 sender verification.
4129 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4130 was its contents. (It was OK if the option was not defined at all.)
4132 11. A "Completed" log line is now written for messages that are removed from
4133 the spool by the -Mrm option.
4135 12. $host_address is now set to the target address during the checking of
4136 ignore_target_hosts.
4138 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4139 being passed; this would have caused $sender_host_name to have been used if
4140 matching the list had actually called for a host name (not very likely,
4141 since this list is usually IP addresses). A host name is now passed as
4144 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4145 code that specifies a non-restarting handler (typically sigaction() in
4146 modern systems) in an attempt to fix a rare and obscure crash bug.
4148 15. Narrowed the window for a race in the daemon that could cause it to ignore
4149 SIGCHLD signals. This is not a major problem, because they are used only to
4150 wake it up if nothing else does.
4152 16. A malformed maildirsize file could cause Exim to calculate negative values
4153 for the mailbox size or file count. Odd effects could occur as a result.
4154 The maildirsize information is now recalculated if the size or filecount
4157 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4158 support for a long time. Removed HAVE_SYS_VFS_H.
4160 18. Updated exipick to current release from John Jetmore.
4162 19. Allow an empty sender to be matched against a lookup in an address list.
4163 Previously the only cases considered were a regular expression, or an
4166 20. Exim went into a mad DNS lookup loop when doing a callout where the
4167 host was specified on the transport, if the DNS lookup yielded more than
4170 21. The RFC2047 encoding function was originally intended for short strings
4171 such as real names; it was not keeping to the 75-character limit for
4172 encoded words that the RFC imposes. It now respects the limit, and
4173 generates multiple encoded words if necessary. To be on the safe side, I
4174 have increased the buffer size for the ${rfc2047: expansion operator from
4177 22. Failure to deliver a bounce message always caused it to be frozen, even if
4178 there was an errors_to setting on the router. The errors_to setting is now
4181 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4182 canonical form (fully expanded) before being placed in
4183 $sender_host_address.
4185 24. Updated eximstats to version 1.33
4187 25. Include certificate and key file names in error message when GnuTLS fails
4188 to set them up, because the GnuTLS error message doesn't include the name
4189 of the failing file when there is a problem reading it.
4191 26. Expand error message when OpenSSL has problems setting up cert/key files.
4194 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4195 (this can affect the format of dates).
4197 28. exim_tidydb, when checking for the continued existence of a message for
4198 which it has found a message-specific retry record, was not finding
4199 messages that were in split spool directories. Consequently, it was
4200 deleting retry records that should have stayed in existence.
4202 29. eximstats updated to version 1.35
4203 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4204 1.35 - bugfix such that pie charts by volume are generated correctly
4206 30. The SPA authentication driver was not abandoning authentication and moving
4207 on to the next authenticator when an expansion was forced to fail,
4208 contradicting the general specification for all authenticators. Instead it
4209 was generating a temporary error. It now behaves as specified.
4211 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4212 (the order specifies the preference for clients). The order is now AES256,
4213 AES128, 3DES, ARCFOUR128.
4215 31. Small patch to Sieve code - explicitly set From: when generating an
4218 32. Exim crashed if a remote delivery caused a very long error message to be
4219 recorded - for instance if somebody sent an entire SpamAssassin report back
4220 as a large number of 550 error lines. This bug was coincidentally fixed by
4221 increasing the size of one of Exim's internal buffers (big_buffer) that
4222 happened as part of the Exiscan merge. However, to be on the safe side, I
4223 have made the code more robust (and fixed the comments that describe what
4226 33. Some experimental protocols are using DNS PTR records for new purposes. The
4227 keys for these records are domain names, not reversed IP addresses. The
4228 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4229 leaves it alone. Component reversal etc. now happens only for IP addresses.
4232 34. The host_aton() function is supposed to be passed a string that is known
4233 to be a valid IP address. However, in the case of IPv6 addresses, it was
4234 not checking this. This is a hostage to fortune. Exim now panics and dies
4235 if the condition is not met. A case was found where this could be provoked
4236 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4237 components; fortuitously, this particular loophole had already been fixed
4238 by change 4.50/55 or 4.44/33 above.
4240 If there are any other similar loopholes, the new check in host_aton()
4241 itself should stop them being exploited. The report I received stated that
4242 data on the command line could provoke the exploit when Exim was running as
4243 exim, but did not say which command line option was involved. All I could
4244 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4245 running as the user.
4248 35. There was a buffer overflow vulnerability in the SPA authentication code
4249 (which came originally from the Samba project). I have added a test to the
4250 spa_base64_to_bits() function which I hope fixes it.
4253 36. The daemon start-up calls getloadavg() while still root for those OS that
4254 need the first call to be done as root, but it missed one case: when
4255 deliver_queue_load_max is set with deliver_drop_privilege. This is
4256 necessary for the benefit of the queue runner, because there is no re-exec
4257 when deliver_drop_privilege is set.
4259 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4260 was in use, was not putting the data itself into the right store pool;
4261 consequently, it could be overwritten for a subsequent message in the same
4262 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4265 38. Sometimes the final signoff response after QUIT could fail to get
4266 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4267 < 0 before doing a fflush(). This bug looks as though it goes back to the
4268 introduction of TLS in release 3.20, but "sometimes" must have been rare
4269 because the tests only now provoked it.
4275 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4276 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4277 Now it returns an integer. A number of calls were still expecting a T/F
4278 return. Fortuitously, in all cases, the tests worked in OK situations,
4279 which is the norm. However, things would have gone wrong on any write
4280 failures on the smtp file descriptor. This function is used when sending
4281 messages over SMTP and also when doing verify callouts.
4283 2. When Exim is called to do synchronous delivery of a locally submitted
4284 message (the -odf or -odi options), it no longer closes stderr before doing
4287 3. Implemented the mua_wrapper option.
4289 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4291 5. Implemented the functions header_remove(), header_testname(),
4292 header_add_at_position(), and receive_remove_recipient(), and exported them
4295 6. If an ACL "warn" statement specified the addition of headers, Exim already
4296 inserted X-ACL-Warn: at the start if there was no header name. However, it
4297 was not making this test for the second and subsequent header lines if
4298 there were newlines in the string. This meant that an invalid header could
4299 be inserted if Exim was badly configured.
4301 7. Allow an ACL "warn" statement to add header lines at the start or after all
4302 the Received: headers, as well as at the end.
4304 8. Added the rcpt_4xx retry error code.
4306 9. Added postmaster_mailfrom=xxx to callout verification option.
4308 10. Added mailfrom=xxxx to the callout verification option, for verify=
4311 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4312 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4314 12. Inserted some casts to stop certain compilers warning when using pointer
4315 differences as field lengths or precisions in printf-type calls (mostly
4316 affecting debugging statements).
4318 13. Added optional readline() support for -be (dynamically loaded).
4320 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4321 same clock tick as a message's arrival, so that its received time was the
4322 same as the "first fail" time on the retry record, and that message
4323 remained on the queue past the ultimate address timeout, every queue runner
4324 would try a delivery (because it was past the ultimate address timeout) but
4325 after another failure, the ultimate address timeout, which should have then
4326 bounced the address, did not kick in. This was a "< instead of <=" error;
4327 in most cases the first failure would have been in the next clock tick
4328 after the received time, and all would be well.
4330 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4331 being recognized when the domain list was tested by the match_domain
4332 condition in an expansion string.
4334 16. Added the ${str2b64: operator.
4336 17. Exim was always calling setrlimit() to set a large limit for the number of
4337 processes, without checking whether the existing limit was already
4338 adequate. (It did check for the limit on file descriptors.) Furthermore,
4339 errors from getrlimit() and setrlimit() were being ignored. Now they are
4340 logged to the main and panic logs, but Exim does carry on, to try to do its
4341 job under whatever limits there are.
4343 18. Imported PCRE 5.0.
4345 19. Trivial typo in log message " temporarily refused connection" (the leading
4348 20. If the log selector return_path_on_delivery was set and an address was
4349 redirected to /dev/null, the delivery process crashed because it assumed
4350 that a return path would always be set for a "successful" delivery. In this
4351 case, the whole delivery is bypassed as an optimization, and therefore no
4354 21. Internal re-arrangement: the function for sending a challenge and reading
4355 a response while authentication was assuming a zero-terminated challenge
4356 string. It's now changed to take a pointer and a length, to allow for
4357 binary data in such strings.
4359 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4361 23. Exim was not respecting finduser_retries when seeking the login of the
4362 uid under which it was called; it was always trying 10 times. (The default
4363 setting of finduser_retries is zero.) Also, it was sleeping after the final
4364 failure, which is pointless.
4366 24. Implemented tls_on_connect_ports.
4368 25. Implemented acl_smtp_predata.
4370 26. If the domain in control=submission is set empty, Exim assumes that the
4371 authenticated id is a complete email address when it generates From: or
4372 Sender: header lines.
4374 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4375 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4376 chown and chgrp in /bin and hostname in /usr/bin.
4378 28. Exim was keeping the "process log" file open after each use, just as it
4379 does for the main log. This opens the possibility of it remaining open for
4380 long periods when the USR1 signal hits a daemon. Occasional processlog
4381 errors were reported, that could have been caused by this. Anyway, it seems
4382 much more sensible not to leave this file open at all, so that is what now
4385 29. The long-running daemon process does not normally write to the log once it
4386 has entered its main loop, and it closes the log before doing so. This is
4387 so that log files can straightforwardly be renamed and moved. However,
4388 there are a couple of unusual error situations where the daemon does write
4389 log entries, and I had neglected to close the log afterwards.
4391 30. The text of an SMTP error response that was received during a remote
4392 delivery was being truncated at 512 bytes. This is too short for some of
4393 the long messages that one sometimes sees. I've increased the limit to
4396 31. It is now possible to make retry rules that apply only when a message has a
4397 specific sender, in particular, an empty sender.
4399 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4400 it possible to be selective about when SMTP synchronization is enforced.
4402 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4404 32. Implemented hosts_connection_nolog.
4406 33. Added an ACL for QUIT.
4408 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4411 35. Added mailbox_size and mailbox_filecount to appendfile.
4413 36. Added control = no_multiline_responses to ACLs.
4415 37. There was a bug in the logic of the code that waits for the clock to tick
4416 in the case where the clock went backwards by a substantial amount such
4417 that the microsecond fraction of "now" was more than the microsecond
4418 fraction of "then" (but the whole seconds number was less).
4420 38. Added support for the libradius Radius client library this is found on
4421 FreeBSD (previously only the radiusclient library was supported).
4427 1. When certain lookups returned multiple values in the form name=value, the
4428 quoting of the values was not always being done properly. Specifically:
4429 (a) If the value started with a double quote, but contained no whitespace,
4431 (b) If the value contained whitespace other than a space character (i.e.
4432 tabs or newlines or carriage returns) it was not quoted.
4433 This fix has been applied to the mysql and pgsql lookups by writing a
4434 separate quoting function and calling it from the lookup code. The fix
4435 should probably also be applied to nisplus, ibase and oracle lookups, but
4436 since I cannot test any of those, I have not disturbed their existing code.
4438 2. A hit in the callout cache for a specific address caused a log line with no
4439 reason for rejecting RCPT. Now it says "Previous (cached) callout
4440 verification failure".
4442 3. There was an off-by-one bug in the queryprogram router. An over-long
4443 return line was truncated at 256 instead of 255 characters, thereby
4444 overflowing its buffer with the terminating zero. As well as fixing this, I
4445 have increased the buffer size to 1024 (and made a note to document this).
4447 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4448 when Exim is waiting for an SMTP response from a remote server, Exim
4449 restarts its select() call on the socket, thereby resetting its timeout.
4450 This is not a problem when such interrupts are rare. Somebody set up a cron
4451 job to run exiwhat every 2 minutes, which is less than the normal select()
4452 timeout (5 or 10 minutes). This meant that the select() timeout never
4453 kicked in because it was always reset. I have fixed this by comparing the
4454 time when an interrupt arrives with the time at the start of the first call
4455 to select(). If more time than the timeout has elapsed, the interrupt is
4456 treated as a timeout.
4458 5. Some internal re-factoring in preparation for the addition of Sieve
4459 extensions (by MH). In particular, the "personal" test is moved to a
4460 separate function, and given an option for scanning Cc: and Bcc: (which is
4461 not set for Exim filters).
4463 6. When Exim created an email address using the login of the caller as the
4464 local part (e.g. when creating a From: or Sender: header line), it was not
4465 quoting the local part when it contained special characters such as @.
4467 7. Installed new OpenBSD configuration files.
4469 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4470 try to make them clearer.
4472 9. Callout options, other than the timeout value, were being ignored when
4473 verifying sender addresses in header lines. For example, when using
4475 verify = header_sender/callout=no_cache
4477 the cache was (incorrectly) being used.
4479 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4480 only the Cygwin environment.
4482 11. When return_path_on_delivery was set as a log selector, if different remote
4483 addresses in the same message used different return paths and parallel
4484 remote delivery occurred, the wrong values would sometimes be logged.
4485 (Whenever a remote delivery process finished, the return path value from
4486 the most recently started remote delivery process was logged.)
4488 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4489 lines when AUTH and/or TLS are in use. This is the "received protocol"
4490 field. Exim used to use "asmtp" for authenticated SMTP, without any
4491 indication (in the protocol name) for TLS use. Now it follows the RFC and
4492 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4493 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4494 names appear in log lines as well as in Received: header lines.
4496 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4497 extensions, and comparison tests, and to fix some bugs.
4499 14. Changes to the "personal" filter test:
4501 (1) The test was buggy in that it was just doing the equivalent of
4502 "contains" tests on header lines. For example, if a user's address was
4503 anne@some.where, the "personal" test would incorrectly be true for
4505 To: susanne@some.where
4507 This test is now done by extracting each address from the header in turn,
4508 and checking the entire address. Other tests that are part of "personal"
4509 are now done using regular expressions (for example, to check local parts
4510 of addresses in From: header lines).
4512 (2) The list of non-personal local parts in From: addresses has been
4513 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4514 taken from the Sieve specification recommendations.
4516 (3) If the message contains any header line starting with "List-" it is
4517 treated as non-personal.
4519 (4) The test for "circular" in the Subject: header line has been removed
4520 because it now seems ill-conceived.
4522 15. Minor typos in src/EDITME comments corrected.
4524 16. Installed latest exipick from John Jetmore.
4526 17. If headers_add on a router specified a text string that was too long for
4527 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4528 of string_sprintf() is now avoided.
4530 18. $message_body_size was not set (it was always zero) when running the DATA
4531 ACL and the local_scan() function.
4533 19. For the "mail" command in an Exim filter, no default was being set for
4534 the once_repeat time, causing a random time value to be used if "once" was
4535 specified. (If the value happened to be <= 0, no repeat happened.) The
4536 default is now 0s, meaning "never repeat". The "vacation" command was OK
4537 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4538 (I found it when inspecting the code).
4540 20. There is now an overall timeout for performing a callout verification. It
4541 defaults to 4 times the callout timeout, which applies to individual SMTP
4542 commands during the callout. The overall timeout applies when there is more
4543 than one host that can be tried. The timeout is checked before trying the
4544 next host. This prevents very long delays if there are a large number of
4545 hosts and all are timing out (e.g. when the network connections are timing
4546 out). The value of the overall timeout can be changed by specifying an
4547 additional sub-option for "callout", called "maxwait". For example:
4549 verify = sender/callout=5s,maxwait=20s
4551 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4552 to the end before writing, but this should make it even safer).
4554 22. Exim was forgetting that it had advertised PIPELINING for the second and
4555 subsequent messages on an SMTP connection. It was also not resetting its
4556 memory on STARTTLS and an internal HELO.
4558 23. When Exim logs an SMTP synchronization error within a session, it now
4559 records whether PIPELINING has been advertised or not.
4561 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4562 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4563 rather than long int.
4565 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4567 26. Added the never_mail option to autoreply.
4573 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4574 crash if the getsockname() call failed; this can happen if a connection is
4575 closed very soon after it is established. The problem was simply in the
4576 order in which certain operations were done, causing Exim to try to write
4577 to the SMTP stream before it had set up the file descriptor. The bug has
4578 been fixed by making things happen in the correct order.
4584 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4585 before the connection was closed, thus losing the rejection response.
4587 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4588 some early Solaris releases, but causes trouble in current releases where
4589 socklen_t is defined.
4591 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4594 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4597 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4598 that releasing the top of it at the end releases what was used for sub-
4599 expansions (unless the block got too big). However, discard this block if
4600 the first thing is a variable or header, so that we can use its block when
4601 it is dynamic (useful for very large $message_headers, for example).
4603 6. Lookups now cache *every* query, not just the most recent. A new, separate
4604 store pool is used for this. It can be recovered when all lookup caches are
4605 flushed. Lookups now release memory at the end of their result strings.
4606 This has involved some general refactoring of the lookup sources.
4608 7. Some code has been added to the store_xxx() functions to reduce the amount
4609 of flapping under certain conditions.
4611 8. log_incoming_interface used to affect only the <= reception log lines. Now
4612 it causes the local interface and port to be added to several more SMTP log
4613 lines, for example "SMTP connection from", and rejection lines.
4615 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4617 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4619 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4620 could be overwritten at the end of the current message (or the start of a
4621 new message if it was set in a HELO ACL). The value is now preserved for
4622 the duration of the SMTP connection.
4624 12. If a transport had a headers_rewrite setting, and a matching header line
4625 contained an unqualified address, that address was qualified, even if it
4626 did not match any rewriting rules. The underlying bug was that the values
4627 of the flags that permit the existence of unqualified sender and recipient
4628 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4629 non-local messages, and by -bnq for local messages) were not being
4630 preserved with the message after it was received.
4632 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4633 "next input=" as part of the text comprising the host identity instead of
4634 the correct text. The code was using the same buffer for two different
4635 strings. However, depending on which order the printing function evaluated
4636 its arguments, the bug did not always show up. Under Linux, for example, my
4637 test suite worked just fine.
4639 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4640 doesn't work with some older versions of Perl. It has been changed to "my",
4641 which in any case is probably the better facility to use.
4643 15. A really picky compiler found some instances of statements for creating
4644 error messages that either had too many or two few arguments for the format
4647 16. The size of the buffer for calls to the DNS resolver has been increased
4648 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4649 for addresses that have a lot of PTR records. This alleviates a problem; it
4650 does not fully solve it.
4652 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4653 buffer now truncates the list and logs the incident, which is the same
4654 action as happens when Exim is looking up a host name and its aliases.
4655 Previously in this situation something unpredictable would happen;
4656 sometimes it was "internal error: store_reset failed".
4658 18. If a server dropped the connection unexpectedly when an Exim client was
4659 using GnuTLS and trying to read a response, the client delivery process
4660 crashed while trying to generate an error log message.
4662 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4663 string, for example:
4665 warn message = H1: something\nH2: something
4667 the text was added as a single header line from Exim's point of view
4668 though it ended up OK in the delivered message. However, searching for the
4669 second and subsequent header lines using $h_h2: did not work. This has been
4670 fixed. Similarly, if a system filter added multiple headers in this way,
4671 the routers could not see them.
4673 20. Expanded the error message when iplsearch is called with an invalid key to
4674 suggest using net-iplsearch in a host list.
4676 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4677 ACLs are no longer actually imposed (and a message to that effect is
4680 22. If a "gecos" field in a passwd entry contained escaped characters, in
4681 particular, if it contained a \" sequence, Exim got it wrong when building
4682 a From: or a Sender: header from that name. A second bug also caused
4683 incorrect handling when an unquoted " was present following a character
4684 that needed quoting.
4686 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4687 was not being matched caselessly.
4689 24. Arranged for all hyphens in the exim.8 source to be escaped with
4692 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4693 properly. Recipient callout cache records were still being keyed to include
4694 the sender, even when use_sender was set false. This led to far more
4695 callouts that were necessary. The sender is no longer included in the key
4696 when use_sender is false.
4698 26. Added "control = submission" modifier to ACLs.
4700 27. Added the ${base62d: operator to decode base 62 numbers.
4702 28. dnsdb lookups can now access SRV records.
4704 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4705 the configuration file.
4707 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4708 (-v) mode. This makes the output for a verbose queue run more intelligible.
4710 31. Added a use_postmaster feature to recipient callouts.
4712 32. Added the $body_zerocount variable, containing the number of binary zero
4713 bytes in the message body.
4715 33. The time of last modification of the "new" subdirectory is now used as the
4716 "mailbox time last read" when there is a quota error for a maildir
4719 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4721 35. Added +ignore_unknown as a special item in host lists.
4723 36. Code for decoding IPv6 addresses in host lists is now included, even if
4724 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4725 address was recognized as an IP address, but was then not correctly decoded
4726 into binary, causing unexpected and incorrect effects when compared with
4733 1. Very minor rewording of debugging text in manualroute to say "list of
4734 hosts" instead of "hostlist".
4736 2. If verify=header_syntax was set, and a header line with an unqualified
4737 address (no domain) and a large number of spaces between the end of the
4738 name and the colon was received, the reception process suffered a buffer
4739 overflow, and (when I tested it) crashed. This was caused by some obsolete
4740 code that should have been removed. The fix is to remove it!
4742 3. When running in the test harness, delay a bit after writing a bounce
4743 message to get a bit more predictability in the log output.
4745 4. Added a call to search_tidyup() just before forking a reception process. In
4746 theory, someone could use a lookup in the expansion of smtp_accept_max_
4747 per_host which, without the tidyup, could leave open a database connection.
4749 5. Added the variables $recipient_data and $sender_data which get set from a
4750 lookup success in an ACL "recipients" or "senders" condition, or a router
4751 "senders" option, similar to $domain_data and $local_part_data.
4753 6. Moved the writing of debug_print from before to after the "senders" test
4756 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4757 problems for message scanning, either using a data ACL, or using
4758 local_scan() because the Received: header was not generated till after they
4759 were called (in order to set the time as the time of reception completion).
4760 I have revised the way this works. The header is now generated after the
4761 body is received, but before the ACL or local_scan() are called. After they
4762 are run, the timestamp in the header is updated.
4768 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4769 before starting a queue runner without re-exec. This happened only when
4770 deliver_drop_privilege was set or when the Exim user was set to root. The
4771 effect of the bug was that timeouts during subsequent deliveries caused
4772 crashes instead of being properly handled. The handler is now left at its
4773 default (and expected) setting.
4775 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4776 message, again when deliver_drop_privilege is set or Exim is run as root.
4777 The bug described in (1) was not present in this case, but the tidying up
4778 of the other signals was missing. I have made the two cases consistent.
4780 3. The ignore_target_hosts setting on a manualroute router was being ignored
4781 for hosts that were looked up using the /MX notation.
4783 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4786 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4787 operated on the sender address. After changing the $sender_address to <>
4788 for the sender address verify, Exim was re-instated it as the original
4789 (before rewriting) address, but remembering that it had rewritten it, so it
4790 wasn't rewriting it again. This bug also had the effect of breaking the
4791 sender address verification caching when the sender address was rewritten.
4793 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4794 This has been changed so that if the ip literal address matches
4795 ignore_target_hosts, the router declines.
4797 7. Added expansion conditions match_domain, match_address, and match_local_
4798 part (NOT match_host).
4800 8. The placeholder for the Received: header didn't have a length field set.
4802 9. Added code to Exim itself and to exim_lock to test for a specific race
4803 condition that could lead to file corruption when using MBX delivery. The
4804 issue is with the lockfile that is created in /tmp. If this file is removed
4805 after a process has opened it but before that process has acquired a lock,
4806 there is the potential for a second process to recreate the file and also
4807 acquire a lock. This could lead to two Exim processes writing to the file
4808 at the same time. The added code performs the same test as UW imapd; it
4809 checks after acquiring the lock that its file descriptor still refers to
4810 the same named file.
4812 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4813 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4816 11. Added the smtp_active_hostname option. If used, this will typically be made
4817 to depend on the incoming interface address. Because $interface_address is
4818 not set up until the daemon has forked a reception process, error responses
4819 that can happen earlier (such as "too many connections") no longer contain
4822 12. If an expansion in a condition on a "warn" statement fails because a lookup
4823 defers, the "warn" statement is abandoned, and the next ACL statement is
4824 processed. Previously this caused the whole ACL to be aborted.
4826 13. Added the iplsearch lookup type.
4828 14. Added ident_timeout as a log selector.
4830 15. Added tls_certificate_verified as a log selector.
4832 16. Added a global option tls_require_ciphers (compare the smtp transport
4833 option of the same name). This controls incoming TLS connections.
4835 17. I finally figured out how to make tls_require_ciphers do a similar thing
4836 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4837 before starting the TLS session.
4839 18. Tabs are now shown as \t in -bP output.
4841 19. If the log selector return_path_on_delivery was set, Exim crashed when
4842 bouncing a message because it had too many Received: header lines.
4844 20. If two routers both had headers_remove settings, and the first one included
4845 a superfluous trailing colon, the final name in the first list and the
4846 first name in the second list were incorrectly joined into one item (with a
4847 colon in the middle).
4853 1. Added -C and -D options to the exinext utility, mainly to make it easier
4854 to include in the automated testing, but these could be helpful when
4855 multiple configurations are in use.
4857 2. The exinext utility was not formatting the output nicely when there was
4858 an alternate port involved in the retry record key, nor when there was a
4859 message id as well (for retries that were specific to a specific message
4860 and a specific host). It was also confused by IPv6 addresses, because of
4861 the additional colons they contain. I have fixed the IPv4 problem, and
4862 patched it up to do a reasonable job for IPv6.
4864 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4865 delivery, the log line now contains "pipelined" if PIPELINING was used.
4867 4. An SMTP transport process used to panic and die if the bind() call to set
4868 an explicit outgoing interface failed. This has been changed; it is now
4869 treated in the same way as a connect() failure.
4871 5. A reference to $sender_host_name in the part of a conditional expansion
4872 that was being skipped was still causing a DNS lookup. This no longer
4875 6. The def: expansion condition was not recognizing references to header lines
4876 that used bh_ and bheader_.
4878 7. Added the _cache feature to named lists.
4880 8. The code for checking quota_filecount in the appendfile transport was
4881 allowing one more file than it should have been.
4883 9. For compatibility with Sendmail, the command line option
4891 and sets the incoming protocol and host name (for trusted callers). The
4892 host name and its colon can be omitted when only the protocol is to be set.
4893 Note the Exim already has two private options, -pd and -ps, that refer to
4894 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4895 "s", but I don't think that's a major issue.
4897 10. A number of refactoring changes to the code, none of which should affect
4900 (a) The number of logging options was getting close to filling up the
4901 32-bit word that was used as a bit map. I have split them into two classes:
4902 those that are passed in the argument to log_write(), and those that are
4903 only ever tested independently outside of that function. These are now in
4904 separate 32-bit words, so there is plenty of room for expansion again.
4905 There is no change in the user interface or the logging behaviour.
4907 (b) When building, for example, log lines, the code previously used a
4908 macro that called string_cat() twice, in order to add two strings. This is
4909 not really sufficiently general. Furthermore, there was one instance where
4910 it was actually wrong because one of the argument was used twice, and in
4911 one call a function was used. (As it happened, calling the function twice
4912 did not affect the overall behaviour.) The macro has been replaced by a
4913 function that can join an arbitrary number of extra strings onto a growing
4916 (c) The code for expansion conditions now uses a table and a binary chop
4917 instead of a serial search (which was left over from when there were very
4918 few conditions). Also, it now recognizes conditions like "pam" even when
4919 the relevant support is not compiled in: a suitably worded error message is
4920 given if an attempt is made to use such a condition.
4922 11. Added ${time_interval:xxxxx}.
4924 12. A bug was causing one of the ddress fields not to be passed back correctly
4925 from remote delivery subprocesses. The field in question was not being
4926 subsequently used, so this caused to problems in practice.
4928 13. Added new log selectors queue_time and deliver_time.
4930 14. Might have fixed a bug in maildirsizefile handling that threw up
4931 "unexpected character" debug warnings, and recalculated the data
4932 unnecessarily. In any case, I expanded the warning message to give more
4935 15. Added the message "Restricted characters in address" to the statements in
4936 the default ACL that block characters like @ and % in local parts.
4938 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4939 Three changes have been made:
4941 (a) There was a serious bug; a negative response to MAIL caused the whole
4942 recipient domain to be cached as invalid, thereby blocking all messages
4943 to all local parts at the same domain, from all senders. This bug has
4944 been fixed. The domain is no longer cached after a negative response to
4945 MAIL if the sender used is not empty.
4947 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4950 (c) A new callout option, "use_sender" has been added for people who want
4951 the modified behaviour.
4957 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4960 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4961 indeed breaks things for older releases.
4963 3. Added additional logging to the case where there is a problem reading data
4964 from a filter that is running in a subprocess using a pipe, in order to
4965 try to track down a specific problem.
4967 4. Testing facility fudge: when running in the test harness and attempting
4968 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4969 getting "No route to host". Convert this to a timeout.
4971 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4974 6. Some OS don't have socklen_t but use size_t instead. This affects the
4975 fifth argument of getsockopt() amongst other things. This is now
4976 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4977 can be set for individual OS. I have set it for SunOS5, OSF1, and
4978 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4979 some earlier ones do not.
4981 7. Change 4.30/15 was not doing the test caselessly.
4983 8. The standard form for an IPv6 address literal was being rejected by address
4984 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4985 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4986 this, as well as the form without the "IPv6" on the front (but only when
4987 address literals are enabled, of course).
4989 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4991 10. Exim crashed if a message with an empty sender address specified by -f
4992 encountered a router with an errors_to setting. This could be provoked only
4993 by a command such as
4997 where an empty string was supplied; "<>" did not hit this bug.
4999 11. Installed PCRE release 4.5.
5001 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5002 remained set. It is now erased.
5004 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5005 times from message ids (which are base 36 rather than the normal 62).
5007 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5008 were being counted as actual protocol errors, and logged if the log
5009 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5010 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5011 and DATA following a set of rejected RCPTs do not count as protocol errors.
5012 In other words, Exim assumes they were pipelined, though this may not
5013 actually be the case. Of course, in all cases the client gets an
5014 appropriate error code.
5016 15. If a lookup fails in an ACL condition, a message about the failure may
5017 be available; it is used if testing the ACL cannot continue, because most
5018 such messages specify what the cause of the deferral is. However, some
5019 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5020 that caused an old message to be retained and used if a later statement
5021 caused a defer, replacing the real cause of the deferral.
5023 16. If an IP address had so many PTR records that the DNS lookup buffer
5024 was not large enough to hold them, Exim could crash while trying to process
5025 the truncated data. It now detects and logs this case.
5027 17. Further to 4.21/58, another change has been made: if (and only if) the
5028 first line of a message (the first header line) ends with CRLF, a bare LF
5029 in a subsequent header line has a space inserted after it, so as not to
5030 terminate the header.
5032 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5033 unnecessarily, used atoi() instead of strtol(), and didn't check the
5034 termination when getting file sizes from file names by regex.
5036 19. Completely re-implemented the support for maildirsize files, in the light
5037 of a number of problems with the previous contributed implementation
5038 (4.30/29). In particular:
5040 . If the quota is zero, the maildirsize file is maintained, but no quota is
5043 . If the maildir directory does not exist, it is created before any attempt
5044 to write a maildirsize file.
5046 . The quota value in the file is just a cache; if the quota is changed in
5047 the transport, the new value overrides.
5049 . A regular expression is available for excluding directories from the
5052 20. The autoreply transport checks the characters in options that define the
5053 message's headers; it allows continued headers, but it was checking with
5054 isspace() after an embedded newline instead of explicitly looking for a
5057 21. If all the "regular" hosts to which an address was routed had passed their
5058 expiry times, and had not reached their retry times, the address was
5059 bounced, even if fallback hosts were defined. Now Exim should go on to try
5062 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5063 equivalent code in the SMTP transport. Some hosts send humungous responses
5064 to HELO/EHLO, more than 1024 it seems.
5066 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5067 gives compiler warnings in some environments. I've now done it "properly",
5070 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5071 (because of problems with the built-in one) was declared to return uschar *
5072 instead of char *, causing compiler failure.
5074 25. Fixed a file descriptor leak when processing alias/forward files.
5076 26. Fixed a minor format string issue in dbfn.c.
5078 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5080 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5081 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5084 29. When a sender address is verified, it is cached, to save repeating the test
5085 when there is more than one recipient in a message. However, when the
5086 verification involves a callout, it is possible for different callout
5087 options to be set for different recipients. It is too complicated to keep
5088 track of this in the cache, so now Exim always runs a verification when a
5089 callout is required, relying on the callout cache for the optimization.
5090 The overhead is duplication of the address routing, but this should not be
5093 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5094 to be verified with callout=postmaster, and the main callout worked but the
5095 postmaster check failed, the verification correctly failed. However, if a
5096 subsequent RCPT command asked for sender verification *without* the
5097 postmaster check, incorrect caching caused this verification also to fail,
5100 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5101 it was not caching the DNS options (qualify_single, search_parents) that
5102 were used when the lookup failed. A subsequent lookup with different
5103 options therefore always gave the same answer, though there were cases
5104 where it should not have. (Example: a "domains = !$mx_any" option on a
5105 dnslookup router: the "domains" option is always processed without any
5106 widening, but the router might have qualify_single set.) Now Exim uses the
5107 cached value only when the same options are set.
5109 32. Added John Jetmore's "exipick" utility to the distribution.
5111 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5112 than a timeout (e.g. a certificate is required, and is not provided), an
5113 Exim server now closes the connection immediately. Previously it waited for
5114 the client to close - but if the client is SSL, it seems that they each
5115 wait for each other, leading to a delay before one of them times out.
5117 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5118 maintained 0.8.x compatibility because I don't think many are using it, and
5119 it is clearly obsolete.
5121 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5124 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5125 client certificate was expired. A simple patch fixes this, though I don't
5126 understand the full logic of why the verify callback is called multiple
5129 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5130 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5131 which causes problems with some clients (such as the Certicom SSL Plus
5132 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5133 disables the coutermeasure allowing Eudora to connect."
5135 38. Exim was not checking that a write() to a log file succeeded. This could
5136 lead to Bad Things if a log got too big, in particular if it hit a file
5137 size limit. Exim now panics and dies if it cannot write to a log file, just
5138 as it does if it cannot open a log file.
5140 39. Modified OS/Makefile-Linux so that it now contains
5142 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5144 The two -D definitions ensure that Exim is compiled with large file
5145 support, which makes it possible to handle log files that are bigger than
5148 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5149 instance) a domain was checked against a named list that involved a lookup,
5150 causing $domain_data to be set, then another domain was checked against the
5151 same list, then the first domain was re-checked, the value of $domain_data
5152 after the final check could be wrong. In particular, if the second check
5153 failed, it could be set empty. This bug probably also applied to
5156 41. The strip_trailing_dot option was not being applied to the address given
5157 with the -f command-line option.
5159 42. The code for reading a message's header from the spool was incrementing
5160 $received_count, but never initializing it. This meant that the value was
5161 incorrect (doubled) while delivering a message in the same process in which
5162 it was received. In the most common configuration of Exim, this never
5163 happens - a fresh exec is done - but it can happen when
5164 deliver_drop_privilege is set.
5166 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5167 it now includes up to 150 characters of the unexpected data in the log
5170 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5171 and building data strings. The size of both of these buffers was 10 000
5172 bytes - far larger than anybody would *ever* want, thought I. Needless to
5173 say, somebody hit the limit. I have increased the maximum line length to
5174 20 000 and the maximum data length of concatenated lines to 100 000. I have
5175 also fixed two bugs, because there was no checking on these buffers. Tsk,
5176 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5177 buffer is too small.
5179 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5180 lsearch lookups. Now it does.
5182 46. When parsing a route_list item in a manualroute router, a fixed-length
5183 buffer was used for the list of hosts. I made this 1024 bytes long,
5184 thinking that nobody would ever have a list of hosts that long. Wrong.
5185 Somebody had a whole pile of complicated expansion conditions, and the
5186 string was silently truncated, leading to an expansion error. It turns out
5187 that it is easier to change to an unlimited length (owing to other changes
5188 that have happened since this code was originally written) than to build
5189 structure for giving a limitation error. The length of the item that
5190 expands into the list of hosts is now unlimited.
5192 47. The lsearch lookup could not handle data where the length of text line was
5193 more than 4095 characters. Such lines were truncated, leading to shortened
5194 data being returned. It should now handle lines of any length.
5196 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5197 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5200 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5201 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5202 no longer happens in comment lines. A list of replacements is now placed
5203 at the head of all of the source files, except those whose only change is
5204 to replace PERL_COMMAND in the very first #! line.
5206 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5207 messages on the queue, with a bottom-up merge sort, using code contributed
5208 by Michael Haardt. This should make operations like -bp somewhat faster on
5209 large queues. It won't affect queue runners, except when queue_run_in_order
5212 51. Installed eximstats 1.31 in the distribution.
5214 52. Added support for SRV lookups to the dnslookup router.
5216 53. If an ACL referred to $message_body or $message_body_end, the value was not
5217 reset for any messages that followed in the same SMTP session.
5219 54. The store-handling optimization for building very long strings was not
5220 differentiating between the different store pools. I don't think this
5221 actually made any difference in practice, but I've tidied it.
5223 55. While running the routers to verify a sender address, $sender_address
5224 was still set to the sender address. This is wrong, because when routing to
5225 send a bounce to the sender, it would be empty. Therefore, I have changed
5226 it so that, while verifying a sender address, $sender_address is set to <>.
5227 (There is no change to what happens when verifying a recipient address.)
5229 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5230 target A or AAAA records (if not already returned) without resetting the
5231 qualify_single or search_parents options of the DNS resolver. These are
5232 inappropriate in this case because the targets of MX and SRV records must
5233 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5234 target that, when qualified, matched something in the local domain. These
5235 two options are now turned off when doing these lookups.
5237 57. It seems that at least some releases of Reiserfs (which does not have the
5238 concept of a fixed number of inodes) returns zero and not -1 for the
5239 number of available inodes. This interacted badly with check_spool_inodes,
5240 which assumed that -1 was the "no such thing" setting. What I have done is
5241 to check that the total number of inodes is greater than zero before doing
5242 the test of how many are available.
5244 58. When a "warn" ACL statement has a log_message modifier, the message is
5245 remembered, and not repeated. This is to avoid a lot of repetition when a
5246 message has many recipients that cause the same warning to be written.
5247 Howewer, Exim was preserving the list of already written lines for an
5248 entire SMTP session, which doesn't seem right. The memory is now reset if a
5249 new message is started.
5251 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5252 debugging output unless log_rewrite was also set.
5254 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5255 of (int)(handle) when we know that handle contains (void *)(-1).
5257 61. The Exim daemon panic-logs an error return when it closes the incoming
5258 connection. However "connection reset by peer" seems to be common, and
5259 isn't really an error worthy of noting specially, so that particular error
5262 62. When Exim is trying to find all the local interfaces, it used to panic and
5263 die if the ioctl to get the interface flags failed. However, it seems that
5264 on at least one OS (Solaris 9) it is possible to have an interface that is
5265 included in the list of interfaces, but for which you get a failure error
5266 for this call. This happens when the interface is not "plumbed" into a
5267 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5268 failure of the "get flags" call assumes that the interface is down.
5270 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5271 makes life easier for people who are doing arithmetic on fields extracted
5272 from dates, where you often get leading zeros that should not be
5273 interpreted as octal.
5275 64. Added qualify_domain to the redirect router, to override the global
5278 65. If a pathologically long header line contained very many addresses (the
5279 report of this problem mentioned 10 000) and each of them was rewritten,
5280 Exim could use up a very large amount of memory. (It kept on making new
5281 copies of the header line as it rewrote, and never released the old ones.)
5282 At the expense of a bit more processing, the header rewriting function has
5283 been changed so that it no longer eats memory in this way.
5285 66. The generation of the Received: header has been moved from the time that a
5286 message starts to be received, to the time that it finishes. The timestamp
5287 in the Received: header should now be very close to that of the <= log
5288 line. There are two side-effects of this change:
5290 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5291 logged header lines no longer include the local Received: line, because
5292 it has not yet been created. The same applies to a copy of the message
5293 that is returned to a non-SMTP sender when a message is rejected.
5295 (b) When a filter file is tested using -bf, no additional Received: header
5296 is added to the test message. After some thought, I decided that this
5299 This change does not affect the value of $received_for. It is still set
5300 after address rewriting, but before local_scan() is called.
5302 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5304 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5305 gave an unhelpful panic error message, and a defer error. I have managed to
5306 change this behaviour so that it now rejects any supplied certificate,
5307 which seems right, as the list of acceptable certificates is empty.
5309 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5310 gave an unhelpful defer error. I have not managed to make this reject any
5311 supplied certificates, but the error message it gives is "no certificate
5312 supplied", which is not helpful.
5314 70. exigrep's output now also includes lines that are not associated with any
5315 message, but which match the given pattern. Implemented by a patch from
5316 Martin Sluka, which also tidied up the Perl a bit.
5318 71. Recipient callout verification, like sender verification, was using <> in
5319 the MAIL FROM command. This isn't really the right thing, since the actual
5320 sender may affect whether the remote host accepts the recipient or not. I
5321 have changed it to use the actual sender in the callout; this means that
5322 the cache record is now keyed on a recipient/sender pair, not just the
5323 recipient address. There doesn't seem to be a real danger of callout loops,
5324 since a callout by the remote host to check the sender would use <>.
5325 [SEE ABOVE: changed after hitting problems.]
5327 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5328 temporary errors. However, in the case of such a code being given after
5329 the end of a data transmission (i.e. after ".") Exim was failing to write
5330 a retry record for the message. (Yes, there was some broken host that was
5331 actually sending 8xx at this point.)
5333 73. An unknown lookup type in a host list could cause Exim to panic-die when
5334 the list was checked. (An example that provoked this was putting <; in the
5335 middle of a list instead of at the start.) If this happened during a DATA
5336 ACL check, a -D file could be left lying around. This kind of configuration
5337 error no longer causes Exim to die; instead it causes a defer errror. The
5338 incident is still logged to the main and panic logs.
5340 74. Buglet left over from Exim 3 conversion. The message "too many messages
5341 in one connection" was written to the rejectlog but not the mainlog, except
5342 when address rewriting (yes!) was being logged.
5344 75. Added write_rejectlog option.
5346 76. When a system filter was run not as root (that is, when system_filter_user
5347 was set), the values of the $n variables were not being returned to the
5348 main process; thus, they were not subsequently available in the $sn
5351 77. Added +return_path_on_delivery log selector.
5353 78. A connection timeout was being treated differently from recipients deferred
5354 when testing hosts_max_try with a message that was older than the host's
5355 retry timeout. (The host should not be counted, thus allowing all hosts to
5356 be tried at least once before bouncing.) This may have been the cause of an
5357 occasionally reported bug whereby a message would remain on the queue
5358 longer than the retry timeout, but would be bounced if a delivery was
5359 forced. I say "may" because I never totally pinned down the problem;
5360 setting up timeout/retry tests is difficult. See also the next item.
5362 79. The ultimate address timeout was not being applied to errors that involved
5363 a combination of host plus message (for example, a timeout on a MAIL
5364 command). When an address resolved to a number of possible hosts, and they
5365 were not all tried for each delivery (e.g. because of hosts_max_try), a
5366 message could remain on the queue longer than the retry timeout.
5368 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5371 81. Fixed an obscure SMTP outgoing bug which required at least the following
5372 conditions: (a) there was another message waiting for the same server;
5373 (b) the server returned 5xx to all RCPT commands in the first message so
5374 that the message was not completed; (c) the server dropped the connection
5375 or gave a negative response to the RSET that Exim sends to abort the
5376 transaction. The observed case was a dropped connection after DATA that had
5377 been sent in pipelining mode. That is, the server had advertised PIPELINING
5378 but was not implementing it correctly. The effect of the bug was incorrect
5379 behaviour, such as trying another host, and this could lead to a crash.
5385 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5386 and daemon.c were passed as pointers to ints; they should have been
5387 pointers to socklen_t variables (which are typically unsigned ints).
5389 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5392 3. Fixed a really odd bug that affected only the testing scheme; patching a
5393 certain fixed string in the binary changed the value of another string that
5394 happened to be identical to the end of the original first string.
5396 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5397 name", it returns that address as the IP address. On some operating
5398 systems (e.g. Solaris), it also passes back the IP address string as the
5399 "host name". However, on others (e.g. Linux), it passes back an empty
5400 string. Exim wasn't checking for this, and was changing the host name to an
5401 empty string, assuming it had been canonicized.
5403 5. Although rare, it is permitted to have more than one PTR record for a given
5404 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5405 all the names associated with an address, because they do in Solaris.
5406 However, it seems that they do not in Linux for data that comes from the
5407 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5408 I found this out when I moved to a new Linux workstation and tried to run
5409 the Exim test suite.
5411 To get round this problem I have changed the code so that it now does its
5412 own call to the DNS to look up PTR records when searching for a host name.
5413 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5414 addresses that are only in /etc/hosts are still found.
5416 This behaviour is, however, controlled by an option called host_lookup_
5417 order, which defaults to "bydns:byaddr". If people want to use the other
5418 order, or indeed, just use one or the other means of lookup, they can
5419 specify it in this variable.
5421 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5422 some operating systems, this comes back from gethostbyaddr() as an empty
5423 string, and this is what Exim used to test for. However, it seems that in
5424 other systems, "." is yielded. Exim now tests for this case too.
5426 7. The values of check_spool_space and check_log_space are now held internally
5427 as a number of kilobytes instead of an absolute number of bytes. If a
5428 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5429 kilobyte. This means that much larger values can be stored.
5431 8. Exim monitor: an attempt to get the action menu when not actually pointing
5432 at a message produces an empty menu entitled "No message selected". This
5433 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5434 no entries in it ("Shell widget menu has zero width and/or height"). So I
5435 have added a single, blank menu entry in this case.
5437 9. Added ${quote_local_part.
5439 10. MIME decoding is now applied to the contents of Subject: header lines when
5442 11. Now that a reference to $sender_host_address automatically causes a reverse
5443 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5444 host lookup before query-style lookups in lists that might use this
5445 variable. This has therefore been abolished, and the "net-" prefix is no
5446 longer necessary for query-style lookups.
5448 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5449 have been a typo for LFLAGS, so it has been changed.
5451 13. The install script calls Exim with "-C /dev/null" in order to find the
5452 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5453 to be output. Howeve, since Exim outputs its version number before the
5454 error, it didn't break the script. It just looked ugly. I fixed this by
5455 always allowing "-C /dev/null" if the caller is root.
5457 14. Ignore overlarge ACL variable number when reading spool file - insurance
5458 against a later release with more variables having written the file.
5460 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5461 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5462 this, as well as the form without the "IPv6" on the front.
5464 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5465 OS/Makefile-Darwin file.
5467 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5468 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5470 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5471 variable. However, some OS use ints for inodes. I've added cast to long int
5472 to get rid of the compiler warning.
5474 19. I had forgotten to lock out "/../" in configuration file names when
5475 ALT_CONFIG_PREFIX was set.
5477 20. Routers used for verification do not need to specify transports. However,
5478 if such a router generated a host list, and callout was configured, Exim
5479 crashed, because it could not find a port number from the (non-existent)
5480 transport. It now assumes port 25 in this circumstance.
5482 21. Added the -t option to exigrep.
5484 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5485 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5486 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5487 them is set, LOOKUP_LSEARCH is forced.
5489 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5491 24. Added sender and host information to the "rejected by local_scan()" log
5492 line; previously there was no indication of these.
5494 25. Added .include_if_exists.
5496 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5497 Linux. It turns out that not all file systems support this. Apparently some
5498 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5499 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5500 not supported on the file descriptor, is now ignored when Exim is trying to
5501 sync a directory. This applies only to Linux.
5503 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5505 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5506 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5509 29. Added support for maildirsize files from supplied patch (modified a bit).
5511 30. The use of :fail: followed by an empty string could lead Exim to respond to
5512 sender verification failures with (e.g.):
5514 550 Verification failed for <xxx>
5515 550 Sender verify failed
5517 where the first response line was missing the '-' that indicates it is not
5518 the final line of the response.
5520 31. The loop for finding the name of the user that called Exim had a hardwired
5521 limit of 10; it now uses the value of finduser_retries, which is used for
5522 all other user lookups.
5524 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5527 33. Exim was neglecting to zero errno before one call of strtol() when
5528 expanding a string and expecting an integer value. On some systems this
5529 resulted in spurious "integer overflow" errors. Also, it was casting the
5530 result into an int without checking.
5532 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5533 did not work. The code looks as if it has *never* worked, though it appears
5534 to have been documented since at least releast 1.62. I have made it work.
5536 35. The "timeout_DNS" error in retry rules, also documented since at least
5537 1.62, also never worked. As it isn't clear exactly what this means, and
5538 clearly it isn't a major issue, I have abolished the feature by treating it
5539 as "timeout", and writing a warning to the main and panic logs.
5541 36. The display of retry rules for -brt wasn't always showing the error code
5544 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5545 timeout_connect_A, timeout_connect_MX.
5547 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5548 to the empty sender.
5550 39. The daemon was not analysing the content of -oX till after it had closed
5551 stderr and disconnected from the controlling terminal. This meant that any
5552 syntax errors were only noted on the panic log, and the return code from
5553 the command was 0. By re-arranging the code a little, I've made the
5554 decoding happen first, so such errors now appear on stderr, and the return
5555 code is 1. However, the actual setting up of the sockets still happens in
5556 the disconnected process, so errors there are still only recorded on the
5559 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5560 connections (as happens on some IP stacks) was logged at start up time as
5561 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5562 it from "IPv6 and IPv4", which means that two separate sockets are being
5565 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5566 says whether AF_INET or AF_INET6 was passed as an argument.
5568 42. Exiwhat output was messed up when time zones were included in log
5571 43. Exiwhat now gives more information about the daemon's listening ports,
5572 and whether -tls-on-connect was used.
5574 44. The "port" option of the smtp transport is now expanded.
5576 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5577 silently ignored. Now an error message is written to the main and panic
5580 46. There's a new ACL modifier called "logwrite" which writes to a log file
5581 as soon as it is encountered.
5583 47. Added $local_user_uid and $local_user_gid at routing time.
5585 48. Exim crashed when trying to verify a sender address that was being
5588 49. Exim was recognizing only a space character after ".include". It now also
5589 recognizes a tab character.
5591 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5592 extracting the relevant information from the specification. The man page no
5593 longer contains scrambled data for the -d option, and I've added a section
5594 at the front about calling Exim under different names.
5596 51. Added "extra_headers" argument to the "mail" command in filter files.
5598 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5601 53. Installed eximstats 1.29.
5603 54. Added transport_filter_timeout as a generic transport option.
5605 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5606 Cc: header lines. This was required by RFC 822, but it not required by RFC
5609 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5610 incoming messages that did not have them. Now it does so only if the
5611 message originates locally, that is, if there is no associated remote host
5612 address. When Resent- header lines are present, this applies to the Resent-
5613 lines rather than the non-Resent- lines.
5615 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5616 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5618 58. Messages for configuration errors now include the name of the main
5619 configuration file - useful now that there may be more than one file in a
5620 list (.included file names were always shown).
5622 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5623 for those rare installations that do not start the daemon as root or run it
5624 setuid root. I've cut out the call to initgroups() if the daemon is not
5627 60. The Exim user and group can now be bound into the binary as text strings
5628 that are looked up at the start of Exim's processing.
5630 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5632 62. Added $mailstore_basename variable.
5634 63. Installed patch to sieve.c from Michael Haardt.
5636 64. When Exim failed to open the panic log after failing to open the main log,
5637 the original message it was trying to log was written to stderr and debug
5638 output, but if they were not available (the usual case in production), it
5639 was lost. Now it is written to syslog before the two lines that record the
5640 failures to open the logs.
5642 65. Users' Exim filters run in subprocesses under the user's uid. It is
5643 possible for a "deliver" command or an alias in a "personal" command to
5644 provoke an address rewrite. If logging of address rewriting is configured,
5645 this fails because the process is not running as root or exim. There may be
5646 a better way of dealing with this, but for the moment (because 4.30 needs
5647 to be released), I have disabled address rewrite logging when running a
5648 filter in a non-root, non-exim process.
5654 1. The buildconfig auxiliary program wasn't quoting the value set for
5655 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5656 not defined. This bug was masked in 4.22 by the effect that was fixed in
5659 2. Some messages that were rejected after a message id was allocated were
5660 shown as "incomplete" by exigrep. It no longer does this for messages that
5661 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5663 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5664 have allow_domain_literals set, the ID did not get logged in the <= line.
5665 Domain literals are now always recognized in Message-ID: header lines.
5667 4. The first argument for a ${extract expansion item is the key name or field
5668 number. Leading and trailing spaces in this item were not being ignored,
5669 causing some misleading effects.
5671 5. When deliver_drop_privilege was set, single queue runner processes started
5672 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5673 same command in the process it spins off) were not dropping privilege.
5675 6. When the daemon running as "exim" started a queue runner, it always
5676 re-executed Exim in the spun-off process. This is a waste of effort when
5677 deliver_drop_privilege is set. The new process now just calls the
5678 queue-runner function directly.
5684 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5687 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5688 ignored. Though the use of -oP was forcing the writing of a pid file, it
5689 was always written to the default place.
5691 3. If the message "no IP address found for host xxxx" is generated during
5692 incoming verification, it is now followed by identification of the incoming
5693 connection (so you can more easily find what provoked it).
5695 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5697 5. Added some features to "harden" Exim a bit more against certain attacks:
5699 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5700 be put in Local/Makefile. This is like the never_users runtime option,
5701 but it cannot be overridden. The default setting is "root".
5703 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5704 prefix string with which any file named in a -C command line option
5707 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5708 is retained for -C and -D only if the caller of Exim is root. Without
5709 it, the exim user may also use -C and -D and retain privilege.
5711 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5712 command line option is disabled.
5714 6. Macro names set by the -D option must start with an upper case letter, just
5715 like macro names defined in the configuration file.
5717 7. Added "dereference=" facility to LDAP.
5719 8. Two instances of the typo "uknown" in the source files are fixed.
5721 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5722 the Configure-Makefile script screwed up while processing it.
5724 10. Incorporated PCRE 4.4.
5726 11. The SMTP synchronization check was not operating right at the start of an
5727 SMTP session. For example, it could not catch a HELO sent before the client
5728 waited for the greeting. There is now a check for outstanding input at the
5729 point when the greeting is written. Because of the duplex, asynchronous
5730 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5731 way, but not yet received, when the check is performed.
5733 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5734 on TCP/IP sockets, because this apparently causes some broken clients to
5737 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5738 unchanged) from the Cygwin maintainer.
5740 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5741 was supported instead of testing for mbx. Effectively a typo.
5743 15. The spa authenticator server code was not checking that the input it
5744 received was valid base64.
5746 16. The debug output line for the "set" modifier in ACLs was not showing the
5747 name of the variable that was being set.
5749 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5751 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5752 lookup on its own. Something else was needed to trigger the lookup. For
5753 example, a match in host_lookup or the need for a host name in a host list.
5754 Now, if $sender_host_name is referenced and the host name has not yet been
5755 looked up, a lookup is performed. If the lookup fails, the variable remains
5756 empty, and $host_lookup_failed is set to "1".
5758 19. Added "eqi" as a case-independent comparison operator.
5760 20. The saslauthd authentication condition could segfault if neither service
5761 nor realm was specified.
5763 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5764 error message that was logged was misleading, and incoming SMTP
5765 connections were dropped. The message is now more accurate, and temporary
5766 errors are given to SMTP connections.
5768 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5769 (except RSET) with a 421 error, until QUIT is received. However, it was
5770 failing to send a response to QUIT.
5772 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5773 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5774 ACL is now run after the helo_try_verify_hosts code.
5776 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5777 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5778 case-independent, so other case variants are also recognized). Apparently
5779 some systems use these upper case variants.
5781 25. If more than two messages were waiting for the same host, and a transport
5782 filter was specified for the transport, Exim sent two messages over the
5783 same TCP/IP connection, and then failed with "socket operation on non-
5784 socket" when it tried to send the third.
5786 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5788 27. The extern definition of crypt16() in expand.c was not being excluded when
5789 the OS had its own crypt16() function.
5791 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5792 as a preferred synonym for return_size_limit, both as an option and as an
5795 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5797 30. Changed the default configuration ACL to relax the local part checking rule
5798 for addresses that are not in any local domains. For these addresses,
5799 slashes and pipe symbols are allowed within local parts, but the sequence
5800 /../ is explicitly forbidden.
5802 31. SPA server authentication was not clearing the challenge buffer before
5805 32. log_message in a "warn" ACL statement was writing to the reject log as
5806 well as to the main log, which contradicts the documentation and doesn't
5807 seem right (because no rejection is happening). So I have stopped it.
5809 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5810 However, I am unable to do any testing of this.
5812 34. Fixed an infelicity in the appendfile transport. When checking directories
5813 for a mailbox, to see if any needed to be created, it was accidentally
5814 using path names with one or more superfluous leading slashes; tracing
5815 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5817 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5818 contents are added to the log line that is written for every discarded
5819 recipient. (Previously a log_message setting was ignored.)
5821 36. The ${quote: operator now quotes the string if it is empty.
5823 37. The install script runs exim in order to find its version number. If for
5824 some reason other than non-existence or emptiness, which it checks, it
5825 could not run './exim', it was installing it with an empty version number,
5826 i.e. as "exim-". This error state is now caught, and the installation is
5829 38. An argument was missing from the function that creates an error message
5830 when Exim fails to connect to the socket for saslauthd authentication.
5831 This could cause Exim to crash, or give a corrupted message.
5833 39. Added isip, isip4, and isip6 to ${if conditions.
5835 40. The ACL variables $acl_xx are now saved with the message, and can be
5836 accessed later in routers, transports, and filters.
5838 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5839 strings in the file are not string-expanded.
5841 42. If a MAIL command specified a SIZE value that was too large to fit into an
5842 int variable, the check against message_size_limit failed. Such values are
5843 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5844 day this will have to be increased, but I don't think I want to be around
5845 when emails are that large.
5852 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5853 iconv() is not standard in FreeBSD.
5855 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5856 IPv6 enabled. The observed symptom was a segmentation fault on return from
5857 the function os_common_find_running_interfaces() in src/os.c.
5859 3. In the check_special_case() function in daemon.c I had used "errno" as an
5860 argument name, which causes warnings on some systems. This was basically a
5861 typo, since it was named "eno" in the comments!
5863 4. The code that waits for the clock to tick (at a resolution of some fraction
5864 of a second) so as to ensure message-id uniqueness was always waiting for
5865 at least one whole tick, when it could have waited for less. [This is
5866 almost certainly not relevant at current processor speeds, where it is
5867 unlikely to ever wait at all. But we try to future-proof.]
5869 5. The function that sleeps for a time interval that includes fractions of a
5870 second contained a race. It did not block SIGALRM between setting the
5871 timer, and suspending (a couple of lines later). If the interval was short
5872 and the sigsuspend() was delayed until after it had expired, the suspension
5873 never ended. On busy systems this could lead to processes getting stuck for
5876 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5877 process, before it forks any delivery processes. The open lookup caching
5878 mechanism meant that the open file or database connection was passed into
5879 the delivery process. The problem was that delivery processes always tidy
5880 up cached lookup data. This could cause a problem for the next delivery
5881 process started by the queue runner, because the external queue runner
5882 process does not know about the closure. So the next delivery process
5883 still has data in the lookup cache. In the case of a file lookup, there was
5884 no problem because closing a file descriptor in a subprocess doesn't affect
5885 the parent. However, if the lookup was caching a connection to a database,
5886 the connection was closed, and the second delivery process was likely to
5887 see errors such as "PGSQL: query failed: server closed the connection
5888 unexpectedly". The problem has been fixed by closing all cached lookups
5889 in a queue runner before running a delivery process.
5891 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5892 seem to have the "const" qualifier which it has on other OS. I've
5895 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5896 *of the same type* (client or server) with the same public name that an
5897 error should be diagnosed.
5899 9. When Exim looked up a host name for an IP address, but failed to find the
5900 original IP address when looking up the host name (a safety check), it
5901 output the message "<ip address> does not match any IP for NULL", which was
5902 confusing, to say the least. The bug was that the host name should have
5903 appeared instead of "NULL".
5905 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5906 user that is built into the binary, and the -C or -D options is used, root
5907 privilege is dropped before the configuration file is read. In addition,
5908 logging is switched to stderr instead of the normal log files. If the
5909 configuration then re-defines the Exim user, the unprivileged environment
5910 is probably not what is expected, so Exim logs a panic warning message (but
5913 However, if deliver_drop_privilege is set, the unprivileged state may well
5914 be exactly what is intended, so the warning has been cut out in that case,
5915 and Exim is allowed to try to write to its normal log files.
5921 1. smtp_return_error_details was not giving details for temporary sender
5922 or receiver verification errors.
5924 2. Diagnose a configuration error if two authenticators have the same public
5927 3. Exim used not to create the message log file for a message until the first
5928 delivery attempt. This could be confusing when incoming messages were held
5929 for policy or load reasons. The message log file is now created at the time
5930 the message is received, and an initial "Received" line is written to it.
5932 4. The automatically generated man page for command line options had a minor
5933 bug that caused no ill effects; however, a more serious problem was that
5934 the procedure for building the man page automatically didn't always
5935 operate. Consequently, release 4.20 contains an out-of-date version. This
5936 shouldn't happen again.
5938 5. When building Exim with embedded Perl support, the script that builds the
5939 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5940 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5942 6. The freeze_tell option was not being used for messages that were frozen on
5943 arrival, either by an ACL or by local_scan().
5945 7. Added the smtp_incomplete_transaction log selector.
5947 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5948 was accepting AUTH without a new EHLO.
5950 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5951 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5954 10. Logging of TCP/IP connections (when configured) now happens in the main
5955 daemon process instead of the child process, so that the TCP/IP connection
5956 count is more accurate (but it can never be perfect).
5958 11. The use of "drop" in a nested ACL was not being handled correctly in the
5959 outer ACL. Now, if condition failure induced by the nested "drop" causes
5960 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5961 or "require"), the connection is dropped.
5963 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5964 that yield "discard" can now be used with an "accept" or a "discard" verb,
5965 but an error is generated for any others (because I can't see a useful way
5966 to define what should happen).
5968 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5969 are now processed in the same way as lines in the Exim configuration file.
5970 In particular, continuation lines are supported.
5972 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5974 15. Added -ti meaning -t -i.
5976 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5977 domains, and warn by logging if others are found.
5979 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5980 returned from GIFCONF when Exim is trying to find the list of interfaces on
5981 a host. The code in os.c has been modified to copy each ifreq to an aligned
5982 structure in all cases.
5984 Also, in some cases, the returned ifreq's were being copied to a 'struct
5985 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5986 means the last couple of bytes of an IPv6 address could be chopped if the
5987 ifreq contained only a normal sockaddr (14 bytes storage).
5989 18. Named domain lists were not supported in the hosts_treat_as_local option.
5990 An entry such as +xxxx was not recognized, and was treated as a literal
5993 19. Ensure that header lines added by a DATA ACL are included in the reject log
5994 if the ACL subsequently rejects the message.
5996 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5997 MD5 (which is deprecated).
5999 21. When testing a filter file using -bf, Exim was writing a message when it
6000 took the sender from a "From " line in the message, but it was not doing so
6001 when it took $return_path from a Return-Path: header line. It now does.
6003 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6004 with a valid header line field name (a series of printing characters
6005 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6007 23. Changed "disc" in the source to "disk" to conform to the documentation and
6008 the book and for uniformity.
6010 24. Ignore Sendmail's -Ooption=value command line item.
6012 25. When execve() failed while trying to run a command in a pipe transport,
6013 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6014 could be confused with a return value of 69 from the command itself. This
6015 has been changed to 127, the value the shell returns if it is asked to run
6016 a non-existent command. The wording for the related log line suggests a
6017 non-existent command as the problem.
6019 26. If received_header_text expands to an empty string, do not add a Received:
6020 header line to the message. (Well, it adds a token one on the spool, but
6021 marks it "old" so that it doesn't get used or transmitted.)
6023 27. Installed eximstats 1.28 (addition of -nt option).
6025 28. There was no check for failure on the call to getsockname() in the daemon
6026 code. This can fail if there is a shortage of resources on the system, with
6027 ENOMEM, for example. A temporary error is now given on failure.
6029 29. Contrary to the C standard, it seems that in some environments, the
6030 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6031 program. Exim now does this explicitly; it affects the formatting of
6032 timestamps using strftime().
6034 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6035 complaints. I've now initialized all the variables, to avoid this.
6037 32. Header lines added by a system filter were not being "seen" during
6038 transport-time rewrites.
6040 33. The info_callback() function passed to OpenSSL is set up with type void
6041 (*)(SSL *, int, int), as described somewhere. However, when calling the
6042 function (actually a macro) that sets it up, the type void(*)() is
6043 expected. I've put in a cast to prevent warnings from picky compilers.
6045 34. If a DNS black list lookup found a CNAME record, but there were no A
6046 records associated with the domain it pointed at, Exim crashed.
6048 35. If a DNS black list lookup returned more than one A record, Exim ignored
6049 all but the first. It now scans all returned addresses if a particular IP
6050 value is being sought. In this situation, the contents of the
6051 $dnslist_value variable are a list of all the addresses, separated by a
6054 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6055 to accept a host name and all its aliases if the forward lookup for any of
6056 them yielded the IP address of the incoming connection. Now it accepts only
6057 those names whose forward lookup yields the correct IP address. Any other
6058 names are discarded. This closes a loophole whereby a rogue DNS
6059 administrator could create reverse DNS records to break through a
6060 wildcarded host restriction in an ACL.
6062 37. If a user filter or a system filter that ran in a subprocess used any of
6063 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6064 the wrong values were passed to the pipe command ($thisaddress had the
6065 value of $0, $0 had the value of $1, etc). This bug was introduced by
6066 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6068 38. Improved the line breaking for long SMTP error messages from ACLs.
6069 Previously, if there was no break point between 40 and 75 characters, Exim
6070 left the rest of the message alone. Two changes have been made: (a) I've
6071 reduced the minimum length to 35 characters; (b) if it can't find a break
6072 point between 35 and 75 characters, it looks ahead and uses the first one
6073 that it finds. This may give the occasional overlong line, but at least the
6074 remaining text gets split now.
6076 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6077 file descriptors might be low, and that setting 1000 would always raise it.
6078 It turns out that in some environments, the limit is already over 1000 and
6079 that lowering it causes trouble. So now Exim takes care not to decrease it.
6081 40. When delivering a message, the value of $return_path is set to $sender_
6082 address at the start of routing (routers may change the value). By an
6083 oversight, this default was not being set up when an address was tested by
6084 -bt or -bv, which affected the outcome if any router or filter referred to
6087 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6088 log only when "message" or "log_message" are set. However, if one of the
6089 conditions was an address verification, or a call to a nested ACL, the
6090 messages generated by the underlying test were being passed through. This
6091 no longer happens. The underlying message is available in $acl_verify_
6092 message for both "message" and "log_message" expansions, so it can be
6093 passed through if needed.
6095 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6096 new expansion $bh_ to give the encoded byte string without charset
6097 translation. Translation happens only if iconv() is available; HAVE_ICONV
6098 indicates this at build time. HEADERS_CHARSET gives the charset to
6099 translate to; headers_charset can change it in the configuration, and
6100 "headers charset" can change it in an individual filter file.
6102 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6103 that creates RFC 2047 encoded "words" labels them as that charset instead
6104 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6105 expansion operator; (ii) when Exim creates a From: line for a local
6106 message; (iii) when a header line is rewritten to include a "phrase" part.
6108 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6109 buggy, causing it to skip the first lines of messages whose message ID
6110 ended in 'D'. This would not have bitten before Exim release 4.14, because
6111 message IDs were unlikely to end in 'D' before then. The effect was to have
6112 incorrect size information for certain domains.
6114 45. #include "config.h" was missing at the start of the crypt16.c module. This
6115 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6118 46. If there was a timeout during a "random" callout check, Exim treated it as
6119 a failure of the random address, and carried on sending RSET and the real
6120 address. If the delay was just some slowness somewhere, the response to the
6121 original RCPT would be taken as a response to RSET and so on, causing
6122 mayhem of various kinds.
6124 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6125 when I implemented it. It didn't allow for the fact that some option values
6126 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6127 the right test for positive values.
6129 48. Domain names in DNS records are case-independent. Exim always looks them up
6130 in lower case. Some resolvers return domain names in exactly the case they
6131 appear in the zone file, that is, they may contain uppercase letters. Not
6132 all resolvers do this - some return always lower case. Exim was treating a
6133 change of case by a resolver as a change of domain, similar to a widening
6134 of a domain abbreviation. This triggered its re-routing code and so it was
6135 trying to route what was effectively the same domain again. This normally
6136 caused routing to fail (because the router wouldn't handle the domain
6137 twice). Now Exim checks for this case specially, and just changes the
6138 casing of the domain that it ultimately uses when it transmits the message
6141 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6144 50. If a filter generated a file delivery with a non-absolute name (possible if
6145 no home directory exists for the router), the forbid_file option was not
6148 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6149 the existing equality matching.
6151 52. Exim was using ints instead of ino_t variables in some places where it was
6152 dealing with inode numbers.
6154 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6155 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6156 called TMPDIR, and if it finds it is different, it changes its value.
6158 54. The smtp_printf() function is now made available to local_scan() so
6159 additional output lines can be written before returning. There is also an
6160 smtp_fflush() function to enable the detection of a dropped connection.
6161 The variables smtp_input and smtp_batched_input are exported to
6164 55. Changed the default runtime configuration: the message "Unknown user"
6165 has been removed from the ACL, and instead placed on the localuser router,
6166 using the cannot_route_message feature. This means that any verification
6167 failures that generate their own messages won't get overridden. Similarly,
6168 the "Unrouteable address" message that was in the ACL for unverifiable
6169 relay addresses has also been removed.
6171 56. Added hosts_avoid_esmtp to the smtp transport.
6173 57. The exicyclog script was not checking for the esoteric option
6174 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6175 will work only if exicyclog is run under the appropriate euid.
6177 58. Following a discussion on the list, the rules by which Exim recognises line
6178 endings on incoming messages have been changed. The -dropcr and drop_cr
6179 options are now no-ops, retained only for backwards compatibility. The
6180 following line terminators are recognized: LF CRLF CR. However, special
6181 processing applies to CR:
6183 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6184 nor a local message in the state where . is a terminator.
6186 (ii) If a bare CR is encountered in a header line, an extra space is added
6187 after the line terminator so as not to end the header. The reasoning
6188 behind this is that bare CRs in header lines are most likely either
6189 to be mistakes, or people trying to play silly games.
6191 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6192 was being incorrectly given as 18 bytes larger than it should have been.
6193 This is a VOB (very old bug).
6195 60. This may never have affected anything current, but just in case it has:
6196 When the local host is found other than at the start of a list of hosts,
6197 the local host, those with the same MX, and any that follow, are discarded.
6198 When the list in question was part of a longer list of hosts, the following
6199 hosts (not currently being processed) were also being discarded. This no
6200 longer happens. I'm not sure if this situation could ever has previously
6203 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6206 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6207 header. This is something that is recommended in a new Internet Draft, and
6208 is something that is documented as being done by Sendmail. There are two
6209 possible values. For messages generated by the autoreply transport, Exim
6212 Auto-Submitted: auto-replied
6214 whereas for all other generated messages (e.g. bounces) it adds
6216 Auto-Submitted: auto-generated
6218 63. The "personal" condition in filters now includes a test for the
6219 Auto-Submitted: header. If it contains the string "auto-" the message it
6220 not considered personal.
6222 64. Added rcpt_include_affixes as a generic transport option.
6224 65. Added queue_only_override (default true).
6226 66. Added the syslog_duplication option.
6228 67. If what should have been the first header line of a message consisted of
6229 a space followed by a colon, Exim was mis-interpreting it as a header line.
6230 It isn't of course - it is syntactically invalid and should therefore be
6231 treated as the start of the message body. The misbehaviour could have
6232 caused a number of strange effects, including loss of data in subsequent
6233 header lines, and spool format errors.
6235 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6236 client host had authenticated. This control can now be exercised by an ACL
6237 for more flexibility.
6239 69. By default, callouts do not happen when testing with -bh. There is now a
6240 variant, -bhc, which does actually run the callout code, including
6241 consulting and updating the callout cache.
6243 70. Added support for saslauthd authentication, courtesy of Alexander
6246 71. If statvfs() failed on the spool or log directories while checking their
6247 size for availability, Exim confusingly gave the error "space shortage".
6248 Furthermore, in debugging mode it crashed with a floating point exception.
6249 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6250 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6251 serious problem, Exim now writes to the main and panic logs when this
6252 happens, with details of the failure. It then refuses to accept the
6253 incoming message, giving the message "spool directory problem" or "log
6254 directory problem" with a 421 code for SMTP messages.
6256 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6257 0, 1, and 2 exist, because some OS complain for execs without them (see
6258 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6259 descriptors. However, the code omitted to check that the open succeeded,
6260 causing mysterious errors if for some reason the permissions on /dev/null
6261 got screwed. Now Exim writes a message to the main and panic logs, and
6262 bombs out if it can't open /dev/null.
6264 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6265 interact so that it is all more flexible. It is supposed to remain
6266 backwards compatible. Also added extra_local_interfaces.
6268 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6269 to bomb out with an assertion failure - to the client this appears as a
6270 connection drop. This problem occurs in the part of the code that was taken
6271 from the Samba project. Fortunately, the assertion is in a very simple
6272 function, so I have fixed this by reproducing the function inline in the
6273 one place where it is called, and arranging for authentication to fail
6274 instead of killing the process with assert().
6276 75. The SPA client code was not working when the server requested OEM rather
6277 than Unicode encoding.
6279 76. Added code to make require_files with a specific uid setting more usable in
6280 the case where statting the file as root fails - usually a non-root-mounted
6281 NFS file system. When this happens and the failure is EACCES, Exim now
6282 forks a subprocess and does the per-uid checking as the relevant uid.
6284 77. Added process_log_path.
6286 78. If log_file_path was not explicitly set, a setting of check_log_space or
6287 check_log_inodes was ignored.
6289 79. If a space check for the spool or log partitions fails, the incident is now
6290 logged. Of course, in the latter case the data may get lost...
6292 80. Added the %p formatting code to string_format() so that it can be used to
6293 print addresses in debug_print(). Adjusted all the address printing in the
6294 debugging in store.c to use %p rather than %d.
6296 81. There was a concern that a line of code in smtp_in.c could overflow a
6297 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6298 initially expressed, the concern was not well-founded, because trailing
6299 spaces are removed early. However, if the trailing spaces were followed by
6300 a NULL, they did not get removed, so the overflow was possible. Two fixes
6303 (a) I re-wrote the offending code in a cleaner fashion.
6304 (b) If an incoming SMTP command contains a NULL character, it is rejected
6307 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6308 runs initgroups(), so that if the Exim user is in any additional groups,
6309 they will be used during message reception.
6315 The change log for 4.20 and earlier releases has been archived.