1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way. Bug 68.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and priority
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
61 and remote servers. Remote servers can be IPv6 addresses, and
64 JH/17 Bug 68: The spamd_address main option now supports an optional
65 timeout value per server.
67 JH/18 Bug 1581: Router and transport options headers_add/remove can
68 now have the list separator specified.
70 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
73 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
76 JH/21 Support for the A6 type of dns record is withdrawn.
78 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
79 rather than the verbs used.
81 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
82 from 255 to 1024 chars.
84 JH/24 Verification callouts now attempt to use TLS by default.
86 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
87 are generic router options now. The defaults didn't change.
89 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
90 Original patch from Alexander Shikoff, worked over by JH.
92 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
93 files if ZCAT_COMMAND is not executable.
95 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
97 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
102 TL/01 When running the test suite, the README says that variables such as
103 no_msglog_check are global and can be placed anywhere in a specific
104 test's script, however it was observed that placement needed to be near
105 the beginning for it to behave that way. Changed the runtest perl
106 script to read through the entire script once to detect and set these
107 variables, reset to the beginning of the script, and then run through
108 the script parsing/test process like normal.
110 TL/02 The BSD's have an arc4random API. One of the functions to induce
111 adding randomness was arc4random_stir(), but it has been removed in
112 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
113 function when detected.
115 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
116 cause callback expansion.
118 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
119 syntax errors in an expansion can be treated as a string instead of
120 logging or causing an error, due to the internal use of bool_lax
121 instead of bool when processing it.
123 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
124 server certificates when making smtp deliveries.
126 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
128 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
130 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
132 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
133 Merged patch from Sebastian Wiedenroth.
135 JH/05 Fix results-pipe from transport process. Several recipients, combined
136 with certificate use, exposed issues where response data items split
137 over buffer boundaries were not parsed properly. This eventually
138 resulted in duplicates being sent. This issue only became common enough
139 to notice due to the introduction of conection certificate information,
140 the item size being so much larger. Found and fixed by Wolfgang Breyha.
142 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
143 size buffer was used, resulting in syntax errors when an expansion
146 JH/07 Add support for directories of certificates when compiled with a GnuTLS
147 version 3.3.6 or later.
149 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
150 is EXPERIMENTAL_EVENT, the main-configuration and transport options
151 both become "event_action", the variables become $event_name, $event_data
152 and $event_defer_errno. There is a new variable $verify_mode, usable in
153 routers, transports and related events. The tls:cert event is now also
154 raised for inbound connections, if the main configuration event_action
157 TL/06 In test suite, disable OCSP for old versions of openssl which contained
158 early OCSP support, but no stapling (appears to be less than 1.0.0).
160 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
161 server certificate names available under the smtp transport option
162 "tls_verify_cert_hostname" now do not permit multi-component wildcard
165 JH/10 Time-related extraction expansions from certificates now use the main
166 option "timezone" setting for output formatting, and are consistent
167 between OpenSSL and GnuTLS compilations. Bug 1541.
169 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
170 encoded parameter in the incoming message. Bug 1558.
172 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
173 include certificate info, eximon was claiming there were spoolfile
176 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
178 JH/14 Log delivery-related information more consistently, using the sequence
179 "H=<name> [<ip>]" wherever possible.
181 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
182 are problematic for Debian distribution, omit them from the release
185 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
187 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
190 JH/17 Fix a null-indirection in certextract expansions when a nondefault
191 output list separator was used.
196 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
197 checkers that were complaining about end of non-void function with no
200 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
201 This was a regression intruduced in 4.83 by another bugfix.
203 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
205 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
206 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
212 TF/01 Correctly close the server side of TLS when forking for delivery.
214 When a message was received over SMTP with TLS, Exim failed to clear up
215 the incoming connection properly after forking off the child process to
216 deliver the message. In some situations the subsequent outgoing
217 delivery connection happened to have the same fd number as the incoming
218 connection previously had. Exim would try to use TLS and fail, logging
219 a "Bad file descriptor" error.
221 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
222 utilities have not been installed.
224 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
225 temporary space as the ACL may create new global variables.
227 TL/01 LDAP support uses per connection or global context settings, depending
228 upon the detected version of the libraries at build time.
230 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
231 to extract and use the src ip:port in logging and expansions as if it
232 were a direct connection from the outside internet. PPv2 support was
233 updated based on HAProxy spec change in May 2014.
235 JH/02 Add ${listextract {number}{list}{success}{fail}}.
237 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
238 Properly escape header and check for NULL return.
240 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
243 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
245 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
246 characters in header names, implemented as a verify condition.
247 Contributed by Michael Fischer v. Mollard.
249 TL/05 Rename SPF condition results err_perm and err_temp to standardized
250 results permerror and temperror. Previous values are deprecated but
251 still accepted. In a future release, err_perm and err_temp will be
252 completely removed, which will be a backward incompatibility if the
253 ACL tests for either of these two old results. Patch contributed by
254 user bes-internal on the mailing list.
256 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
258 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
259 selectors, in both main and reject logs.
261 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
264 JH/07 Add malware type "sock" for talking to simple daemon.
266 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
268 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
269 routers/transports under cutthrough routing.
271 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
272 numbers. Touch up "bool" conditional to keep the same definition.
274 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
276 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
278 JH/12 Expand items in router/transport headers_add or headers_remove lists
279 individually rather than the list as a whole. Bug 1452.
281 Required for reasonable handling of multiple headers_ options when
282 they may be empty; requires that headers_remove items with embedded
283 colons must have them doubled (or the list-separator changed).
285 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
286 view the policy declared in the DMARC record. Currently, $dmarc_status
287 is a combined value of both the record presence and the result of the
290 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
292 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
293 dnslookup router and the smtp transport (applying to the forward
296 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
297 of ldap servers used for a specific lookup. Patch provided by Heiko
300 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
301 New variable $lookup_dnssec_authenticated for observability.
303 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
304 Patch submitted by Lars Timman.
306 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
308 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
309 Requires trusted mode and valid format message id, aborts otherwise.
310 Patch contributed by Heiko Schlichting.
312 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
313 certextract with support for various fields. Bug 1358.
315 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
316 is requested by default, modifiable by smtp transport option
319 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
320 operate on certificate variables to give certificate fingerprints
321 Also new ${sha256:cert_variable}.
323 JH/23 The PRDR feature is moved from being Experimental into the mainline.
325 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
326 Christian Aistleitner.
328 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
330 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
331 file. Patch from Wolfgang Breyha.
333 JH/25 Expand the coverage of the delivery $host and $host_address to
334 client authenticators run in verify callout. Bug 1476.
336 JH/26 Port service names are now accepted for tls_on_connect_ports, to
337 align with daemon_smtp_ports. Bug 72.
339 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
340 support and error reporting did not work properly.
342 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
343 and is readable. Patch from Andrew Colin Kissa.
345 TL/14 Enhance documentation of ${run expansion and how it parses the
346 commandline after expansion, particularly in the case when an
347 unquoted variable expansion results in an empty value.
349 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
351 PP/02 Fix internal collision of T_APL on systems which support RFC3123
352 by renaming away from it. Addresses GH issue 15, reported by
355 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
357 TL/15 SECURITY: prevent double expansion in math comparison functions
358 (can expand unsanitized data). Not remotely exploitable.
365 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
367 PP/02 Make -n do something, by making it not do something.
368 When combined with -bP, the name of an option is not output.
370 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
373 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
374 $sender_host_name and config options to manage this, and basic check
377 PP/05 DSCP support for outbound connections and control modifier for inbound.
379 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
380 (Only plugin which currently uses this is kerberos4, which nobody should
381 be using, but we should make it available and other future plugins might
382 conceivably use it, even though it would break NAT; stuff *should* be
383 using channel bindings instead).
385 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
386 name; added for Sendmail compatibility; requires admin caller.
387 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
388 just ignore it); requires trusted caller.
389 Also parse but ignore: -Ac -Am -X<logfile>
392 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
394 TL/02 Add +smtp_confirmation as a default logging option.
396 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
397 Patch by Magnus Holmgren from 2007-02-20.
399 TL/04 Bugzilla 1281 - Spec typo.
400 Bugzilla 1283 - Spec typo.
401 Bugzilla 1290 - Spec grammar fixes.
403 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
405 TL/06 Add Experimental DMARC support using libopendmarc libraries.
407 TL/07 Fix an out of order global option causing a segfault. Reported to dev
408 mailing list by by Dmitry Isaikin.
410 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
412 JH/02 Support "G" suffix to numbers in ${if comparisons.
414 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
416 NM/01 Bugzilla 1197 - Spec typo
417 Bugzilla 1196 - Spec examples corrections
419 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
421 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
422 gnutls_enable_pkcs11, but renamed to more accurately indicate its
425 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
426 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
428 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
429 "acl {{name}{arg}...}", and optional args on acl condition
432 JH/05 Permit multiple router/transport headers_add/remove lines.
434 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
436 JH/07 Avoid using a waiting database for a single-message-only transport.
437 Performance patch from Paul Fisher. Bugzilla 1262.
439 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
442 JH/09 Add $headers_added variable, with content from use of ACL modifier
443 add_header (but not yet added to the message). Bugzilla 199.
445 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
446 Pulled from Bugzilla 817 by Wolfgang Breyha.
448 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
450 (nb: this is the same fix as in Exim 4.80.1)
452 JH/11 Add A= logging on delivery lines, and a client_set_id option on
455 JH/12 Add optional authenticated_sender logging to A= and a log_selector
458 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
460 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
461 advertise SMTP AUTH mechanism to us, instead of a generic
462 protocol violation error. Also, make Exim more robust to bad
463 data from the Dovecot auth socket.
465 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
467 When a queue runner is handling a message, Exim first routes the
468 recipient addresses, during which it prunes them based on the retry
469 hints database. After that it attempts to deliver the message to
470 any remaining recipients. It then updates the hints database using
473 So if a recipient address works intermittently, it can get repeatedly
474 deferred at routing time. The retry hints record remains fresh so the
475 address never reaches the final cutoff time.
477 This is a fairly common occurrence when a user is bumping up against
478 their storage quota. Exim had some logic in its local delivery code
479 to deal with this. However it did not apply to per-recipient defers
480 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
482 This change adds a proper retry rule check during routing so that the
483 final cutoff time is checked against the message's age. We only do
484 this check if there is an address retry record and there is not a
485 domain retry record; this implies that previous attempts to handle
486 the address had the retry_use_local_parts option turned on. We use
487 this as an approximation for the destination being like a local
488 delivery, as in LMTP.
490 I suspect this new check makes the old local delivery cutoff check
491 redundant, but I have not verified this so I left the code in place.
493 TF/02 Correct gecos expansion when From: is a prefix of the username.
495 Test 0254 submits a message to Exim with the header
499 When I ran the test suite under the user fanf2, Exim expanded
500 the header to contain my full name, whereas it should have added
501 a Resent-Sender: header. It erroneously treats any prefix of the
502 username as equal to the username.
504 This change corrects that bug.
506 GF/01 DCC debug and logging tidyup
507 Error conditions log to paniclog rather than rejectlog.
508 Debug lines prefixed by "DCC: " to remove any ambiguity.
510 TF/03 Avoid unnecessary rebuilds of lookup-related code.
512 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
513 Bug spotted by Jeremy Harris; was flawed since initial commit.
514 Would have resulted in OCSP responses post-SNI triggering an Exim
515 NULL dereference and crash.
517 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
519 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
520 Bug detection, analysis and fix by Samuel Thibault.
521 Bugzilla 1331, Debian bug #698092.
523 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
525 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
526 Server implementation by Todd Lyons, client by JH.
527 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
528 config variable "prdr_enable" controls whether the server
529 advertises the facility. If the client requests PRDR a new
530 acl_data_smtp_prdr ACL is called once for each recipient, after
531 the body content is received and before the acl_smtp_data ACL.
532 The client is controlled by bolth of: a hosts_try_prdr option
533 on the smtp transport, and the server advertisement.
534 Default client logging of deliveries and rejections involving
535 PRDR are flagged with the string "PRDR".
537 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
538 fclose(). Diagnosis by Todd Lyons.
540 PP/17 Update configure.default to handle IPv6 localhost better.
541 Patch by Alain Williams (plus minor tweaks).
544 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
545 This is now consistent with GnuTLS, and is now documented: the
546 previous undocumented portable approach to treating the option as
547 unset was to force an expansion failure. That still works, and
548 an empty string is now equivalent.
550 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
551 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
552 not performing validation itself.
554 PP/20 Added force_command boolean option to pipe transport.
555 Patch from Nick Koston, of cPanel Inc.
557 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
560 TF/04 Added udpsend ACL modifer and hexquote expansion operator
562 PP/21 Fix eximon continuous updating with timestamped log-files.
563 Broken in a format-string cleanup in 4.80, missed when I repaired the
564 other false fix of the same issue.
565 Report and fix from Heiko Schlichting.
568 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
569 Report from Prashanth Katuri.
571 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
572 It's SecureTransport, so affects any MacOS clients which use the
573 system-integrated TLS libraries, including email clients.
575 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
576 using a MIME ACL for non-SMTP local injection.
577 Report and assistance in diagnosis by Warren Baker.
579 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
581 JH/16 Fix comparisons for 64b. Bugzilla 1385.
583 TL/09 Add expansion variable $authenticated_fail_id to keep track of
584 last id that failed so it may be referenced in subsequent ACL's.
586 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
589 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
590 ldap library initialization, allowing self-signed CA's to be
591 used. Also properly sets require_cert option later in code by
592 using NULL (global ldap config) instead of ldap handle (per
593 session). Bug diagnosis and testing by alxgomz.
595 TL/12 Enhanced documentation in the ratelimit.pl script provided in
596 the src/util/ subdirectory.
598 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
599 renamed to Transport Post Delivery Action by Jeremy Harris, as
602 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
603 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
604 redis_servers = needs to be configured which will be used by the redis
605 lookup. Patch from Warren Baker, of The Packet Hub.
607 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
609 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
610 hostname or reverse DNS when processing a host list. Used suggestions
611 from multiple comments on this bug.
613 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
615 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
616 Missed a few lines, added it to make the runtest require no keyboard
619 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
620 contains upper case chars. Make router use caseful_local_part.
622 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
623 support when GnuTLS has been built with p11-kit.
629 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
631 This, or similar/improved, will also be change PP/11 of 4.82.
637 PP/01 Handle short writes when writing local log-files.
638 In practice, only affects FreeBSD (8 onwards).
639 Bugzilla 1053, with thanks to Dmitry Isaikin.
641 NM/01 Bugzilla 949 - Documentation tweak
643 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
646 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
648 PP/02 Implemented gsasl authenticator.
650 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
652 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
653 `pkg-config foo` for cflags/libs.
655 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
656 with rest of GSASL and with heimdal_gssapi.
658 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
659 `pkg-config foo` for cflags/libs for the TLS implementation.
661 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
662 properties get this fed in as external SSF. A number of robustness
663 and debugging improvements to the cyrus_sasl authenticator.
665 PP/08 cyrus_sasl server now expands the server_realm option.
667 PP/09 Bugzilla 1214 - Log authentication information in reject log.
668 Patch by Jeremy Harris.
670 PP/10 Added dbmjz lookup type.
672 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
674 PP/12 MAIL args handles TAB as well as SP, for better interop with
675 non-compliant senders.
676 Analysis and variant patch by Todd Lyons.
678 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
679 Bug report from Lars Müller <lars@samba.org> (via SUSE),
680 Patch from Dirk Mueller <dmueller@suse.com>
682 PP/13 tls_peerdn now print-escaped for spool files.
683 Observed some $tls_peerdn in wild which contained \n, which resulted
684 in spool file corruption.
686 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
687 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
688 or write after TLS renegotiation, which otherwise led to messages
691 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
692 as a tracking header (ie: a signed header comes before the signature).
693 Patch from Wolfgang Breyha.
695 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
696 comma-sep list; embedded commas doubled.
698 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
700 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
702 Report and patch from Dmitry Banschikov.
704 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
705 Removed SSL_clear() after SSL_new() which led to protocol negotiation
706 failures. We appear to now support TLS1.1+ with Exim.
708 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
709 lets Exim select keys and certificates based upon TLS SNI from client.
710 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
711 before an outbound SMTP session. New log_selector, +tls_sni.
713 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
714 NULL dereference. Report and patch from Alun Jones.
716 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
717 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
718 Not seeing resolver debug output on NetBSD, but suspect this is a
719 resolver implementation change.
721 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
722 Left warnings. Added "eximon gdb" invocation mode.
724 PP/21 Defaulting "accept_8bitmime" to true, not false.
726 PP/22 Added -bw for inetd wait mode support.
728 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
729 locate the relevant includes and libraries. Made this the default.
731 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
732 Bugzilla 1246, report and most of solution from Tomasz Kusy.
734 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
735 This may cause build issues on older platforms.
737 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
738 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
739 gnutls_require_mac & gnutls_require_protocols (no longer supported).
740 Added SNI support via GnuTLS too.
741 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
743 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
745 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
746 Applied second patch from Janne, implementing suggestion to default
747 multiple-strings-in-record handling to match SPF spec.
749 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
751 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
752 read-only, out of scope).
753 Patch from Wolfgang Breyha, report from Stuart Northfield.
755 PP/29 Fix three issues highlighted by clang analyser static analysis.
756 Only crash-plausible issue would require the Cambridge-specific
757 iplookup router and a misconfiguration.
758 Report from Marcin Mirosław.
760 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
762 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
763 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
764 As part of this, removing so much warning spew let me fix some minor
765 real issues in debug logging.
767 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
768 assignment on my part. Fixed.
770 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
771 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
772 Janne Snabb (who went above and beyond: thank you).
774 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
775 string otherwise requires a connection and a bunch more work and it's
776 relatively easy to get wrong. Should also expose TLS library linkage
779 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
780 64-bit ${eval} (JH/03).
782 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
783 GNU libc to support some of the 64-bit stuff, should not lead to
784 conflicts. Defined before os.h is pulled in, so if a given platform
785 needs to override this, it can.
787 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
788 protection layer was required, which is not implemented.
789 Bugzilla 1254, patch from Wolfgang Breyha.
791 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
792 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
793 tls_dhparam take prime identifiers. Also unbreak combination of
794 OpenSSL+DH_params+TLSSNI.
796 PP/39 Disable SSLv2 by default in OpenSSL support.
802 PP/01 Solaris build fix for Oracle's LDAP libraries.
803 Bugzilla 1109, patch from Stephen Usher.
805 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
807 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
810 TF/02 Fix a couple more cases where we did not log the error message
811 when unlink() failed. See also change 4.74-TF/03.
813 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
814 lock up or crash if it happened to be inside a call to libc when it
815 got a SIGUSR1 from exiwhat.
817 The SIGUSR1 handler appends the current process status to the process
818 log which is later printed by exiwhat. It used to use the general
819 purpose logging code to do this, but several functions it calls are
820 not safe for signals.
822 The new output code in the SIGUSR1 handler is specific to the process
823 log, and simple enough that it's easy to inspect for signal safety.
824 Removing some special cases also simplifies the general logging code.
825 Removing the spurious timestamps from the process log simplifies
828 TF/04 Improved ratelimit ACL condition.
830 The /noupdate option has been deprecated in favour of /readonly which
831 has clearer semantics. The /leaky, /strict, and /readonly update modes
832 are mutually exclusive. The update mode is no longer included in the
833 database key; it just determines when the database is updated. (This
834 means that when you upgrde Exim will forget old rate measurements.)
836 Exim now checks that the per_* options are used with an update mode that
837 makes sense for the current ACL. For example, when Exim is processing a
838 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
839 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
840 must specify per_mail/readonly. If you omit the update mode it defaults to
841 /leaky where that makes sense (as before) or /readonly where required.
843 The /noupdate option is now undocumented but still supported for
844 backwards compatibility. It is equivalent to /readonly except that in
845 ACLs where /readonly is required you may specify /leaky/noupdate or
846 /strict/noupdate which are treated the same as /readonly.
848 A useful new feature is the /count= option. This is a generalization
849 of the per_byte option, so that you can measure the throughput of other
850 aggregate values. For example, the per_byte option is now equivalent
851 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
853 The per_rcpt option has been generalized using the /count= mechanism
854 (though it's more complicated than the per_byte equivalence). When it is
855 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
856 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
857 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
858 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
859 Note that using per_rcpt with a non-readonly update mode in more than
860 one ACL will cause the recipients to be double-counted. (The per_mail
861 and per_byte options don't have this problem.)
863 The handling of very low rates has changed slightly. If the computed rate
864 is less than the event's count (usually one) then this event is the first
865 after a long gap. In this case the rate is set to the same as this event's
866 count, so that the first message of a spam run is counted properly.
868 The major new feature is a mechanism for counting the rate of unique
869 events. The new per_addr option counts the number of different
870 recipients that someone has sent messages to in the last time period. It
871 behaves like per_rcpt if all the recipient addresses are different, but
872 duplicate recipient addresses do not increase the measured rate. Like
873 the /count= option this is a general mechanism, so the per_addr option
874 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
875 example, measure the rate that a client uses different sender addresses
876 with the options per_mail/unique=$sender_address. There are further
877 details in the main documentation.
879 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
881 TF/06 Removed a few PCRE remnants.
883 TF/07 Automatically extract Exim's version number from tags in the git
884 repository when doing development or release builds.
886 PP/02 Raise smtp_cmd_buffer_size to 16kB.
887 Bugzilla 879. Patch from Paul Fisher.
889 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
890 Heavily based on revision 40f9a89a from Simon Arlott's tree.
893 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
895 PP/05 Variable $av_failed, true if the AV scanner deferred.
896 Bugzilla 1078. Patch from John Horne.
898 PP/06 Stop make process more reliably on build failure.
899 Bugzilla 1087. Patch from Heiko Schlittermann.
901 PP/07 Make maildir_use_size_file an _expandable_ boolean.
902 Bugzilla 1089. Patch from Heiko Schlittermann.
904 PP/08 Handle ${run} returning more data than OS pipe buffer size.
905 Bugzilla 1131. Patch from Holger Weiß.
907 PP/09 Handle IPv6 addresses with SPF.
908 Bugzilla 860. Patch from Wolfgang Breyha.
910 PP/10 GnuTLS: support TLS 1.2 & 1.1.
912 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
915 PP/11 match_* no longer expand right-hand-side by default.
916 New compile-time build option, EXPAND_LISTMATCH_RHS.
917 New expansion conditions, "inlist", "inlisti".
919 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
921 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
923 PP/14 fix log_write() format string regression from TF/03.
924 Bugzilla 1152. Patch from Dmitry Isaikin.
930 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
932 PP/02 Harmonised TLS library version reporting; only show if debugging.
933 Layout now matches that introduced for other libraries in 4.74 PP/03.
935 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
937 PP/04 New "dns_use_edns0" global option.
939 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
942 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
943 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
945 TK/01 Updated PolarSSL code to 0.14.2.
946 Bugzilla 1097. Patch from Andreas Metzler.
948 PP/07 Catch divide-by-zero in ${eval:...}.
951 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
954 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
955 format-string attack -- SECURITY: remote arbitrary code execution.
957 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
958 time unintentionally subject to list matching rules, letting the header
959 cause arbitrary Exim lookups (of items which can occur in lists, *not*
960 arbitrary string expansion). This allowed for information disclosure.
962 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
963 INT_MIN/-1 -- value coerced to INT_MAX.
969 NM/01 Workround for PCRE version dependency in version reporting
972 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
973 This fixes portability to compilers other than gcc, notably
974 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
976 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
977 makefiles for portability to HP-UX and POSIX correctness.
979 PP/01 Permit LOOKUP_foo enabling on the make command-line.
980 Also via indented variable definition in the Makefile.
981 (Debugging by Oliver Heesakkers).
983 PP/02 Restore caching of spamd results with expanded spamd_address.
984 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
986 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
987 Improves build reliability. Fix from: Frank Elsner
989 NM/02 Fix wide character breakage in the rfc2047 coding
990 Fixes bug 1064. Patch from Andrey N. Oktyabrski
992 NM/03 Allow underscore in dnslist lookups
993 Fixes bug 1026. Patch from Graeme Fowler
995 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
996 Code patches from Adam Ciarcinski of NetBSD.
998 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1001 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1002 is logged, to avoid truncation. Patch from John Horne.
1004 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1005 Patch from Jakob Hirsch.
1007 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1008 SQL string expansion failure details.
1009 Patch from Andrey Oktyabrski.
1011 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1012 Patch from Simon Arlott.
1014 PP/09 New lookups functionality failed to compile on old gcc which rejects
1015 extern declarations in function scope.
1016 Patch from Oliver Fleischmann
1018 PP/10 Use sig_atomic_t for flags set from signal handlers.
1019 Check getgroups() return and improve debugging.
1020 Fixed developed for diagnosis in bug 927 (which turned out to be
1023 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1024 Patch from Mark Zealey.
1026 PP/12 Bugzilla 1056: Improved spamd server selection.
1027 Patch from Mark Zealey.
1029 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1030 Based on patch from Heiko Schlittermann.
1032 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1033 Patch from Uwe Doering, sign-off by Michael Haardt.
1035 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1036 variable declaration deep within a block. Bug and patch from
1039 PP/15 lookups-Makefile IRIX compatibilty coercion.
1041 PP/16 Make DISABLE_DKIM build knob functional.
1043 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1044 Patch by Simon Arlott
1046 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1047 variable argument macros. Our copy now differs from upstream.
1053 TF/01 Failure to get a lock on a hints database can have serious
1054 consequences so log it to the panic log.
1056 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1057 controlled using the smtp_confirmation log selector.
1059 TF/03 Include the error message when we fail to unlink a spool file.
1061 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1062 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1063 for maintaining out-of-tree patches for some time.
1065 PP/01 Bugzilla 139: Documentation and portability issues.
1066 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1067 Handle per-OS dynamic-module compilation flags.
1069 PP/02 Let /dev/null have normal permissions.
1070 The 4.73 fixes were a little too stringent and complained about the
1071 permissions on /dev/null. Exempt it from some checks.
1072 Reported by Andreas M. Kirchwitz.
1074 PP/03 Report version information for many libraries, including
1075 Exim version information for dynamically loaded libraries. Created
1076 version.h, now support a version extension string for distributors
1077 who patch heavily. Dynamic module ABI change.
1079 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1080 privilege escalation vulnerability whereby the Exim run-time user
1081 can cause root to append content of the attacker's choosing to
1084 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1087 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1088 If dropping privileges for untrusted macros, we disabled normal logging
1089 on the basis that it would fail; for the Exim run-time user, this is not
1090 the case, and it resulted in successful deliveries going unlogged.
1091 Fixed. Reported by Andreas Metzler.
1097 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1098 only prepend for the Resent-* case. Fixes regression introduced in
1099 Exim 4.70 by NM/22 for Bugzilla 607.
1101 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1102 increasing numbers of administrators be bitten by this.
1104 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1106 PP/03 Bugzilla 994: added openssl_options main configuration option.
1108 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1110 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1112 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1114 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1115 without a peer certificate, leading to a segfault because of an
1116 assumption that peers always have certificates. Be a little more
1117 paranoid. Problem reported by Martin Tscholak.
1119 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1120 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1121 NB: ClamAV planning to remove STREAM in "middle of 2010".
1122 CL also introduces -bmalware, various -d+acl logging additions and
1123 more caution in buffer sizes.
1125 PP/09 Implemented reverse_ip expansion operator.
1127 PP/10 Bugzilla 937: provide a "debug" ACL control.
1129 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1131 PP/12 Bugzilla 973: Implement --version.
1133 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1135 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1137 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1139 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1140 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1141 ignore trailing whitespace.
1143 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1145 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1148 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1149 Notification from Dr Andrew Aitchison.
1151 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1152 ExtendedDetectionInfo response format.
1153 Notification from John Horne.
1155 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1158 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1159 XSL and documented dependency on system catalogs, with examples of how
1162 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1165 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1166 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1167 configuration file which is writeable by the Exim user or group.
1169 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1170 of configuration files to cover files specified with the -C option if
1171 they are going to be used with root privileges, not just the default
1174 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1175 option (effectively making it always true).
1177 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1178 files to be used while preserving root privileges.
1180 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1181 that rogue child processes cannot use them.
1183 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1184 run-time user, instead of root.
1186 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1187 Exim run-time user without dropping privileges.
1189 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1190 result string, instead of calling string_vformat() twice with the same
1193 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1194 for other users. Others should always drop root privileges if they use
1195 -C on the command line, even for a whitelisted configure file.
1197 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1199 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1205 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1206 $data_path, and $header_path variables; fixed documentation bugs and
1209 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1210 exipick to access non-standard spools, including the "frozen" queue
1213 NM/01 Bugzilla 965: Support mysql stored procedures.
1214 Patch from Alain Williams
1216 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1218 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1219 Patch from Andreas Metzler
1221 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1222 Patch from Kirill Miazine
1224 NM/05 Bugzilla 671: Added umask to procmail example.
1226 JJ/03 installed exipick 20100323.0, fixing doc bug
1228 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1229 directory. Notification and patch from Dan Rosenberg.
1231 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1233 TK/02 Improve log output when DKIM signing operation fails.
1235 MH/01 Treat the transport option dkim_domain as a colon separated
1236 list, not as a single string, and sign the message with each element,
1237 omitting multiple occurences of the same signer.
1239 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1240 Bugzilla 985, 986. Patch by Simon Arlott
1242 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1243 Patch by Simon Arlott
1245 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1246 MBX locking. Notification from Dan Rosenberg.
1252 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1254 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1256 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1258 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1260 NM/04 Bugzilla 915: Flag broken perl installation during build.
1266 TK/01 Added patch by Johannes Berg that expands the main option
1267 "spamd_address" if it starts with a dollar sign.
1269 TK/02 Write list of recipients to X-Envelope-Sender header when building
1270 the mbox-format spool file for content scanning (suggested by Jakob
1273 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1274 (http://www.dcc-servers.net/) support via dccifd. Activated by
1275 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1277 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1278 by Mark Daniel Reidel <mr@df.eu>.
1280 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1281 When building exim an external PCRE library is now needed -
1282 PCRE is a system library on the majority of modern systems.
1283 See entry on PCRE_LIBS in EDITME file.
1285 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1286 conversation. Added nologin parameter to request.
1287 Patch contributed by Kirill Miazine.
1289 TF/01 Do not log submission mode rewrites if they do not change the address.
1291 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1293 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1294 log files in place. Contributed by Roberto Lima.
1296 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1298 TF/03 Bugzilla 615: When checking the local_parts router precondition
1299 after a local_part_suffix or local_part_prefix option, Exim now
1300 does not use the address's named list lookup cache, since this
1301 contains cached lookups for the whole local part.
1303 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1304 Robert Millan. Documentation is in experimental-spec.txt.
1306 TF/04 Bugzilla 668: Fix parallel build (make -j).
1308 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1310 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1311 Patch provided by Jan Srzednicki.
1313 TF/05 Leading white space used to be stripped from $spam_report which
1314 wrecked the formatting. Now it is preserved.
1316 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1317 that they are available at delivery time.
1319 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1321 TF/08 TLS error reporting now respects the incoming_interface and
1322 incoming_port log selectors.
1324 TF/09 Produce a more useful error message if an SMTP transport's hosts
1325 setting expands to an empty string.
1327 NM/06 Bugzilla 744: EXPN did not work under TLS.
1328 Patch provided by Phil Pennock.
1330 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1331 Patch provided by Richard Godbee.
1333 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1334 acl_smtp_notquit, added index entry.
1336 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1337 Patch provided by Eugene Bujak.
1339 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1340 accept(). Patch provided by Maxim Dounin.
1342 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1343 Patch provided by Phil Pennock.
1345 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1347 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1348 Patch provided by Brad "anomie" Jorsch.
1350 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1351 Patch provided by Dean Brooks.
1353 TK/05 Add native DKIM support (does not depend on external libraries).
1355 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1356 Patch provided by Graeme Fowler.
1358 NM/16 Bugzilla 851: Documentation example syntax fix.
1360 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1362 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1365 NM/19 Bugzilla 745: TLS version reporting.
1366 Patch provided by Phil Pennock.
1368 NM/20 Bugzilla 167: bool: condition support.
1369 Patch provided by Phil Pennock.
1371 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1372 clients. Patch provided by Phil Pennock.
1374 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1375 Patch provided by Brad "anomie" Jorsch.
1377 NM/23 Bugzilla 687: Fix misparses in eximstats.
1378 Patch provided by Heiko Schlittermann.
1380 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1381 Patch provided by Heiko Schlittermann.
1383 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1384 plus update to original patch.
1386 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1388 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1389 Patch provided by David Brownlee.
1391 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1393 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1395 NM/30 Bugzilla 888: TLS documentation bugfixes.
1397 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1399 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1400 Unlike the original bugzilla I have changed all shell scripts in src tree.
1402 NM/33 Bugzilla 898: Transport filter timeout fix.
1403 Patch by Todd Rinaldo.
1405 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1406 Patch by Serge Demonchaux.
1408 NM/35 Bugzilla 39: Base64 decode bug fixes.
1409 Patch by Jakob Hirsch.
1411 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1413 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1415 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1417 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1423 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1424 ALT-N's libdkim that I have put here:
1425 http://duncanthrax.net/exim-experimental/
1427 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1428 were called 'true' and it seems that C99 defines that as a reserved
1429 keyword to be used with 'bool' variable types. That means you could
1430 not include C99-style headers which use bools without triggering
1431 build errors in sieve.c.
1433 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1434 as mailq or other aliases. Changed the --help handling significantly
1435 to do whats expected. exim_usage() emits usage/help information.
1437 SC/01 Added the -bylocaldomain option to eximstats.
1439 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1441 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1443 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1449 PH/01 Another patch from the Sieve maintainer.
1451 PH/02 When an IPv6 address is converted to a string for single-key lookup
1452 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1453 dots are used instead of colons so that keys in lsearch files need not
1454 contain colons. This was done some time before quoting was made available
1455 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1456 (notated using the quote facility) so as to distinguish them from IPv4
1457 keys. This meant that lookups for IP addresses in host lists did not work
1458 for iplsearch lookups.
1460 This has been fixed by arranging for IPv6 addresses to be expressed with
1461 colons if the lookup type is iplsearch. This is not incompatible, because
1462 previously such lookups could never work.
1464 The situation is now rather anomolous, since one *can* have colons in
1465 ordinary lsearch keys. However, making the change in all cases is
1466 incompatible and would probably break a number of configurations.
1468 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1471 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1472 conversion specification without a maximum field width, thereby enabling
1473 a rogue spamd server to cause a buffer overflow. While nobody in their
1474 right mind would setup Exim to query an untrusted spamd server, an
1475 attacker that gains access to a server running spamd could potentially
1476 exploit this vulnerability to run arbitrary code as the Exim user.
1478 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1479 $primary_hostname instead of what libspf2 thinks the hosts name is.
1481 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1482 a directory entry by the name of the lookup key. Previously, if a
1483 symlink pointed to a non-existing file or a file in a directory that
1484 Exim lacked permissions to read, a lookup for a key matching that
1485 symlink would fail. Now it is enough that a matching directory entry
1486 exists, symlink or not. (Bugzilla 503.)
1488 PH/03 The body_linecount and body_zerocount variables are now exported in the
1491 PH/04 Added the $dnslist_matched variable.
1493 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1494 This means they are set thereafter only if the connection becomes
1497 PH/06 Added the client_condition to authenticators so that some can be skipped
1498 by clients under certain conditions.
1500 PH/07 The error message for a badly-placed control=no_multiline_responses left
1501 "_responses" off the end of the name.
1503 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1505 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1506 (without spaces) instead of just copying the configuration text.
1508 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1510 PH/11 Added $max_received_linelength.
1512 PH/12 Added +ignore_defer and +include_defer to host lists.
1514 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1515 way in which PCRE > 7.0 is built.
1517 PH/14 Implemented queue_only_load_latch.
1519 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1520 MAIL command. The effect was to mangle the value on 64-bit systems.
1522 PH/16 Another patch from the Sieve maintainer.
1524 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1526 PH/18 If a system quota error occurred while trying to create the file for
1527 a maildir delivery, the message "Mailbox is full" was not appended to the
1528 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1529 only to a quota excession during the actual writing of the file.
1531 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1532 characters?) which causes problems in log lines. The DN values are now
1533 passed through string_printing() before being added to log lines.
1535 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1536 and InterBase are left for another time.)
1538 PH/21 Added message_body_newlines option.
1540 PH/22 Guard against possible overflow in moan_check_errorcopy().
1542 PH/23 POSIX allows open() to be a macro; guard against that.
1544 PH/24 If the recipient of an error message contained an @ in the local part
1545 (suitably quoted, of course), incorrect values were put in $domain and
1546 $local_part during the evaluation of errors_copy.
1552 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1553 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1556 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1557 issue a MAIL command.
1559 PH/02 In an ACL statement such as
1561 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1563 if a client was not listed at all, or was listed with a value other than
1564 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1565 the condition was not true (as it should be), so access was not denied.
1566 The bug was that the ! inversion was incorrectly passed on to the second
1567 item. This has been fixed.
1569 PH/03 Added additional dnslists conditions == and =& which are different from
1570 = and & when the dns lookup returns more than one IP address.
1572 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1573 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1575 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1576 FSYNC, which compiles an option called disable_fsync that allows for
1577 bypassing fsync(). The documentation is heavily laced with warnings.
1579 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1581 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1582 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1583 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1584 including adding "make clean"; (3) Added -fPIC when compiling the test
1585 dynamically loaded module, to get rid of a warning.
1587 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1588 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1589 = 0s. The bug is otherwise harmless.
1591 PH/07 There was a bug in the dovecot authenticator such that the value of
1592 $auth1 could be overwritten, and so not correctly preserved, after a
1593 successful authentication. This usually meant that the value preserved by
1594 the server_setid option was incorrect.
1596 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1598 PH/09 Installed PCRE release 7.0.
1600 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1601 run for batched SMTP input. It is now run at the start of every message
1602 in the batch. While fixing this I discovered that the process information
1603 (output by running exiwhat) was not always getting set for -bs and -bS
1604 input. This is fixed, and it now also says "batched" for BSMTP.
1606 PH/11 Added control=no_pipelining.
1608 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1609 patch, slightly modified), and move the expansion of helo_data till after
1610 the connection is made in the smtp transport (so it can use these
1613 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1615 PH/14 Added log_selector = +pid.
1617 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1619 PH/16 Add ${if forany and ${if forall.
1621 PH/17 Added dsn_from option to vary the From: line in DSNs.
1623 PH/18 Flush SMTP output before performing a callout, unless control =
1624 no_callout_flush is set.
1626 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1627 was true (the default) a successful delivery failed to delete the retry
1628 item, thus causing premature timeout of the address. The bug is now
1631 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1633 PH/21 Long custom messages for fakedefer and fakereject are now split up
1634 into multiline reponses in the same way that messages for "deny" and
1635 other ACL rejections are.
1637 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1638 with slight modification.
1640 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1641 draft, changing the syntax and factoring some duplicate code".
1643 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1644 for deliveries of the second and subsequent messages over the same SMTP
1647 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1648 ${reduce, with only minor "tidies".
1650 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1652 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1653 expansion side effects.
1655 PH/27 When a message times out after an over-quota error from an Exim-imposed
1656 quota, the bounce message says "mailbox is full". This message was not
1657 being given when it was a system quota that was exceeded. It now should
1660 MH/03 Made $recipients available in local_scan(). local_scan() already has
1661 better access to the recipient list through recipients_list[], but
1662 $recipients can be useful in postmaster-provided expansion strings.
1664 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1665 in the case of a MAIL command with additional options following the
1666 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1667 were accidentally chopped off.
1669 PH/29 SMTP synchronization checks are implemented when a command is read -
1670 there is a check that no more input is waiting when there shouldn't be
1671 any. However, for some commands, a delay in an ACL can mean that it is
1672 some time before the response is written. In this time, more input might
1673 arrive, invalidly. So now there are extra checks after an ACL has run for
1674 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1675 pipelining has not been advertised.
1677 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1679 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1680 returned in the SMTP response when smtp_return_error_details was false.
1681 This has been fixed.
1683 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1684 instead of the C I/O that was originally supplied, because problems were
1685 reported on Solaris.
1687 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1688 Exim which did not show up earlier: it was assuming that a call to
1689 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1690 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1691 was a macro that became an assignment, so it seemed to work. This has
1692 changed to a proper function call with a void return, hence the compile
1693 error. Exim's code has been fixed.
1695 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1698 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1700 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1702 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1703 because it turns out to be too big or there is a timeout) there is no
1704 "Completed" line in the log. When some messages of this type were
1705 selected by exigrep, they were listed as "not completed". Others were
1706 picked up by some special patterns. I have improved the selection
1707 criteria to be more general.
1709 PH/38 The host_find_failed option in the manualroute router can now be set
1710 to "ignore", to completely ignore a host whose IP address cannot be
1711 found. If all hosts are ignored, the behaviour is controlled by the new
1712 host_all_ignored option.
1714 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1715 homing or because of multiple MX records with /mx) generated more than
1716 one IP address, and the following item turned out to be the local host,
1717 all the secondary addresses of the first item were incorrectly removed
1718 from the list, along with the local host and any following hosts (which
1719 is what is supposed to happen).
1721 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1722 whoever called Exim into the -H file. In the case of the daemon it was
1723 behaving confusingly. When first started, it used values for whoever
1724 started the daemon, but after a SIGHUP it used the Exim user (because it
1725 calls itself on a restart). I have changed the code so that it now always
1728 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1729 message are rejected with the same error (e.g. no authentication or bad
1730 sender address), and a DATA command is nevertheless sent (as can happen
1731 with PIPELINING or a stupid MUA), the error message that was given to the
1732 RCPT commands is included in the rejection of the DATA command. This is
1733 intended to be helpful for MUAs that show only the final error to their
1736 PH/42 Another patch from the Sieve maintainer.
1738 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1739 Eximstats - Fixed some broken HTML links and added missing column headers
1741 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1742 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1744 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1746 PH/43 Yet another patch from the Sieve maintainer.
1748 PH/44 I found a way to check for a TCP/IP connection going away before sending
1749 the response to the final '.' that terminates a message, but only in the
1750 case where the client has not sent further data following the '.'
1751 (unfortunately, this is allowed). However, in many cases there won't be
1752 any further data because there won't be any more messages to send. A call
1753 to select() can be used: if it shows that the input is "ready", there is
1754 either input waiting, or the socket has been closed. An attempt to read
1755 the next input character can distinguish the two cases. Previously, Exim
1756 would have sent an OK response which the client would never have see.
1757 This could lead to message repetition. This fix should cure that, at
1758 least in a lot of common cases.
1760 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1761 advertised in response to EHLO.
1767 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1768 fixed by 4.65/MH/01 (is this a record?) are fixed:
1770 (i) An empty string was always treated as zero by the numeric comparison
1771 operators. This behaviour has been restored.
1773 (ii) It is documented that the numeric comparison operators always treat
1774 their arguments as decimal numbers. This was broken in that numbers
1775 starting with 0 were being interpreted as octal.
1777 While fixing these problems I realized that there was another issue that
1778 hadn't been noticed. Values of message_size_limit (both the global option
1779 and the transport option) were treated as octal if they started with 0.
1780 The documentation was vague. These values are now always treated as
1781 decimal, and I will make that clear in the documentation.
1787 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1788 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1791 MH/01 Don't check that the operands of numeric comparison operators are
1792 integers when their expansion is in "skipping" mode (fixes bug
1793 introduced by 4.64-PH/07).
1795 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1796 child addresses, Exim now panics and dies. Previously, because the count
1797 is held in a short int, deliveries were likely to be lost. As such a
1798 large number of recipients for a single message is ridiculous
1799 (performance will be very, very poor), I have chosen to impose a limit
1800 rather than extend the field.
1806 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1807 leftover -K file (the existence of which was triggered by #402).
1808 While we were at it, introduced process PID as part of the -K
1809 filename. This should rule out race conditions when creating
1812 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1813 processing considerably. Previous code took too long for large mails,
1814 triggering a timeout which in turn triggers #401.
1816 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1817 in the DK code in transports.c. sendfile() is not really portable,
1818 hence the _LINUX specificness.
1820 TF/01 In the add_headers option to the mail command in an Exim filter,
1821 there was a bug that Exim would claim a syntax error in any
1822 header after the first one which had an odd number of characters
1825 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1826 callout verification, Exim cached a "reject" for the entire domain. This
1827 is correct for most verifications, but it is not correct for a recipient
1828 verification with use_sender or use_postmaster set, because in that case
1829 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1830 case of MAIL FROM:<> rejection from other early rejections (e.g.
1831 rejection of HELO). When verifying a recipient using a non-null MAIL
1832 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1833 Whatever the result of the callout, the value of the domain cache is
1834 left unchanged (for any other kind of callout, getting as far as trying
1835 RCPT means that the domain itself is ok).
1837 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1840 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1841 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1842 session, the daemon ignores ECONNECT errors and logs others; it now
1843 ignores EPIPE as well.
1845 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1846 (quoted-printable decoding).
1848 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1849 later the small subsequent patch to fix an introduced bug.
1851 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1853 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1855 PH/08 An error is now given if message_size_limit is specified negative.
1857 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1858 to be given (somewhat) arbitrary names.
1860 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1863 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1864 miscellaneous code fixes
1866 PH/10 Added the log_reject_target ACL modifier to specify where to log
1869 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1870 hostname. This is wrong, because it relates to the incoming message (and
1871 probably the interface on which it is arriving) and not to the outgoing
1872 callout (which could be using a different interface). This has been
1873 changed to use the value of the helo_data option from the smtp transport
1874 instead - this is what is used when a message is actually being sent. If
1875 there is no remote transport (possible with a router that sets up host
1876 addresses), $smtp_active_hostname is used.
1878 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1879 tweaks were necessary in order to get it to work (see also 21 below):
1880 (a) The code assumed that strncpy() returns a negative number on buffer
1881 overflow, which isn't the case. Replaced with Exim's string_format()
1883 (b) There were several signed/unsigned issues. I just did the minimum
1884 hacking in of casts. There is scope for a larger refactoring.
1885 (c) The code used strcasecmp() which is not a standard C function.
1886 Replaced with Exim's strcmpic() function.
1887 (d) The code set only $1; it now sets $auth1 as well.
1888 (e) A simple test gave the error "authentication client didn't specify
1889 service in request". It would seem that Dovecot has changed its
1890 interface. Fortunately there's a specification; I followed it and
1891 changed what the client sends and it appears to be working now.
1893 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1896 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1897 address A is aliased to B and C, where B exists and C does not. Without
1898 -v the output is "A verified" because verification stops after a
1899 successful redirection if more than one address is generated. However,
1900 with -v the child addresses are also verified. Exim was outputting "A
1901 failed to verify" and then showing the successful verification for C,
1902 with its parentage. It now outputs "B failed to verify", showing B's
1903 parentage before showing the successful verification of C.
1905 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1906 look up a TXT record in a specific list after matching in a combined
1909 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1910 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1911 they consult the DNS. I had assumed they would set it the way they
1912 wanted; and indeed my experiments on Linux seem to show that in some
1913 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1914 To be on the safe side, however, I have now made the interface to
1915 host_find_byname() similar to host_find_bydns(), with an argument
1916 containing the DNS resolver options. The host_find_byname() function now
1917 sets these options at its start, just as host_find_bydns() does. The smtp
1918 transport options dns_qualify_single and dns_search_parents are passed to
1919 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1920 of host_find_byname() use the default settings of RES_DEFNAMES
1921 (qualify_single) but not RES_DNSRCH (search_parents).
1923 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1924 spool_read_header() do less string testing, by means of a preliminary
1925 switch on the second character of optional "-foo" lines. (This is
1926 overdue, caused by the large number of possibilities that now exist.
1927 Originally there were few.) While I was there, I also converted the
1928 str(n)cmp tests so they don't re-test the leading "-" and the first
1929 character, in the hope this might squeeze out yet more improvement.
1931 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1932 flag allowing group syntax was set by the header_syntax check but not
1933 turned off, possible causing trouble later; (2) The flag was not being
1934 set at all for the header_verify test, causing "group"-style headers to
1935 be rejected. I have now set it in this case, and also caused header_
1936 verify to ignore an empty address taken from a group. While doing this, I
1937 came across some other cases where the code for allowing group syntax
1938 while scanning a header line wasn't quite right (mostly, not resetting
1939 the flag correctly in the right place). These bugs could have caused
1940 trouble for malformed header lines. I hope it is now all correct.
1942 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1943 with the "reply" argument non-NULL. The code, however (which originally
1944 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1945 but it didn't always do it. This confused somebody who was copying the
1946 code for some other use. I have removed all the tests.
1948 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1949 feature that was used to support insecure browsers during the U.S. crypto
1950 embargo. It requires special client support, and Exim is probably the
1951 only MTA that supported it -- and would never use it because real RSA is
1952 always available. This code has been removed, because it had the bad
1953 effect of slowing Exim down by computing (never used) parameters for the
1954 RSA_EXPORT functionality.
1956 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1957 authenticator to fail if there's a tab character in the incoming data
1958 (there should never be unless someone is messing about, as it's supposed
1959 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1960 if the connection is using TLS or if the remote IP is the same as the
1961 local IP, and the "valid-client-cert option" if a client certificate has
1964 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1965 authenticators. This can be used for authorization after authentication
1966 succeeds. (In the case of plaintext, it servers for both authentication
1969 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1970 if any retry times were supplied.
1972 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1973 connection, where there is no client IP address to check. In this
1974 situation, the verify now always succeeds.
1976 PH/25 Applied John Jetmore's -Mset patch.
1978 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1980 PH/27 In a string expansion for a processed (not raw) header when multiple
1981 headers of the same name were present, leading whitespace was being
1982 removed from all of them, but trailing whitespace was being removed only
1983 from the last one. Now trailing whitespace is removed from each header
1984 before concatenation. Completely empty headers in a concatenation (as
1985 before) are ignored.
1987 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1988 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1990 PH/29 [Removed. This was a change that I later backed out, and forgot to
1991 correct the ChangeLog entry (that I had efficiently created) before
1992 committing the later change.]
1994 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1995 address errors (4xx response to RCPT) over the same connection as other
1996 messages routed to the same hosts. Such deliveries are always "forced",
1997 so retry times are not inspected. This resulted in far too many retries
1998 for the affected addresses. The effect occurred only when there were more
1999 hosts than the hosts_max_try setting in the smtp transport when it had
2000 the 4xx errors. Those hosts that it had tried were not added to the list
2001 of hosts for which the message was waiting, so if all were tried, there
2002 was no problem. Two fixes have been applied:
2004 (i) If there are any address or message errors in an SMTP delivery, none
2005 of the hosts (tried or untried) are now added to the list of hosts
2006 for which the message is waiting, so the message should not be a
2007 candidate for sending over the same connection that was used for a
2008 successful delivery of some other message. This seems entirely
2009 reasonable: after all the message is NOT "waiting for some host".
2010 This is so "obvious" that I'm not sure why it wasn't done
2011 previously. Hope I haven't missed anything, but it can't do any
2012 harm, as the worst effect is to miss an optimization.
2014 (ii) If, despite (i), such a delivery is accidentally attempted, the
2015 routing retry time is respected, so at least it doesn't keep
2016 hammering the server.
2018 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2019 in ${readsocket because some servers need this prod.
2021 PH/32 Added some extra debug output when updating a wait-xxx database.
2023 PH/33 The hint "could be header name not terminated by colon", which has been
2024 given for certain expansion errors for a long time, was not being given
2025 for the ${if def:h_colon_omitted{... case.
2027 PH/34 The spec says: "With one important exception, whenever a domain list is
2028 being scanned, $domain contains the subject domain." There was at least
2029 one case where this was not true.
2031 PH/35 The error "getsockname() failed: connection reset by peer" was being
2032 written to the panic log as well as the main log, but it isn't really
2033 panic-worthy as it just means the connection died rather early on. I have
2034 removed the panic log writing for the ECONNRESET error when getsockname()
2037 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2038 runs only) independently of the message's sender address. This meant
2039 that, if the 4xx error was in fact related to the sender, a different
2040 message to the same recipient with a different sender could confuse
2041 things. In particualar, this can happen when sending to a greylisting
2042 server, but other circumstances could also provoke similar problems.
2043 I have changed the default so that the retry time for these errors is now
2044 based a combination of the sender and recipient addresses. This change
2045 can be overridden by setting address_retry_include_sender=false in the
2048 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2049 remote server are returned as part of bounce messages. This was not
2050 happening for LMTP over a pipe (the lmtp transport), but now it is the
2051 same for both kinds of LMTP.
2053 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2054 in header lines that were in fact CNAMEs. This is no longer the case.
2056 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2057 and queue runs started by the daemon processed all messages. This has
2058 been fixed so that -R and -S can now usefully be given with -q<time>.
2060 PH/40 Import PCRE release 6.7 (fixes some bugs).
2062 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2064 PH/42 Give an error if -q is specified more than once.
2066 PH/43 Renamed the variables $interface_address and $interface_port as
2067 $received_ip_address and $received_port, to make it clear that these
2068 values apply to message reception, and not to the outgoing interface when
2069 a message is delivered. (The old names remain recognized, of course.)
2071 PH/44 There was no timeout on the connect() call when using a Unix domain
2072 socket in the ${readsocket expansion. There now is.
2074 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2075 be meaningful with "accept".
2077 SC/01 Eximstats V1.43
2078 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2080 SC/02 Eximstats V1.44
2081 Use a glob alias rather than an array ref in the generated
2082 parser. This improves both readability and performance.
2084 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2085 Collect SpamAssassin and rejection statistics.
2086 Don't display local sender or destination tables unless
2087 there is data to show.
2088 Added average volumes into the top table text output.
2090 SC/04 Eximstats V1.46
2091 Collect data on the number of addresses (recipients)
2092 as well as the number of messages.
2094 SC/05 Eximstats V1.47
2095 Added 'Message too big' to the list of mail rejection
2096 reasons (thanks to Marco Gaiarin).
2098 SC/06 Eximstats V1.48
2099 Mainlog lines which have GMT offsets and are too short to
2100 have a flag are now skipped.
2102 SC/07 Eximstats V1.49 (Alain Williams)
2103 Added the -emptyok flag.
2105 SC/08 Eximstats V1.50
2106 Fixes for obtaining the IP address from reject messages.
2108 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2109 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2110 whitesspace changes from 4.64-PH/27
2112 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2115 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2116 are found, allow negative numbers in numeric criteria)
2118 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2120 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2121 to match changes made in 4.64-PH/43
2123 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2125 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2126 "domain" is actually an IP address, and force a failure. This locks out
2127 those revolvers/nameservers that support "A-for-A" lookups, in
2128 contravention of the specifications.
2130 PH/48 When a host name was looked up from an IP address, and the subsequent
2131 forward lookup of the name timed out, the host name was left in
2132 $sender_host_name, contrary to the specification.
2134 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2135 restricted to single-key lookups, Exim was not diagnosing an error if
2136 * or *@ was used with a query-style lookup.
2138 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2140 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2141 long ago, but noone interested enough thought of it. Let's just say that
2142 the "1.1" means that there are some new functions that weren't there at
2143 some point in the past.
2145 PH/51 Error processing for expansion failure of helo_data from an smtp
2146 transport during callout processing was broken.
2148 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2149 tested/used via the -bh/-bhc/-bs options.
2151 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2152 bug, fixed in subsequent PCRE releases).
2154 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2155 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2157 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2163 SC/01 Use a glob alias rather than an array ref in eximstats generated
2164 parser. This improves both readability and performance.
2166 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2167 Don't display local sender or destination tables in eximstats unless
2168 there is data to show.
2169 Added average volumes into the eximstats top table text output.
2171 SC/03 Collect data on the number of addresses (recipients) as well
2172 as the number of messages in eximstats.
2174 TF/01 Correct an error in the documentation for the redirect router. Exim
2175 does (usually) call initgroups() when daemonizing.
2177 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2178 with consistent privilege compared to when running as a daemon.
2180 TF/03 Note in the spec that $authenticated_id is not set for local
2181 submissions from trusted users.
2183 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2184 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2186 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2187 by adding some example configuration directives to the default
2188 configuration file. A little bit of work is required to uncomment the
2189 directives and define how usernames and passwords are checked, but
2190 there is now a framework to start from.
2192 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2193 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2194 without this. I don't know how relevant this is to other LDAP libraries.
2196 PH/02 Add the verb name to the "unknown ACL verb" error.
2198 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2200 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2202 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2203 directory not expanded when it should be if an expanded home directory
2204 was set for the address (which is overridden by the transport).
2206 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2209 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2210 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2211 because it is too late at that time, and has no effect.
2213 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2214 security issue with \' (bugzilla #107). I could not use the
2215 PQescapeStringConn() function, because it needs a PGconn value as one of
2218 PH/08 When testing addresses using -bt, indicate those final addresses that
2219 are duplicates that would not cause an additional delivery. At least one
2220 person was confused, thinking that -bt output corresponded to deliveries.
2221 (Suppressing duplicates isn't a good idea as you lose the information
2222 about possibly different redirections that led to the duplicates.)
2224 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2225 systems where poll() doesn't work, in particular OS X.
2227 PH/10 Added more information to debugging output for retry time not reached.
2229 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2230 operations in malware.c.
2232 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2235 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2236 syslog_duplication set false, log lines that would normally be written
2237 both the the main log and to the reject log were not written to syslog at
2240 PH/14 In the default configuration, change the use of "message" in ACL warn
2241 statements to "add_header".
2243 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2244 not followed by a command (e.g. "seen endif").
2246 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2247 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2250 PH/17 Added extra conditions to the default value of delay_warning_condition
2254 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2255 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2256 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2259 The Auto-Submitted: and various List- headers are standardised, whereas I
2260 don't think Precedence: ever was.
2262 PH/18 Refactored debugging code in route_finduser() to show more information,
2263 in particular, the error code if getpwnam() issues one.
2265 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2266 This is apparently needed in addition to the PH/07 change above to avoid
2267 any possible encoding problems.
2269 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2270 but not after initializing Perl.
2272 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2273 output them only if debugging. By default they are written stderr,
2274 apparently, which is not desirable.
2276 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2279 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2282 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2284 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2285 authenticated or an ident call has been made. Suppress the default
2286 values for $authenticated_id and $authenticated_sender (but permit -oMai
2287 and -oMas) when testing with -bh.
2289 PH/24 Re-jigged the order of the tests in the default configuration so that the
2290 tests for valid domains and recipients precede the DNS black list and CSA
2291 tests, on the grounds that those ones are more expensive.
2293 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2294 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2295 This bug exists in every version of Exim that I still have, right back to
2298 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2299 However, an attempt to turn on case-sensitivity in a regex key by
2300 including (?-i) didn't work because the subject string was already
2301 lowercased, and the effects were non-intuitive. It turns out that a
2302 one-line patch can be used to allow (?-i) to work as expected.
2308 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2309 other effects) broke the use of negated acl sub-conditions.
2311 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2314 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2315 "Deny" causes Exim to reject the incoming connection with a 554 error.
2316 Unfortunately, if there is a major crisis, such as a disk failure,
2317 tcp-wrappers gives "deny", whereas what one would like would be some
2318 kind of temporary error. A kludge has been added to help with this.
2319 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2320 554 error is used if errno is still zero or contains ENOENT (which occurs
2321 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2324 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2326 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2327 errors. Otherwise a message that provokes a temporary error (when other
2328 messages do not) can cause a whole host to time out.
2330 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2331 addresses were routed directly to files or pipes from a redirect router.
2332 File deliveries just didn't batch; pipe deliveries might have suffered
2335 PH/06 A failure to get a lock for a hints database would erroneously always say
2336 "Failed to get write lock", even when it was really a read lock.
2338 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2339 of 0600. This has been changed to use the value of the lockfile_mode
2340 option (which defaults to 0600).
2342 PH/08 Applied small patch from the Sieve maintainer.
2344 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2345 folder from quota calculations, a direct delivery into this folder messed
2346 up the contents of the maildirsize file. This was because the regex was
2347 used only to exclude .Trash (or whatever) when the size of the mailbox
2348 was calculated. There was no check that a delivery was happening into an
2349 excluded directory. This bug has been fixed by ignoring all quota
2350 processing for deliveries into excluded directories.
2352 PH/10 Added the maildirfolder_create_regex option to appendfile.
2358 PH/01 The code for finding all the local interface addresses on a FreeBSD
2359 system running IPv6 was broken. This may well have applied to all BSD
2360 systems, as well as to others that have similar system calls. The broken
2361 code found IPv4 interfaces correctly, but gave incorrect values for the
2362 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2363 that it would not match correctly against @[] and not recognize the IPv6
2366 PH/02 The ipliteral router was not recognizing addresses of the form user@
2367 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2369 PH/03 Added disable_ipv6.
2371 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2372 decoded form, because it is most often used to construct To: headers
2373 lines in autoreplies, and the decoded form may well be syntactically
2374 invalid. However, $reply_address has leading white space removed, and all
2375 newlines turned into spaces so that the autoreply transport does not
2378 PH/05 If group was specified without a user on a router, and no group or user
2379 was specified on a transport, the group from the router was ignored.
2381 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2382 for visible compile-time settings that can be used to change these
2383 numbers, for those that want even more. Backwards compatibility with old
2384 spool files has been maintained. However, going back to a previous Exim
2385 release will lost any variables that are in spool files.
2387 PH/07 Two small changes when running in the test harness: increase delay when
2388 passing a TCP/IP connection to a new process, in case the original
2389 process has to generate a bounce, and remove special handling of
2390 127.0.0.2 (sic), which is no longer necessary.
2392 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2393 be the same on different OS.
2395 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2398 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2399 whether --show-vars was specified or not
2401 JJ/02 exipick: Added support for new ACL variable spool format introduced
2404 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2405 syntactically invalid From: or Reply-to: line, and a filter used this to
2406 generate an autoreply, and therefore failed to obtain an address for the
2407 autoreply, Exim could try to deliver to a non-existent relative file
2408 name, causing unrelated and misleading errors. What now happens is that
2409 it logs this as a hard delivery error, but does not attempt to create a
2412 PH/11 The exinext utility has a -C option for testing purposes, but although
2413 the given file was scanned by exinext itself; it wasn't being passed on
2414 when Exim was called.
2416 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2417 an end-of-file indication when reading a command response.
2419 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2420 compiled. In many other places in Exim, IPv6 addresses are always
2421 recognized, so I have changed this. It also means that IPv4 domain
2422 literals of the form [IPV4:n.n.n.n] are now always recognized.
2424 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2425 used if the router is not running as root, for example, when verifying at
2426 ACL time, or when using -bh. The debugging output from this situation was
2427 non-existent - all you got was a failure to exec. I have made two
2430 (a) Failures to set uid/gid, the current directory, or a process leader
2431 in a subprocess such as that created by queryprogram now generate
2432 suitable debugging ouput when -d is set.
2434 (b) The queryprogram router detects when it is not running as root,
2435 outputs suitable debugging information if -d is set, and then runs
2436 the subprocess without attempting to change uid/gid.
2438 PH/15 Minor change to Makefile for building test_host (undocumented testing
2441 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2442 additional section of a DNS packet that returns MX or SRV records.
2443 Instead, it always explicitly searches for A/AAAA records. This avoids
2444 major problems that occur when a DNS server includes only records of one
2445 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2446 fixed another bug: if SRV records were looked up and the corresponding
2447 address records were *not* found in the additional section, the port
2448 values from the SRV records were lost.
2450 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2451 using the correct key (the original address) when searching the retry
2452 rules in order to find which one to use for generating the retry hint.
2454 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2455 adding the default one. Similarly, if it contains a Reply-To: header, the
2456 errors_reply_to option, if set, is not used.
2458 PH/19 When calculating a retry time, Exim used to measure the "time since
2459 failure" by looking at the "first failed" field in the retry record. Now
2460 it does not use this if it is later than than the arrival time of the
2461 message. Instead it uses the arrival time. This makes for better
2462 behaviour in cases where some deliveries succeed, thus re-setting the
2463 "first failed" field. An example is a quota failure for a huge message
2464 when small messages continue to be delivered. Without this change, the
2465 "time since failure" will always be short, possible causing more frequent
2466 delivery attempts for the huge message than are intended.
2467 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2469 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2470 $1, $2, $3) because the numerical variables can be reset during some
2471 expansion items (e.g. "match"), thereby losing the authentication data.
2473 PH/21 Make -bV show the size of off_t variables so that the test suite can
2474 decide whether to run tests for quotas > 2G.
2476 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2477 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2478 filecount value is greater than 2G or if a quota value is greater than 2G
2479 on a system where the size of off_t is not greater than 4, a panic error
2482 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2483 never match. The debug and -bh output now contains an explicit error
2484 message indicating a malformed IPv4 address or mask.
2486 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2487 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2488 PH/23 above applies.
2490 PH/25 Do not write to syslog when running in the test harness. The only
2491 occasion when this arises is a failure to open the main or panic logs
2492 (for which there is an explicit test).
2494 PH/26 Added the /no_tell option to "control=freeze".
2496 PH/27 If a host name lookup failed very early in a connection, for example, if
2497 the IP address matched host_lookup and the reverse lookup yielded a name
2498 that did not have a forward lookup, an error message of the form "no IP
2499 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2500 could be logged. Now it outputs the IP address instead of "NULL".
2502 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2503 allows the sender and the authenticated sender to be set when
2504 submitting a message from within Exim. Since child_open_exim() is
2505 documented for local_scan(), the new function should be too.
2507 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2508 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2509 results in an empty string is now treated as unset.
2511 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2513 PH/31 Added sender_verify_fail logging option.
2515 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2516 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2517 tidied the source and removed it altogether.
2519 PH/33 When a queue run was abandoned because the load average was too high, a
2520 log line was always written; now it is written only if the queue_run log
2521 selector is set. In addition, the log line for abandonment now contains
2522 information about the queue run such as the pid. This is always present
2523 in "start" and "stop" lines but was omitted from the "abandon" line.
2525 PH/34 Omit spaces between a header name and the colon in the error message that
2526 is given when verify = headers_syntax fails (if there are lots of them,
2527 the message gets confusing).
2529 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2530 names, as there are now some PTR records that contain slashes. This check
2531 is only to protect against broken name servers that fall over on strange
2532 characters, so the fact that it applies to all lookups doesn't matter.
2534 PH/36 Now that the new test suite is complete, we can remove some of the
2535 special code in Exim that was needed for the old test suite. For example,
2536 sorting DNS records because real resolvers return them in an arbitrary
2537 order. The new test suite's fake resolver always returns records in the
2540 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2541 bounces) except when queue_only is set, to avoid logging races between
2542 the different processes.
2544 PH/38 Panic-die if .include specifies a non-absolute path.
2546 PH/39 A tweak to the "H" retry rule from its user.
2548 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2549 a label. They prevented compilation on older perls.
2551 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2552 a warning to be raised on newish perls.
2554 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2555 on queue. Changes to match documented behaviour of showing count of
2556 messages matching specified criteria.
2558 PH/40 Changed the default ident timeout from 30s to 5s.
2560 PH/41 Added support for the use of login_cap features, on those BSD systems
2561 that have them, for controlling the resources used by pipe deliveries.
2563 PH/42 The content-scanning code uses fopen() to create files in which to put
2564 message data. Previously it was not paying any attention to the mode of
2565 the files. Exim runs with umask(0) because the rest of the code creates
2566 files with open(), and sets the required mode explicitly. Thus, these
2567 files were ending up world-writeable. This was not a big issue, because,
2568 being within the spool directory, they were not world-accessible. I have
2569 created a function called modefopen, which takes an additional mode
2570 argument. It sets umask(777), creates the file, chmods it to the required
2571 mode, then resets the umask. All the relevant calls to fopen() in the
2572 content scanning code have been changed to use this function.
2574 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2575 to 24 hours. This avoids potential overflow problems when processing G
2576 and H retry rules. I suspect nobody ever tinkers with this value.
2578 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2580 PH/45 When the plaintext authenticator is running as a client, the server's
2581 challenges are checked to ensure they are valid base64 strings. By
2582 default, the authentication attempt is cancelled if an invalid string is
2583 received. Setting client_ignore_invalid_base64 true ignores these errors.
2584 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2585 they are received. Thus, the responses can be made to depend on the
2586 challenges. If an invalid string is ignored, an empty string is placed in
2589 PH/46 Messages that are created by the autoreply transport now contains a
2590 References: header, in accordance with RFCs 2822 and 3834.
2592 PH/47 Added authenticated_sender_force to the smtp transport.
2594 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2596 PH/49 Installed latest patch from the Sieve maintainer.
2598 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2599 was also set, the appendfile transport was unnecessarily scanning a
2600 directory of message files (e.g. for maildir delivery) to find the count
2601 of files (along with the size), even though it did not need this
2602 information. It now does the scan only if it needs to find either the
2603 size of the count of files.
2605 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2607 PH/52 Two bugs concerned with error handling when the smtp transport is
2610 (i) Exim was not creating retry information for temporary errors given
2611 for individual recipients after the DATA command when the smtp transport
2612 was used in LMTP mode. This meant that they could be retried too
2613 frequently, and not timed out correctly.
2615 (ii) Exim was setting the flag that allows error details to be returned
2616 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2617 recipients that were returned after the DATA command.
2619 PH/53 This is related to PH/52, but is more general: for any failing address,
2620 when detailed error information was permitted to be returned to the
2621 sender, but the error was temporary, then after the final timeout, only
2622 "retry timeout exceeded" was returned. Now it returns the full error as
2623 well as "retry timeout exceeded".
2625 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2626 do this, and (what is worse) MTAs that accept it.
2628 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2629 will now be deprecated.
2631 PH/56 New os.c-cygwin from the Cygwin maintainer.
2633 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2634 formats (previously only available in exim formats via -bpr, -bpru,
2635 and -bpra. Now also available in native and exiqgrep formats)
2637 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2638 with very large, slow to parse queues
2640 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2642 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2644 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2645 responses to authentication challenges, though it was showing the
2646 challenges; (ii) I've removed the CR characters from the debug output for
2649 PH/58 Allow for the insertion of a newline as well as a space when a string
2650 is turned into more than one encoded-word during RFC 2047 encoding. The
2651 Sieve code now uses this.
2653 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2654 data_4xx, lost_connection, tls_required.
2656 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2657 message was being sent as an SMTP response.
2659 PH/61 Add -l and -k options to exicyclog.
2661 PH/62 When verifying, if an address was redirected to one new address, so that
2662 verification continued, and the new address failed or deferred after
2663 having set something in $address_data, the value of $address_data was not
2664 passed back to the ACL. This was different to the case when no
2665 redirection occurred. The value is now passed back in both cases.
2667 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2668 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2669 use login_cap.h, so on its own it isn't the distinguishing feature. The
2670 new name refers directly to the setclassresources() function.
2672 PH/65 Added configuration files for NetBSD3.
2674 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2676 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2677 is preferred over IPv4.
2679 PH/68 The bounce_return_message and bounce_return_body options were not being
2680 honoured for bounces generated during the reception of non-SMTP messages.
2681 In particular, this applied to messages rejected by the ACL. This bug has
2682 been fixed. However, if bounce_return_message is true and bounce_return_
2683 body is false, the headers that are returned for a non-SMTP message
2684 include only those that have been read before the error was detected.
2685 (In the case of an ACL rejection, they have all been read.)
2687 PH/69 The HTML version of the specification is now built in a directory called
2688 spec_html instead of spec.html, because the latter looks like a path with
2689 a MIME-type, and this confuses some software.
2691 PH/70 Catch two compiler warnings in sieve.c.
2693 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2694 function verify_get_ident() calls ip_connect() to connect a socket, but
2695 if the "connect()" function timed out, ip_connect() used to close the
2696 socket. However, verify_get_ident() also closes the socket later, and in
2697 between Exim writes to the log, which may get opened at this point. When
2698 the socket was closed in ip_connect(), the log could get the same file
2699 descriptor number as the socket. This naturally causes chaos. The fix is
2700 not to close the socket in ip_connect(); the socket should be closed by
2701 the function that creates it. There was only one place in the code where
2702 this was missing, in the iplookup router, which I don't think anybody now
2703 uses, but I've fixed it anyway.
2705 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2706 well as to direct DNS lookups. Otherwise the handling of names in host
2707 lists is inconsistent and therefore confusing.
2713 PH/01 Two changes to the default runtime configuration:
2715 (1) Move the checks for relay_from_hosts and authenticated clients from
2716 after to before the (commented out) DNS black list checks.
2718 (2) Add control=submission to the relay_from_hosts and authenticated
2719 clients checks, on the grounds that messages accepted by these
2720 statements are most likely to be submissions.
2722 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2724 (1) Generate an error if the third argument for the ${prvs expansion is
2727 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2730 (3) Reset the variables that are obtained from the first argument of
2731 ${prvscheck and used in the second argument before leaving the code,
2732 because their memory is reclaimed, so using them afterwards may do
2735 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2736 one (it's much easier than Tom thought :-).
2738 (5) Because of (4), we can now allow for the use of $prvscheck_result
2739 inside the third argument.
2741 PH/03 For some reason, the default setting of PATH when running a command from
2742 a pipe transport was just "/usr/bin". I have changed it to
2745 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2746 anything to be listed in the output from -bV.
2748 PH/05 When a filter generated an autoreply, the entire To: header line was
2749 quoted in the delivery log line, like this:
2751 => >A.N.Other <ano@some.domain> <original@ddress> ...
2753 This has been changed so that it extracts the operative address. There
2754 may be more than one such address. If so, they are comma-separated, like
2757 => >ano@some.domain,ona@other.domain <original@ddress> ...
2759 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2760 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2761 not being looked up in the rDNS to get a host name, Exim was showing the
2762 IP address twice in Received: lines, even though the IP addresses were
2763 identical. For example:
2765 Received: from [1.2.3.4] (helo=[1.2.3.4])
2767 However, if the real host name was known, it was omitting the HELO data
2768 if it matched the actual IP address. This has been tidied up so that it
2769 doesn't show the same IP address twice.
2771 PH/07 When both +timestamp and +memory debugging was on, the value given by
2772 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2773 called by the debug printing, thereby overwriting the timestamp buffer.
2774 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2776 PH/08 When the original message was included in an autoreply transport, it
2777 always said "this is a copy of the message, including all the headers",
2778 even if body_only or headers_only was set. It now gives an appropriate
2781 PH/09 Applied a patch from the Sieve maintainer which:
2783 o fixes some comments
2784 o adds the (disabled) notify extension core
2785 o adds some debug output for the result of if/elsif tests
2786 o points to the current vacation draft in the documentation
2787 and documents the missing references header update
2791 o fixes a bug in processing the envelope test (when testing
2792 multiple envelope elements, the last element determinted the
2795 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2796 Electronic Mail") by including:
2798 Auto-submitted: auto-generated
2800 in the messages that it generates (bounce messages and others, such as
2801 warnings). In the case of bounce messages for non-SMTP mesages, there was
2802 also a typo: it was using "Auto_submitted" (underscore instead of
2803 hyphen). Since every message generated by Exim is necessarily in response
2804 to another message, thes have all been changed to:
2806 Auto-Submitted: auto-replied
2808 in accordance with these statements in the RFC:
2810 The auto-replied keyword:
2812 - SHOULD be used on messages sent in direct response to another
2813 message by an automatic process,
2815 - MUST NOT be used on manually-generated messages,
2817 - MAY be used on Delivery Status Notifications (DSNs) and Message
2818 Disposition Notifications (MDNs),
2820 - MUST NOT be used on messages generated by automatic or periodic
2821 processes, except for messages which are automatic responses to
2824 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2825 to the default Received: header definition.
2827 PH/12 Added log selector acl_warn_skipped (default on).
2829 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2830 variables because (a) they are in the wrong storage pool and (b) even if
2831 they were copied, it wouldn't work properly because of the caching.
2833 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2834 checking when decoding. Apparently there are clients that generate
2835 overlong encoded strings. Why am I not surprised?
2837 PH/15 If the first argument of "${if match_address" was not empty, but did not
2838 contain an "@" character, Exim crashed. Now it writes a panic log message
2839 and treats the condition as false.
2841 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2843 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2844 extension "envelope-auth". The code is finished and in agreement with
2845 other implementations, but there is no documentation so far and in fact,
2846 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2847 not changing the active code.
2849 Print executed "if" and "elsif" statements when debugging is used. This
2850 helps a great deal to understand what a filter does.
2852 Document more things not specified clearly in RFC3028. I had all this
2853 sorted out, when out of a sudden new issues came to my mind. Oops."
2855 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2858 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2859 canonical form (as documented). However, after a host name lookup from
2860 the IP address, check_host() was doing a simple string comparison with
2861 addresses acquired from the DNS when checking that the found name did
2862 have the original IP as one of its addresses. Since any found IPv6
2863 addresses are likely to be in abbreviated form, the comparison could
2864 fail. Luckily, there already exists a function for doing the comparison
2865 by converting both addresses to binary, so now that is used instead of
2866 the text comparison.
2868 PH/20 There was another similar case to PH/19, when a complete host name was
2869 given in a host list; looking up its IP address could give an abbreviated
2870 form, whereas the current host's name might or might not be abbreviated.
2871 The same fix has been applied.
2877 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2878 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2881 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2882 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2884 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2886 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2887 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2888 Also, for "Auto-Submitted", treat anything other than "no" as
2889 non-personal, in accordance with RFC 3834. (Previously it treated
2890 anything starting "auto-" as non-personal.)
2892 TF/01 The control=submission/name=... option had a problem with syntax
2893 errors if the name included a slash character. The /name= option
2894 now slurps the rest of the string, so it can include any characters
2895 but it must come last in the list of options (after /sender_retain
2898 PH/05 Some modifications to the interface to the fake nameserver for the new
2906 TK/01 Added the "success_on_redirect" address verification option. See
2907 NewStuff for rationale and an example.
2909 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2911 PH/02 Patch to exigrep to allow it to work on syslog lines.
2913 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2914 fread() to skip over the body file's header line, because in Cygwin the
2915 header line is locked and is inaccessible.
2917 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2918 co-exist for some time) to make it clear that it is the Exim ID that is
2919 referenced, not the Message-ID: header line.
2921 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2922 string_format() function, because snprintf() does not exist on all
2925 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2926 ${stat: expansion item.
2928 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2929 protocol synchronization error", to keep the pedants happy.
2931 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2932 well as for IRIX systems, when gcc is being used. See the host.c source
2935 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2937 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2940 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2941 transport and to the smtp transport in LMTP mode.
2943 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2945 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2946 run a filter in a subprocess. This could lead to confusion in subsequent
2947 lookups in the parent process. There should also be a search_tidyup() at
2948 the end of the subprocess.
2950 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2951 only if the host matched helo_try_verify_hosts, which caused the
2952 verification to occur when the EHLO/HELO command was issued. The ACL just
2953 tested the remembered result. Now, if a previous verification attempt has
2954 not happened, "verify = helo" does it there and then.
2956 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2958 TK/03 Fix log output including CR from clamd.
2960 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2961 exist provoked a memory error which could cause a segfault.
2963 PH/15 Installed PCRE 6.2
2965 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2967 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2968 of the problem. Specifically, suggested +O2 rather than +O1 for the
2971 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2973 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2974 record, the port was not correctly logged when the outgoing_port log
2975 selector was set (it logged the transort's default port).
2977 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2978 fallback_hosts, and "hosts" in the smtp transport.
2980 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2981 host errors such as "Connection refused".
2983 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2984 authentication with radiusclient 0.4.9:
2986 - Error returned from rc_read_config was caught wrongly
2987 - Username/password not passed on to radius server due to wrong length.
2989 The presumption is that some radiusclient API changes for 4.51/PH/17
2990 were not taken care of correctly. The code is still untested by me (my
2991 Linux distribution still has 0.3.2 of radiusclient), but it was
2992 contributed by a Radius user.
2994 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2995 expanding the "port" option of the smtp transport.
2997 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2998 while reading a MIME header. Thanks to Tom Hughes for a patch.
3000 PH/24 Include config.h inside local_scan.h so that configuration settings are
3003 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3004 that in an ACL for RCPT (for example), you can examine exactly what was
3007 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3008 commands, but it was not correctly comparing the address with the actual
3009 client host address. Thus, it would show the EHLO address in Received:
3010 header lines when this was not necessary.
3012 PH/27 Added the % operator to ${eval:}.
3014 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3015 it should be ignoring failures (because with -C, for example, it has lost
3016 privilege). It wasn't ignoring creation failures other than "already
3019 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3022 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3023 because an input line was too long, either on its own, or by virtue of
3024 too many continuations, the temporary file was not being removed, and the
3025 return code was incorrect.
3027 PH/31 Missing "BOOL" in function definition in filtertest.c.
3029 PH/32 Applied Sieve patches from the maintainer.
3031 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3033 PH/33 Added "verify = not_blind".
3035 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3036 Local/Makefile (with some defaults set). These are used in built scripts
3037 such as exicyclog, but they have never been used in the exim_install
3038 script (though there are many overriding facilities there). I have
3039 arranged that the exim_install script now takes note of these two
3042 PH/35 Installed configuration files for Dragonfly.
3044 PH/36 When a locally submitted message by a trusted user did not contain a
3045 From: header, and the sender address was obtained from -f or from an SMTP
3046 MAIL command, and the trusted user did not use -F to supply a sender
3047 name, $originator_name was incorrectly used when constructing a From:
3048 header. Furthermore, $originator_name was used for submission mode
3049 messages from external hosts without From: headers in a similar way,
3050 which is clearly wrong.
3052 PH/37 Added control=suppress_local_fixups.
3054 PH/38 When log_selector = +received_sender was set, and the addition of the
3055 sender made the log line's construction buffer exactly full, or one byte
3056 less than full, an overflow happened when the terminating "\n" was
3059 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3060 when the result of a list match is failure because a DNS lookup failed.
3062 PH/40 RM_COMMAND is now used in the building process.
3064 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3065 the "build-* directories that it finds.
3067 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3068 address in a domain literal was a prefix of an interface address.
3070 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3071 when verifying a sender address, unless rewrite_headers is false.
3073 PH/44 Wrote a long comment about why errors_to addresses are verified as
3074 recipients, not senders.
3076 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3077 the ratelimit ACL was added.
3079 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3081 PH/46 Added extra information about PostgreSQL errors to the error string.
3083 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3084 avoiding the need to install special zones in a real server. This is
3085 backwards compatible; if it can't find the fake resolver, it drops back.
3086 Thus, both old and new test suites can be run.
3088 TF/02 Added util/ratelimit.pl
3090 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3091 clock is set back in time.
3093 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3094 Candler <B.Candler@pobox.com>.
3096 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3097 OK for addresses that are the result of redirections.
3099 PH/48 A number of further additions for the benefit of the new test suite,
3100 including a fake gethostbyname() that interfaces to the fake DNS resolver
3103 TF/06 The fix for widen_domains has also been applied to qualify_single and
3104 search_parents which are the other dnslookup options that can cause
3107 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3110 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3112 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3113 service" if it encountered an MX record with an empty target hostname.
3114 The message is now "an MX or SRV record indicated no SMTP service".
3116 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3117 if the DNS of the sending site is misconfigured. This is quite a
3118 common situation. This change restores the behaviour of treating a
3119 helo verification defer as a failure.
3121 PH/51 If self=fail was set on a router, the bounce message did not include the
3122 actual error message.
3128 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3130 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3131 command itself ended in error, the underlying message about the transport
3132 filter timeout was being overwritten with the pipe command error. Now the
3133 underlying error message should be appended to the second error message.
3135 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3137 PH/02 Reduce the amount of output that "make" produces by default. Full output
3138 can still be requested.
3140 PH/03 The warning log line about a condition test deferring for a "warn" verb
3141 was being output only once per connection, rather than after each
3142 occurrence (because it was using the same function as for successful
3143 "warn" verbs). This seems wrong, so I have changed it.
3145 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3146 it should not have, which might have caused a crash in the right
3147 circumstances, but probably never did.
3149 PH/04 Installed a modified version of Tony Finch's patch to make submission
3150 mode fix the return path as well as the Sender: header line, and to
3151 add a /name= option so that you can make the user's friendly name appear
3154 TF/03 Added the control = fakedefer ACL modifier.
3156 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3157 Mark Lowes for thorough testing.
3159 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3161 TK/03 Merged latest SRS patch from Miles Wilton.
3163 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3164 with the definition in sysexits.h (which is #included earlier).
3165 Fortunately, Exim does not actually use EX_OK. The code used to try to
3166 preserve the sysexits.h value, by assumimg that macro definitions were
3167 scanned for macro replacements. I have been disabused of this notion,
3168 so now the code just undefines EX_OK before #including unistd.h.
3170 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3171 in the smtp transport. When a block could not be written in a single
3172 write() function, the timeout was being re-applied to each part-write.
3173 This seems wrong - if the receiver was accepting one byte at a time it
3174 would take for ever. The timeout is now adjusted when this happens. It
3175 doesn't have to be particularly precise.
3177 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3178 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3180 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3181 without a domain if the check to <postmaster@domain> fails.
3183 SC/01 Eximstats: added -xls and the ability to specify output files
3184 (patch written by Frank Heydlauf).
3186 SC/02 Eximstats: use FileHandles for outputing results.
3188 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3190 SC/04 Eximstats: fixed display of large numbers with -nvr option
3192 SC/05 Eximstats: fixed merging of reports with empty tables.
3194 SC/06 Eximstats: added the -include_original_destination flag
3196 SC/07 Eximstats: removed tabs and trailing whitespace.
3198 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3200 TK/06 MBOX spool code: Add real "From " MBOX separator line
3201 so the .eml file is really in mbox format (even though
3202 most programs do not really care). Patch from Alex Miller.
3204 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3205 The latter is generated from $received_to and is only set if the
3206 message has one envelope recipient. SA can use these headers,
3207 obviously out-of-the-box. Patch from Alex Miller.
3209 PH/08 The ${def test on a variable was returning false if the variable's
3210 value was "0", contrary to what the specification has always said!
3211 The result should be true unless the variable is empty.
3213 PH/09 The syntax error of a character other than { following "${if
3214 def:variable_name" (after optional whitespace) was not being diagnosed.
3215 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3216 accidental colon was present, for example, could give incorrect results.
3218 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3219 result is used (not including appendfile, where other changes are about
3222 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3223 This involved changing a lot of size variables from int to off_t. It
3224 should work with maildirs and everything.
3226 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3227 spamd dying while we are connected to it.
3229 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3232 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3233 function for BDB 4.3.
3235 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3237 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3238 the sources for PCRE have been re-arranged and more files are now
3241 PH/15 The code I had for printing potentially long long variables in PH/11
3242 above was not the best (it lost precision). The length of off_t variables
3243 is now inspected at build time, and an appropriate printing format (%ld
3244 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3245 to be "long long int" or "long int". This is needed for the internal
3246 formatting function string_vformat().
3248 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3249 the configuration file to be ":syslog", then the script "guesses" where
3250 the logs files are, rather than using the compiled in default. In our
3251 case the guess is not the same as the compiled default, so the script
3252 suddenly stopped working when I started to use syslog. The patch checks
3253 to see if log_file_path is "". If so, it attempts to read it from exim
3254 with no configuration file to get the compiled in version, before it
3255 falls back to the previous guessing code."
3257 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3258 implementing BATV in an Exim configuration. See NewStuff for the gory
3261 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3262 Makefile that are specific to HP-UX.
3264 PH/18 If the "use_postmaster" option was set for a recipient callout together
3265 with the "random" option, the postmaster address was used as the MAIL
3266 FROM address for the random test, but not for the subsequent recipient
3267 test. It is now used for both.
3269 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3270 patch removes a few documentation additions to RFC 3028, because the
3271 latest draft now contains them. It adds the new en;ascii-case comparator
3272 and a new error check for 8bit text in MIME parts. Comparator and
3273 require names are now matched exactly. I enabled the subaddress
3274 extension, but it is not well tested yet (read: it works for me)."
3276 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3277 rework some of the code of TK/09 above to avoid the hardwired use of
3278 "%lld" and "long long". Replaced the call to snprintf() with a call to
3281 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3282 records point to non-existent hosts", "retry timeout exceeded", and
3283 "retry time not reached for any host after a long failure period".
3285 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3286 experimental DomainKeys support:
3288 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3289 (2) On an error such as an illegally used "control", the wrong name for
3290 the control was given.
3292 These problems did NOT occur unless DomainKeys support was compiled.
3294 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3296 PH/24 Added ${if match_ip condition.
3298 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3299 databases so that it will be absolutely obvious if a crash occurs in the
3300 DB library. This is a regular occurrence (often caused by mis-matched
3303 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3304 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3305 picked up on a user's system that detects such things. There doesn't seem
3306 to be a gcc warning option for this - only an attribute that has to be
3307 put on the function's prototype. It seems that in Fedora Core 4 they have
3308 set this on a number of new functions. No doubt there will be more in due
3311 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3312 specify a transport. However, if an address that was verified by such a
3313 router was the subject of a callout, Exim crashed because it tried to
3314 read the rcpt_include_affixes from the non-existent transport. Now it
3315 just assumes that the setting of that option is false. This bug was
3316 introduced by 4.51/PH/31.
3318 PH/28 Changed -d+all to exclude +memory, because that information is very
3319 rarely of interest, but it makes the output a lot bigger. People tend to
3320 do -d+all out of habit.
3322 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3323 code in os-type was giving problems when libc.so lives in lib64, like on
3326 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3327 aren't the modern standard, and it seems that some systems' include files
3328 don't always have them. Exim was already checking for some of the newer
3329 ones like T_AAAA, and defining it itself. I've added checks for all the
3330 record types that Exim uses.
3332 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3333 not automatically generating a new one, as it is supposed to. This
3334 prevented TLS from working. If the file did exist, but contained invalid
3335 data, a new version was generated, as expected. It was only the case of a
3336 non-existent file that was broken.
3338 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3339 with a change in libdomainkeys > 0.64.
3341 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3342 from DNS. If the selector record carries the flag, it now has
3343 precedence over the domain-wide flag.
3345 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3347 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3348 the use of a port name that isn't defined in /etc/services) occurred, the
3349 message was deferred as in a normal delivery, and thus remained on the
3350 spool, instead of being failed because of the mua_wrapper setting. This
3351 is now fixed, and I tidied up some of the mua_wrapper messages at the
3354 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3355 the messages in a hash of arrays rather than using individual hashes.
3356 This is a bit cleaner and results in dramatic memory savings, albeit
3357 at a slight CPU cost.
3359 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3360 as requested by Marc Sherman.
3362 SC/10 Eximstats: added histograms for user specified patterns as requested
3365 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3367 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3368 fopen() in the content-scanning modules that did not already have it.
3374 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3375 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3377 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3379 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3381 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3382 to test Sieve filters that use "vacation".
3384 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3385 that changes the way the GnuTLS parameters are stored in the cache file.
3386 The new format can be generated externally. For backward compatibility,
3387 if the data in the cache doesn't make sense, Exim assumes it has read an
3388 old-format file, and it generates new data and writes a new file. This
3389 means that you can't go back to an older release without removing the
3392 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3393 work if there are any delivery delays because "one_time" forces the
3394 parent to be marked "delivered", so its unseen clone is never tried
3395 again. For this reason, Exim now forbids the simultaneous setting of
3398 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3399 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3400 entry if you want to know the details. The fix, however, neglected to
3401 consider the case when local delivery batching is involved. The test for
3402 "previously delivered" was not happening when checking to see if an
3403 address could be batched with a previous (undelivered) one; under
3404 certain circumstances this could lead to multiple deliveries to the same
3407 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3408 in its include files, and this causes problems building Exim.
3410 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3411 header_syntax) but Exim was just ignoring anything given after a slash.
3412 In particular, this caused confusion with an attempt to use "verify =
3413 reverse_host_lookup/defer_ok". An error is now given when options are
3414 supplied for verify items that do not have them. (Maybe reverse_host_
3415 lookup should have a defer_ok option, but that's a different point.)
3417 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3418 defined by RFC 821) to 2048, because there were problems with some AUTH
3419 commands, and RFC 1869 says the size should be increased for extended
3420 SMTP commands that take arguments.
3422 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3425 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3426 "unknown" error; now it says that the functionality isn't in the binary.
3428 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3429 an address' error message when a string expansion fails (syntax or
3430 whatever). Otherwise the password may appear in the log. Following change
3431 PH/42 below, there is no longer a chance of it appearing in a bounce
3434 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3436 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3437 of its addresses was ever tried. (Bugzilla bug #2.)
3439 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3440 the result incorrectly in the debug output. (It correctly added a newline
3441 to what was transported.)
3443 TF/01 Added $received_time.
3445 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3446 commented out examples of how to interface to a virus scanner and to
3447 SpamAssassin. Also added commented examples of av_scanner and
3448 spamd_address settings.
3450 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3451 and controls are allowed in which ACLs. There were a couple of minor
3452 errors. Some of the entries in the conditions table (which is a table of
3453 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3454 negation of where the condition IS allowed.
3456 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3458 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3459 header file does not have a version number, so I've had to invent a new
3460 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3461 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3462 radiusclient), but it was contributed by a Radius user.
3464 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3465 files or directories, for OpenSSL.
3467 PH/19 When an Exim process that is running as root has to create an Exim log
3468 file, it does so in a subprocess that runs as exim:exim so as to get the
3469 ownership right at creation (otherwise, other Exim processes might see
3470 the file with the wrong ownership). There was no test for failure of this
3471 fork() call, which would lead to the process getting stuck as it waited
3472 for a non-existent subprocess. Forks do occasionally fail when resources
3473 run out. I reviewed all the other calls to fork(); they all seem to check
3476 PH/20 When checking for unexpected SMTP input at connect time (before writing
3477 the banner), Exim was not dealing correctly with a non-positive return
3478 from the read() function. If the client had disconnected by this time,
3479 the result was a log entry for a synchronization error with an empty
3480 string after "input=" when read() returned zero. If read() returned -1
3481 (an event I could not check), uninitialized data bytes were printed.
3482 There were reports of junk text (parts of files, etc) appearing after
3485 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3487 PH/22 Added support for macro redefinition, and (re)definition in between
3488 driver and ACL definitions.
3490 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3491 forgetting to use the resulting value; it was using the unexpanded value.
3493 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3494 hadn't been configured. The fix is from Juergen Kreileder, who
3495 understands it better than I do:
3497 "Here's what I see happening with three configured cyrus_sasl
3498 authenticators configured (plain, login, cram-md5):
3500 On startup auth_cyrus_sasl_init() gets called for each of these.
3501 This means three calls to sasl_listmech() without a specified mech_list.
3502 => SASL tests which mechs of all available mechs actually work
3503 => three warnings about OTP not working
3504 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3506 With the patch, sasl_listmech() also gets called three times. But now
3507 SASL's mech_list option is set to the server_mech specified in the the
3508 authenticator. Or in other words, the answer from sasl_listmech()
3509 gets limited to just the mech you're testing for (which is different
3511 => the return list contains just 'plain' or 'login', 'cram-md5' or
3512 nothing depending on the value of ob->server_mech.
3514 I've just tested the patch: Authentication still works fine,
3515 unavailable mechs specified in the exim configuration are still
3516 caught, and the auth.log warnings about OTP are gone."
3518 PH/25 When debugging is enabled, the contents of the command line are added
3519 to the debugging output, even when log_selector=+arguments is not
3522 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3523 answer is "GNU", and only if the return is "GNU/something" is the answer
3526 PH/27 $acl_verify_message is now set immediately after the failure of a
3527 verification in an ACL, and so is available in subsequent modifiers. In
3528 particular, the message can be preserved by coding like this:
3530 warn !verify = sender
3531 set acl_m0 = $acl_verify_message
3533 Previously, $acl_verify_message was set only while expanding "message"
3534 and "log_message" when a very denied access.
3536 PH/28 Modified OS/os.c-Linux with
3538 -#ifndef OS_LOAD_AVERAGE
3539 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3541 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3542 nomenclature these days.)
3544 PH/29 Installed patch from the Sieve maintainer that adds the options
3545 sieve_useraddress and sieve_subaddress to the redirect router.
3547 PH/30 In these circumstances:
3548 . Two addresses routed to the same list of hosts;
3549 . First host does not offer TLS;
3550 . First host accepts first address;
3551 . First host gives temporary error to second address;
3552 . Second host offers TLS and a TLS session is established;
3553 . Second host accepts second address.
3554 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3555 and peerdn, if requested) that were in fact used only for the second
3558 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3559 attention to any local part prefix or suffix that was matched by the
3560 router that accepted the address. It now behaves in the same way as it
3561 does for delivery: the affixes are removed from the local part unless
3562 rcpt_include_affixes is set on the transport.
3564 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3565 timeout during the DATA phase of an incoming message.
3567 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3568 applied a patch sanctioned by the Sieve maintainer.
3570 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3571 the uid or gid is negative. A case of a negative gid caused this to be
3572 noticed. The fix allows for either to be negative.
3574 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3575 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3578 PH/36 The change PH/12 above was broken. Fixed it.
3580 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3581 the grounds that routing the same address twice would always produce the
3582 same answer. This might have been true once, but it is certainly no
3583 longer true now. Routing a child address may depend on the previous
3584 routing that produced that child. Some complicated redirection strategies
3585 went wrong when messages had multiple recipients, and made Exim's
3586 behaviour dependent on the order in which the addresses were given.
3588 I have moved the duplicate checking until after the routing is complete.
3589 Exim scans the addresses that are assigned to local and remote
3590 transports, and removes any duplicates. This means that more work will be
3591 done, as duplicates will always all be routed, but duplicates are
3592 presumably rare, so I don't expect this is of any significance.
3594 For deliveries to pipes, files, and autoreplies, the duplicate checking
3595 still happens during the routing process, since they are not going to be
3598 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3599 It corrects a timeout issue with spamd. This is Ian's comment: "The
3600 background is that sometimes spamd either never reads data from a
3601 connection it has accepted, or it never writes response data. The exiscan
3602 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3603 blindly assumes that writes won't block so it may never time out."
3605 PH/39 Allow G after quota size as well as K and M.
3607 PH/40 The value set for $authenticated_id in an authenticator may not contain
3608 binary zeroes or newlines because the value is written to log lines and
3609 to spool files. There was no check on this. Now the value is run through
3610 the string_printing() function so that such characters are converted to
3611 printable escape sequences.
3613 PH/41 $message_linecount is a new variable that contains the total number of
3614 lines in the message. Compare $body_linecount, which is the count for the
3617 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3618 bounce and delay warning messages, except in certain special cases, which
3621 (a) An SMTP error message from a remote host;
3622 (b) A message specified in a :fail: redirection;
3623 (c) A message specified in a "fail" command in a system filter;
3624 (d) A message specified in a FAIL return from the queryprogram router;
3625 (e) A message specified by the cannot_route_message router option.
3627 In these cases only, Exim does include the error details in bounce and
3628 warning messages. There are also a few cases where bland messages such
3629 as "unrouteable address" or "local delivery error" are given.
3631 PH/43 $value is now also set for the "else" part of a ${run expansion.
3633 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3634 being worked on, but at least Exim now implements the latest version to
3637 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3638 process to complete was treated as a delivery failure, a timeout while
3639 writing the message to the pipe was logged, but erroneously treated as a
3640 successful delivery. Such timeouts include transport filter timeouts. For
3641 consistency with the overall process timeout, these timeouts are now
3642 treated as errors, giving rise to delivery failures by default. However,
3643 there is now a new Boolean option for the pipe transport called
3644 timeout_defer, which, if set TRUE, converts the failures into defers for
3645 both kinds of timeout. A transport filter timeout is now identified in
3648 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3649 systems where "make" and "gmake" are different, calling "gmake" at top
3650 level broke things. I've arranged for the value of $(MAKE) to be passed
3651 from the Makefile to this script so that it can call the same version of
3655 A note about Exim versions 4.44 and 4.50
3656 ----------------------------------------
3658 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3659 changes of various kinds. As a consequence, a big documentation update was
3660 needed. This delayed the release for rather longer than seemed good, especially
3661 in the light of a couple of (minor) security issues. Therefore, the changes
3662 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3663 release. So 4.44 and 4.50 are in effect two different branches that both start
3666 I have left the 4.50 change log unchanged; it contains all the changes since
3667 4.43. The change log for 4.44 is below; many of its items are identical to
3668 those for 4.50. This seems to be the most sensible way to preserve the
3669 historical information.
3675 1. Minor wording change to the doc/README.SIEVE file.
3677 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3678 computation of the current number of files was incorrect.
3680 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3681 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3684 4. Give more explanation in the error message when the command for a transport
3685 filter fails to execute.
3687 5. There are several places where Exim runs a non-Exim command in a
3688 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3689 was being done only for the command run by the queryprogram router. It is
3690 now done for all such subprocesses. The other cases are: ${run, transport
3691 filters, and the commands run by the lmtp and pipe transports.
3693 6. Added CONFIGURE_GROUP build-time option.
3695 7. Some older OS have a limit of 256 on the maximum number of file
3696 descriptors. Exim was using setrlimit() to set 1000 as a large value
3697 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3698 systems. I've change it so that if it can't get 1000, it tries for 256.
3700 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3701 was an oversight, and furthermore, ever since the addition of extra
3702 controls (e.g. 4.43/32), the checks on when to allow different forms of
3703 "control" were broken. There should now be diagnostics for all cases when a
3704 control that does not make sense is encountered.
3706 9. Added the /retain_sender option to "control=submission".
3708 10. $recipients is now available in the predata ACL (oversight).
3710 11. Tidy the search cache before the fork to do a delivery from a message
3711 received from the command line. Otherwise the child will trigger a lookup
3712 failure and thereby defer the delivery if it tries to use (for example) a
3713 cached ldap connection that the parent has called unbind on.
3715 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3716 of $address_data from the recipient verification was clobbered by the
3717 sender verification.
3719 13. The value of address_data from a sender verification is now available in
3720 $sender_address_data in subsequent conditions in the ACL statement.
3722 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3724 15. Added a new option "connect=<time>" to callout options, to set a different
3727 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3728 was its contents. (It was OK if the option was not defined at all.)
3730 17. A "Completed" log line is now written for messages that are removed from
3731 the spool by the -Mrm option.
3733 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3734 information about exactly what failed.
3736 19. Added -dd to debug only the daemon process.
3738 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3739 handles timeouts, both on the server side and network timeouts. Renamed the
3740 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3742 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3743 It is now set to "smtps".
3745 22. $host_address is now set to the target address during the checking of
3746 ignore_target_hosts.
3748 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3749 being passed; this would have caused $sender_host_name to have been used if
3750 matching the list had actually called for a host name (not very likely,
3751 since this list is usually IP addresses). A host name is now passed as
3754 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3755 code that specifies a non-restarting handler (typically sigaction() in
3756 modern systems) in an attempt to fix a rare and obscure crash bug.
3758 25. Narrowed the window for a race in the daemon that could cause it to ignore
3759 SIGCHLD signals. This is not a major problem, because they are used only to
3760 wake it up if nothing else does.
3762 26. A malformed maildirsize file could cause Exim to calculate negative values
3763 for the mailbox size or file count. Odd effects could occur as a result.
3764 The maildirsize information is now recalculated if the size or filecount
3767 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3768 support for a long time. Removed HAVE_SYS_VFS_H.
3770 28. Installed the latest version of exipick from John Jetmore.
3772 29. In an address list, if the pattern was not a regular expression, an empty
3773 subject address (from a bounce message) matched only if the pattern was an
3774 empty string. Non-empty patterns were not even tested. This was the wrong
3775 because it is perfectly reasonable to use an empty address as part of a
3776 database query. An empty address is now tested by patterns that are
3777 lookups. However, all the other forms of pattern expect the subject to
3778 contain a local part and a domain, and therefore, for them, an empty
3779 address still always fails if the pattern is not itself empty.
3781 30. Exim went into a mad DNS loop when attempting to do a callout where the
3782 host was specified on an smtp transport, and looking it up yielded more
3783 than one IP address.
3785 31. Re-factored the code for checking spool and log partition space into a
3786 function that finds that data and another that does the check. The former
3787 is then used to implement four new variables: $spool_space, $log_space,
3788 $spool_inodes, and $log_inodes.
3790 32. The RFC2047 encoding function was originally intended for short strings
3791 such as real names; it was not keeping to the 75-character limit for
3792 encoded words that the RFC imposes. It now respects the limit, and
3793 generates multiple encoded words if necessary. To be on the safe side, I
3794 have increased the buffer size for the ${rfc2047: expansion operator from
3797 33. It is now permitted to omit both strings after an "if" condition; if the
3798 condition is true, the result is "true". As before, when the second string
3799 is omitted, a false condition yields an empty string. This makes it less
3800 cumbersome to write custom ACL and router conditions.
3802 34. Failure to deliver a bounce message always caused it to be frozen, even if
3803 there was an errors_to setting on the router. The errors_to setting is now
3806 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3807 canonical form (fully expanded) before being placed in
3808 $sender_host_address.
3810 36. The table in the code that translates DNS record types into text (T_A to
3811 "A" for instance) was missing entries for NS and CNAME. It is just possible
3812 that this could have caused confusion if both these types were looked up
3813 for the same domain, because the text type is used as part of Exim's
3814 per-process caching. But the chance of anyone hitting this buglet seems
3817 37. The dnsdb lookup has been extended in a number of ways.
3819 (1) There is a new type, "zns", which walks up the domain tree until it
3820 finds some nameserver records. It should be used with care.
3822 (2) There is a new type, "mxh", which is like "mx" except that it returns
3823 just the host names, not the priorities.
3825 (3) It is now possible to give a list of domains (or IP addresses) to be
3826 looked up. The behaviour when one of the lookups defers can be
3827 controlled by a keyword.
3829 (4) It is now possible to specify the separator character for use when
3830 multiple records are returned.
3832 38. The dnslists ACL condition has been extended: it is now possible to supply
3833 a list of IP addresses and/or domains to be looked up in a particular DNS
3836 39. Added log_selector=+queue_time_overall.
3838 40. When running the queue in the test harness, wait just a tad after forking a
3839 delivery process, to get repeatability of debugging output.
3841 41. Include certificate and key file names in error message when GnuTLS fails
3842 to set them up, because the GnuTLS error message doesn't include the name
3843 of the failing file when there is a problem reading it.
3845 42. Allow both -bf and -bF in the same test run.
3847 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3849 44. The "Exiscan patch" is now merged into the mainline Exim source.
3851 45. Sometimes the final signoff response after QUIT could fail to get
3852 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3853 < 0 before doing a fflush(). This bug looks as though it goes back to the
3854 introduction of TLS in release 3.20, but "sometimes" must have been rare
3855 because the tests only now provoked it.
3857 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3858 (this can affect the format of dates).
3860 47. exim_tidydb, when checking for the continued existence of a message for
3861 which it has found a message-specific retry record, was not finding
3862 messages that were in split spool directories. Consequently, it was
3863 deleting retry records that should have stayed in existence.
3865 48. Steve fixed some bugs in eximstats.
3867 49. The SPA authentication driver was not abandoning authentication and moving
3868 on to the next authenticator when an expansion was forced to fail,
3869 contradicting the general specification for all authenticators. Instead it
3870 was generating a temporary error. It now behaves as specified.
3872 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3873 (the order specifies the preference for clients). The order is now AES256,
3874 AES128, 3DES, ARCFOUR128.
3876 51. Small patch to Sieve code - explicitly set From: when generating an
3879 52. Exim crashed if a remote delivery caused a very long error message to be
3880 recorded - for instance if somebody sent an entire SpamAssassin report back
3881 as a large number of 550 error lines. This bug was coincidentally fixed by
3882 increasing the size of one of Exim's internal buffers (big_buffer) that
3883 happened as part of the Exiscan merge. However, to be on the safe side, I
3884 have made the code more robust (and fixed the comments that describe what
3887 53. Now that there can be additional text after "Completed" in log lines (if
3888 the queue_time_overall log selector is set), a one-byte patch to exigrep
3889 was needed to allow it to recognize "Completed" as not the last thing in
3892 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3893 patch that reportedly fixes this has been added. I am not expert enough to
3894 create a test for it. This is what the patch creator wrote:
3896 "I found a little strange behaviour of ldap code when working with
3897 Windows 2003 AD Domain, where users was placed in more than one
3898 Organization Units. When I tried to give exim partial DN, the exit code
3899 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3900 But simultaneously result of request was absolutely normal ldap result,
3901 so I produce this patch..."
3903 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3904 REFERENCE, so I have modified the code to exclude the patch when that macro
3907 55. Some experimental protocols are using DNS PTR records for new purposes. The
3908 keys for these records are domain names, not reversed IP addresses. The
3909 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3910 leaves it alone. Component reversal etc. now happens only for IP addresses.
3913 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3915 57. Double the size of the debug message buffer (to 2048) so that more of very
3916 long debug lines gets shown.
3918 58. The exicyclog utility now does better if the number of log files to keep
3919 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3921 59. Two changes related to the smtp_active_hostname option:
3923 (1) $smtp_active_hostname is now available as a variable.
3924 (2) The default for smtp_banner uses $smtp_active_hostname instead
3925 of $primary_hostname.
3927 60. The host_aton() function is supposed to be passed a string that is known
3928 to be a valid IP address. However, in the case of IPv6 addresses, it was
3929 not checking this. This is a hostage to fortune. Exim now panics and dies
3930 if the condition is not met. A case was found where this could be provoked
3931 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3932 components; fortuitously, this particular loophole had already been fixed
3933 by change 4.50/55 above.
3935 If there are any other similar loopholes, the new check in host_aton()
3936 itself should stop them being exploited. The report I received stated that
3937 data on the command line could provoke the exploit when Exim was running as
3938 exim, but did not say which command line option was involved. All I could
3939 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3940 running as the user.
3943 61. There was a buffer overflow vulnerability in the SPA authentication code
3944 (which came originally from the Samba project). I have added a test to the
3945 spa_base64_to_bits() function which I hope fixes it.
3948 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3949 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3951 63. The daemon start-up calls getloadavg() while still root for those OS that
3952 need the first call to be done as root, but it missed one case: when
3953 deliver_queue_load_max is set with deliver_drop_privilege. This is
3954 necessary for the benefit of the queue runner, because there is no re-exec
3955 when deliver_drop_privilege is set.
3957 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3958 This has been fixed.
3960 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3961 was in use, was not putting the data itself into the right store pool;
3962 consequently, it could be overwritten for a subsequent message in the same
3963 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3966 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3968 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3969 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3970 the function were treating the return as a boolean value, which happened to
3971 work because 0=false and not-0=true, but is not correct code.
3973 68. The host_aton() function was not handling scoped IPv6 addresses (those
3974 with, for example, "%eth0" on the end) correctly.
3976 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3977 negated items (that is, ~something) in unsigned ints. Some compilers
3978 apparently mutter when there is no cast.
3980 70. If an address verification called from an ACL failed, and did not produce a
3981 user-specific message (i.e. there was only a "system" message), nothing was
3982 put in $acl_verify_message. In this situation, it now puts the system
3985 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3986 session; change 4.31/43 added the unwanted input to the log line - except
3987 that it did not do this in the start of session case. It now does.
3989 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3990 This is wrong and can cause the other end to generate a synchronization
3991 error if it is another Exim or anything else that does the synchronization
3992 check. A QUIT command is no longer sent after a timeout.
3994 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3995 during host lookups.
3997 74. The defer_ok option of callout verification was not working if it was used
3998 when verifying addresses in header lines, that is, for this case:
4000 verify = header_sender/callout=defer_ok
4002 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4003 those file descriptors could be used for SMTP connections. If anything
4004 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4005 could be sent to the SMTP client, causing chaos. The daemon now opens
4006 stdin, stdout, and stderr to /dev/null when it puts itself into the
4009 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4010 log by default. The user can override this with suitable Perl magic.
4012 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4013 the log message when discard triggers, was not working for the DATA ACL or
4014 for the non-SMTP ACL.
4016 78. Error message wording change in sieve.c.
4018 79. If smtp_accept_max_per_host was set, the number of connections could be
4019 restricted to fewer than expected, because the daemon was trying to set up
4020 a new connection before checking whether the processes handling previous
4021 connections had finished. The check for completed processes is now done
4022 earlier. On busy systems, this bug wouldn't be noticed because something
4023 else would have woken the daemon, and it would have reaped the completed
4026 80. If a message was submitted locally by a user whose login name contained one
4027 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4028 It caused a spool format error. I have fixed the spool reading code. A
4029 related problem was that the "from" clause in the Received: line became
4030 illegal because of the space(s). It is now covered by ${quote_local_part.
4032 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4035 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4037 83. Patch from Sieve maintainer to fix unterminated string problem in
4038 "vacation" handling.
4040 84. Some minor changes to the Linux configuration files to help with other
4041 OS variants using glibc.
4043 85. One more patch for Sieve to update vacation handling to latest spec.
4046 ----------------------------------------------------
4047 See the note above about the 4.44 and 4.50 releases.
4048 ----------------------------------------------------
4054 1. Change 4.43/35 introduced a bug that caused file counts to be
4055 incorrectly computed when quota_filecount was set in an appendfile
4058 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4059 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4062 3. Give more explanation in the error message when the command for a transport
4063 filter fails to execute.
4065 4. There are several places where Exim runs a non-Exim command in a
4066 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4067 was being done only for the command run by the queryprogram router. It is
4068 now done for all such subprocesses. The other cases are: ${run, transport
4069 filters, and the commands run by the lmtp and pipe transports.
4071 5. Some older OS have a limit of 256 on the maximum number of file
4072 descriptors. Exim was using setrlimit() to set 1000 as a large value
4073 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4074 systems. I've change it so that if it can't get 1000, it tries for 256.
4076 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4077 was an oversight, and furthermore, ever since the addition of extra
4078 controls (e.g. 4.43/32), the checks on when to allow different forms of
4079 "control" were broken. There should now be diagnostics for all cases when a
4080 control that does not make sense is encountered.
4082 7. $recipients is now available in the predata ACL (oversight).
4084 8. Tidy the search cache before the fork to do a delivery from a message
4085 received from the command line. Otherwise the child will trigger a lookup
4086 failure and thereby defer the delivery if it tries to use (for example) a
4087 cached ldap connection that the parent has called unbind on.
4089 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4090 of $address_data from the recipient verification was clobbered by the
4091 sender verification.
4093 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4094 was its contents. (It was OK if the option was not defined at all.)
4096 11. A "Completed" log line is now written for messages that are removed from
4097 the spool by the -Mrm option.
4099 12. $host_address is now set to the target address during the checking of
4100 ignore_target_hosts.
4102 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4103 being passed; this would have caused $sender_host_name to have been used if
4104 matching the list had actually called for a host name (not very likely,
4105 since this list is usually IP addresses). A host name is now passed as
4108 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4109 code that specifies a non-restarting handler (typically sigaction() in
4110 modern systems) in an attempt to fix a rare and obscure crash bug.
4112 15. Narrowed the window for a race in the daemon that could cause it to ignore
4113 SIGCHLD signals. This is not a major problem, because they are used only to
4114 wake it up if nothing else does.
4116 16. A malformed maildirsize file could cause Exim to calculate negative values
4117 for the mailbox size or file count. Odd effects could occur as a result.
4118 The maildirsize information is now recalculated if the size or filecount
4121 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4122 support for a long time. Removed HAVE_SYS_VFS_H.
4124 18. Updated exipick to current release from John Jetmore.
4126 19. Allow an empty sender to be matched against a lookup in an address list.
4127 Previously the only cases considered were a regular expression, or an
4130 20. Exim went into a mad DNS lookup loop when doing a callout where the
4131 host was specified on the transport, if the DNS lookup yielded more than
4134 21. The RFC2047 encoding function was originally intended for short strings
4135 such as real names; it was not keeping to the 75-character limit for
4136 encoded words that the RFC imposes. It now respects the limit, and
4137 generates multiple encoded words if necessary. To be on the safe side, I
4138 have increased the buffer size for the ${rfc2047: expansion operator from
4141 22. Failure to deliver a bounce message always caused it to be frozen, even if
4142 there was an errors_to setting on the router. The errors_to setting is now
4145 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4146 canonical form (fully expanded) before being placed in
4147 $sender_host_address.
4149 24. Updated eximstats to version 1.33
4151 25. Include certificate and key file names in error message when GnuTLS fails
4152 to set them up, because the GnuTLS error message doesn't include the name
4153 of the failing file when there is a problem reading it.
4155 26. Expand error message when OpenSSL has problems setting up cert/key files.
4158 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4159 (this can affect the format of dates).
4161 28. exim_tidydb, when checking for the continued existence of a message for
4162 which it has found a message-specific retry record, was not finding
4163 messages that were in split spool directories. Consequently, it was
4164 deleting retry records that should have stayed in existence.
4166 29. eximstats updated to version 1.35
4167 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4168 1.35 - bugfix such that pie charts by volume are generated correctly
4170 30. The SPA authentication driver was not abandoning authentication and moving
4171 on to the next authenticator when an expansion was forced to fail,
4172 contradicting the general specification for all authenticators. Instead it
4173 was generating a temporary error. It now behaves as specified.
4175 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4176 (the order specifies the preference for clients). The order is now AES256,
4177 AES128, 3DES, ARCFOUR128.
4179 31. Small patch to Sieve code - explicitly set From: when generating an
4182 32. Exim crashed if a remote delivery caused a very long error message to be
4183 recorded - for instance if somebody sent an entire SpamAssassin report back
4184 as a large number of 550 error lines. This bug was coincidentally fixed by
4185 increasing the size of one of Exim's internal buffers (big_buffer) that
4186 happened as part of the Exiscan merge. However, to be on the safe side, I
4187 have made the code more robust (and fixed the comments that describe what
4190 33. Some experimental protocols are using DNS PTR records for new purposes. The
4191 keys for these records are domain names, not reversed IP addresses. The
4192 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4193 leaves it alone. Component reversal etc. now happens only for IP addresses.
4196 34. The host_aton() function is supposed to be passed a string that is known
4197 to be a valid IP address. However, in the case of IPv6 addresses, it was
4198 not checking this. This is a hostage to fortune. Exim now panics and dies
4199 if the condition is not met. A case was found where this could be provoked
4200 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4201 components; fortuitously, this particular loophole had already been fixed
4202 by change 4.50/55 or 4.44/33 above.
4204 If there are any other similar loopholes, the new check in host_aton()
4205 itself should stop them being exploited. The report I received stated that
4206 data on the command line could provoke the exploit when Exim was running as
4207 exim, but did not say which command line option was involved. All I could
4208 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4209 running as the user.
4212 35. There was a buffer overflow vulnerability in the SPA authentication code
4213 (which came originally from the Samba project). I have added a test to the
4214 spa_base64_to_bits() function which I hope fixes it.
4217 36. The daemon start-up calls getloadavg() while still root for those OS that
4218 need the first call to be done as root, but it missed one case: when
4219 deliver_queue_load_max is set with deliver_drop_privilege. This is
4220 necessary for the benefit of the queue runner, because there is no re-exec
4221 when deliver_drop_privilege is set.
4223 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4224 was in use, was not putting the data itself into the right store pool;
4225 consequently, it could be overwritten for a subsequent message in the same
4226 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4229 38. Sometimes the final signoff response after QUIT could fail to get
4230 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4231 < 0 before doing a fflush(). This bug looks as though it goes back to the
4232 introduction of TLS in release 3.20, but "sometimes" must have been rare
4233 because the tests only now provoked it.
4239 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4240 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4241 Now it returns an integer. A number of calls were still expecting a T/F
4242 return. Fortuitously, in all cases, the tests worked in OK situations,
4243 which is the norm. However, things would have gone wrong on any write
4244 failures on the smtp file descriptor. This function is used when sending
4245 messages over SMTP and also when doing verify callouts.
4247 2. When Exim is called to do synchronous delivery of a locally submitted
4248 message (the -odf or -odi options), it no longer closes stderr before doing
4251 3. Implemented the mua_wrapper option.
4253 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4255 5. Implemented the functions header_remove(), header_testname(),
4256 header_add_at_position(), and receive_remove_recipient(), and exported them
4259 6. If an ACL "warn" statement specified the addition of headers, Exim already
4260 inserted X-ACL-Warn: at the start if there was no header name. However, it
4261 was not making this test for the second and subsequent header lines if
4262 there were newlines in the string. This meant that an invalid header could
4263 be inserted if Exim was badly configured.
4265 7. Allow an ACL "warn" statement to add header lines at the start or after all
4266 the Received: headers, as well as at the end.
4268 8. Added the rcpt_4xx retry error code.
4270 9. Added postmaster_mailfrom=xxx to callout verification option.
4272 10. Added mailfrom=xxxx to the callout verification option, for verify=
4275 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4276 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4278 12. Inserted some casts to stop certain compilers warning when using pointer
4279 differences as field lengths or precisions in printf-type calls (mostly
4280 affecting debugging statements).
4282 13. Added optional readline() support for -be (dynamically loaded).
4284 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4285 same clock tick as a message's arrival, so that its received time was the
4286 same as the "first fail" time on the retry record, and that message
4287 remained on the queue past the ultimate address timeout, every queue runner
4288 would try a delivery (because it was past the ultimate address timeout) but
4289 after another failure, the ultimate address timeout, which should have then
4290 bounced the address, did not kick in. This was a "< instead of <=" error;
4291 in most cases the first failure would have been in the next clock tick
4292 after the received time, and all would be well.
4294 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4295 being recognized when the domain list was tested by the match_domain
4296 condition in an expansion string.
4298 16. Added the ${str2b64: operator.
4300 17. Exim was always calling setrlimit() to set a large limit for the number of
4301 processes, without checking whether the existing limit was already
4302 adequate. (It did check for the limit on file descriptors.) Furthermore,
4303 errors from getrlimit() and setrlimit() were being ignored. Now they are
4304 logged to the main and panic logs, but Exim does carry on, to try to do its
4305 job under whatever limits there are.
4307 18. Imported PCRE 5.0.
4309 19. Trivial typo in log message " temporarily refused connection" (the leading
4312 20. If the log selector return_path_on_delivery was set and an address was
4313 redirected to /dev/null, the delivery process crashed because it assumed
4314 that a return path would always be set for a "successful" delivery. In this
4315 case, the whole delivery is bypassed as an optimization, and therefore no
4318 21. Internal re-arrangement: the function for sending a challenge and reading
4319 a response while authentication was assuming a zero-terminated challenge
4320 string. It's now changed to take a pointer and a length, to allow for
4321 binary data in such strings.
4323 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4325 23. Exim was not respecting finduser_retries when seeking the login of the
4326 uid under which it was called; it was always trying 10 times. (The default
4327 setting of finduser_retries is zero.) Also, it was sleeping after the final
4328 failure, which is pointless.
4330 24. Implemented tls_on_connect_ports.
4332 25. Implemented acl_smtp_predata.
4334 26. If the domain in control=submission is set empty, Exim assumes that the
4335 authenticated id is a complete email address when it generates From: or
4336 Sender: header lines.
4338 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4339 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4340 chown and chgrp in /bin and hostname in /usr/bin.
4342 28. Exim was keeping the "process log" file open after each use, just as it
4343 does for the main log. This opens the possibility of it remaining open for
4344 long periods when the USR1 signal hits a daemon. Occasional processlog
4345 errors were reported, that could have been caused by this. Anyway, it seems
4346 much more sensible not to leave this file open at all, so that is what now
4349 29. The long-running daemon process does not normally write to the log once it
4350 has entered its main loop, and it closes the log before doing so. This is
4351 so that log files can straightforwardly be renamed and moved. However,
4352 there are a couple of unusual error situations where the daemon does write
4353 log entries, and I had neglected to close the log afterwards.
4355 30. The text of an SMTP error response that was received during a remote
4356 delivery was being truncated at 512 bytes. This is too short for some of
4357 the long messages that one sometimes sees. I've increased the limit to
4360 31. It is now possible to make retry rules that apply only when a message has a
4361 specific sender, in particular, an empty sender.
4363 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4364 it possible to be selective about when SMTP synchronization is enforced.
4366 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4368 32. Implemented hosts_connection_nolog.
4370 33. Added an ACL for QUIT.
4372 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4375 35. Added mailbox_size and mailbox_filecount to appendfile.
4377 36. Added control = no_multiline_responses to ACLs.
4379 37. There was a bug in the logic of the code that waits for the clock to tick
4380 in the case where the clock went backwards by a substantial amount such
4381 that the microsecond fraction of "now" was more than the microsecond
4382 fraction of "then" (but the whole seconds number was less).
4384 38. Added support for the libradius Radius client library this is found on
4385 FreeBSD (previously only the radiusclient library was supported).
4391 1. When certain lookups returned multiple values in the form name=value, the
4392 quoting of the values was not always being done properly. Specifically:
4393 (a) If the value started with a double quote, but contained no whitespace,
4395 (b) If the value contained whitespace other than a space character (i.e.
4396 tabs or newlines or carriage returns) it was not quoted.
4397 This fix has been applied to the mysql and pgsql lookups by writing a
4398 separate quoting function and calling it from the lookup code. The fix
4399 should probably also be applied to nisplus, ibase and oracle lookups, but
4400 since I cannot test any of those, I have not disturbed their existing code.
4402 2. A hit in the callout cache for a specific address caused a log line with no
4403 reason for rejecting RCPT. Now it says "Previous (cached) callout
4404 verification failure".
4406 3. There was an off-by-one bug in the queryprogram router. An over-long
4407 return line was truncated at 256 instead of 255 characters, thereby
4408 overflowing its buffer with the terminating zero. As well as fixing this, I
4409 have increased the buffer size to 1024 (and made a note to document this).
4411 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4412 when Exim is waiting for an SMTP response from a remote server, Exim
4413 restarts its select() call on the socket, thereby resetting its timeout.
4414 This is not a problem when such interrupts are rare. Somebody set up a cron
4415 job to run exiwhat every 2 minutes, which is less than the normal select()
4416 timeout (5 or 10 minutes). This meant that the select() timeout never
4417 kicked in because it was always reset. I have fixed this by comparing the
4418 time when an interrupt arrives with the time at the start of the first call
4419 to select(). If more time than the timeout has elapsed, the interrupt is
4420 treated as a timeout.
4422 5. Some internal re-factoring in preparation for the addition of Sieve
4423 extensions (by MH). In particular, the "personal" test is moved to a
4424 separate function, and given an option for scanning Cc: and Bcc: (which is
4425 not set for Exim filters).
4427 6. When Exim created an email address using the login of the caller as the
4428 local part (e.g. when creating a From: or Sender: header line), it was not
4429 quoting the local part when it contained special characters such as @.
4431 7. Installed new OpenBSD configuration files.
4433 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4434 try to make them clearer.
4436 9. Callout options, other than the timeout value, were being ignored when
4437 verifying sender addresses in header lines. For example, when using
4439 verify = header_sender/callout=no_cache
4441 the cache was (incorrectly) being used.
4443 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4444 only the Cygwin environment.
4446 11. When return_path_on_delivery was set as a log selector, if different remote
4447 addresses in the same message used different return paths and parallel
4448 remote delivery occurred, the wrong values would sometimes be logged.
4449 (Whenever a remote delivery process finished, the return path value from
4450 the most recently started remote delivery process was logged.)
4452 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4453 lines when AUTH and/or TLS are in use. This is the "received protocol"
4454 field. Exim used to use "asmtp" for authenticated SMTP, without any
4455 indication (in the protocol name) for TLS use. Now it follows the RFC and
4456 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4457 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4458 names appear in log lines as well as in Received: header lines.
4460 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4461 extensions, and comparison tests, and to fix some bugs.
4463 14. Changes to the "personal" filter test:
4465 (1) The test was buggy in that it was just doing the equivalent of
4466 "contains" tests on header lines. For example, if a user's address was
4467 anne@some.where, the "personal" test would incorrectly be true for
4469 To: susanne@some.where
4471 This test is now done by extracting each address from the header in turn,
4472 and checking the entire address. Other tests that are part of "personal"
4473 are now done using regular expressions (for example, to check local parts
4474 of addresses in From: header lines).
4476 (2) The list of non-personal local parts in From: addresses has been
4477 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4478 taken from the Sieve specification recommendations.
4480 (3) If the message contains any header line starting with "List-" it is
4481 treated as non-personal.
4483 (4) The test for "circular" in the Subject: header line has been removed
4484 because it now seems ill-conceived.
4486 15. Minor typos in src/EDITME comments corrected.
4488 16. Installed latest exipick from John Jetmore.
4490 17. If headers_add on a router specified a text string that was too long for
4491 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4492 of string_sprintf() is now avoided.
4494 18. $message_body_size was not set (it was always zero) when running the DATA
4495 ACL and the local_scan() function.
4497 19. For the "mail" command in an Exim filter, no default was being set for
4498 the once_repeat time, causing a random time value to be used if "once" was
4499 specified. (If the value happened to be <= 0, no repeat happened.) The
4500 default is now 0s, meaning "never repeat". The "vacation" command was OK
4501 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4502 (I found it when inspecting the code).
4504 20. There is now an overall timeout for performing a callout verification. It
4505 defaults to 4 times the callout timeout, which applies to individual SMTP
4506 commands during the callout. The overall timeout applies when there is more
4507 than one host that can be tried. The timeout is checked before trying the
4508 next host. This prevents very long delays if there are a large number of
4509 hosts and all are timing out (e.g. when the network connections are timing
4510 out). The value of the overall timeout can be changed by specifying an
4511 additional sub-option for "callout", called "maxwait". For example:
4513 verify = sender/callout=5s,maxwait=20s
4515 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4516 to the end before writing, but this should make it even safer).
4518 22. Exim was forgetting that it had advertised PIPELINING for the second and
4519 subsequent messages on an SMTP connection. It was also not resetting its
4520 memory on STARTTLS and an internal HELO.
4522 23. When Exim logs an SMTP synchronization error within a session, it now
4523 records whether PIPELINING has been advertised or not.
4525 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4526 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4527 rather than long int.
4529 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4531 26. Added the never_mail option to autoreply.
4537 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4538 crash if the getsockname() call failed; this can happen if a connection is
4539 closed very soon after it is established. The problem was simply in the
4540 order in which certain operations were done, causing Exim to try to write
4541 to the SMTP stream before it had set up the file descriptor. The bug has
4542 been fixed by making things happen in the correct order.
4548 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4549 before the connection was closed, thus losing the rejection response.
4551 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4552 some early Solaris releases, but causes trouble in current releases where
4553 socklen_t is defined.
4555 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4558 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4561 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4562 that releasing the top of it at the end releases what was used for sub-
4563 expansions (unless the block got too big). However, discard this block if
4564 the first thing is a variable or header, so that we can use its block when
4565 it is dynamic (useful for very large $message_headers, for example).
4567 6. Lookups now cache *every* query, not just the most recent. A new, separate
4568 store pool is used for this. It can be recovered when all lookup caches are
4569 flushed. Lookups now release memory at the end of their result strings.
4570 This has involved some general refactoring of the lookup sources.
4572 7. Some code has been added to the store_xxx() functions to reduce the amount
4573 of flapping under certain conditions.
4575 8. log_incoming_interface used to affect only the <= reception log lines. Now
4576 it causes the local interface and port to be added to several more SMTP log
4577 lines, for example "SMTP connection from", and rejection lines.
4579 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4581 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4583 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4584 could be overwritten at the end of the current message (or the start of a
4585 new message if it was set in a HELO ACL). The value is now preserved for
4586 the duration of the SMTP connection.
4588 12. If a transport had a headers_rewrite setting, and a matching header line
4589 contained an unqualified address, that address was qualified, even if it
4590 did not match any rewriting rules. The underlying bug was that the values
4591 of the flags that permit the existence of unqualified sender and recipient
4592 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4593 non-local messages, and by -bnq for local messages) were not being
4594 preserved with the message after it was received.
4596 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4597 "next input=" as part of the text comprising the host identity instead of
4598 the correct text. The code was using the same buffer for two different
4599 strings. However, depending on which order the printing function evaluated
4600 its arguments, the bug did not always show up. Under Linux, for example, my
4601 test suite worked just fine.
4603 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4604 doesn't work with some older versions of Perl. It has been changed to "my",
4605 which in any case is probably the better facility to use.
4607 15. A really picky compiler found some instances of statements for creating
4608 error messages that either had too many or two few arguments for the format
4611 16. The size of the buffer for calls to the DNS resolver has been increased
4612 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4613 for addresses that have a lot of PTR records. This alleviates a problem; it
4614 does not fully solve it.
4616 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4617 buffer now truncates the list and logs the incident, which is the same
4618 action as happens when Exim is looking up a host name and its aliases.
4619 Previously in this situation something unpredictable would happen;
4620 sometimes it was "internal error: store_reset failed".
4622 18. If a server dropped the connection unexpectedly when an Exim client was
4623 using GnuTLS and trying to read a response, the client delivery process
4624 crashed while trying to generate an error log message.
4626 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4627 string, for example:
4629 warn message = H1: something\nH2: something
4631 the text was added as a single header line from Exim's point of view
4632 though it ended up OK in the delivered message. However, searching for the
4633 second and subsequent header lines using $h_h2: did not work. This has been
4634 fixed. Similarly, if a system filter added multiple headers in this way,
4635 the routers could not see them.
4637 20. Expanded the error message when iplsearch is called with an invalid key to
4638 suggest using net-iplsearch in a host list.
4640 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4641 ACLs are no longer actually imposed (and a message to that effect is
4644 22. If a "gecos" field in a passwd entry contained escaped characters, in
4645 particular, if it contained a \" sequence, Exim got it wrong when building
4646 a From: or a Sender: header from that name. A second bug also caused
4647 incorrect handling when an unquoted " was present following a character
4648 that needed quoting.
4650 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4651 was not being matched caselessly.
4653 24. Arranged for all hyphens in the exim.8 source to be escaped with
4656 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4657 properly. Recipient callout cache records were still being keyed to include
4658 the sender, even when use_sender was set false. This led to far more
4659 callouts that were necessary. The sender is no longer included in the key
4660 when use_sender is false.
4662 26. Added "control = submission" modifier to ACLs.
4664 27. Added the ${base62d: operator to decode base 62 numbers.
4666 28. dnsdb lookups can now access SRV records.
4668 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4669 the configuration file.
4671 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4672 (-v) mode. This makes the output for a verbose queue run more intelligible.
4674 31. Added a use_postmaster feature to recipient callouts.
4676 32. Added the $body_zerocount variable, containing the number of binary zero
4677 bytes in the message body.
4679 33. The time of last modification of the "new" subdirectory is now used as the
4680 "mailbox time last read" when there is a quota error for a maildir
4683 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4685 35. Added +ignore_unknown as a special item in host lists.
4687 36. Code for decoding IPv6 addresses in host lists is now included, even if
4688 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4689 address was recognized as an IP address, but was then not correctly decoded
4690 into binary, causing unexpected and incorrect effects when compared with
4697 1. Very minor rewording of debugging text in manualroute to say "list of
4698 hosts" instead of "hostlist".
4700 2. If verify=header_syntax was set, and a header line with an unqualified
4701 address (no domain) and a large number of spaces between the end of the
4702 name and the colon was received, the reception process suffered a buffer
4703 overflow, and (when I tested it) crashed. This was caused by some obsolete
4704 code that should have been removed. The fix is to remove it!
4706 3. When running in the test harness, delay a bit after writing a bounce
4707 message to get a bit more predictability in the log output.
4709 4. Added a call to search_tidyup() just before forking a reception process. In
4710 theory, someone could use a lookup in the expansion of smtp_accept_max_
4711 per_host which, without the tidyup, could leave open a database connection.
4713 5. Added the variables $recipient_data and $sender_data which get set from a
4714 lookup success in an ACL "recipients" or "senders" condition, or a router
4715 "senders" option, similar to $domain_data and $local_part_data.
4717 6. Moved the writing of debug_print from before to after the "senders" test
4720 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4721 problems for message scanning, either using a data ACL, or using
4722 local_scan() because the Received: header was not generated till after they
4723 were called (in order to set the time as the time of reception completion).
4724 I have revised the way this works. The header is now generated after the
4725 body is received, but before the ACL or local_scan() are called. After they
4726 are run, the timestamp in the header is updated.
4732 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4733 before starting a queue runner without re-exec. This happened only when
4734 deliver_drop_privilege was set or when the Exim user was set to root. The
4735 effect of the bug was that timeouts during subsequent deliveries caused
4736 crashes instead of being properly handled. The handler is now left at its
4737 default (and expected) setting.
4739 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4740 message, again when deliver_drop_privilege is set or Exim is run as root.
4741 The bug described in (1) was not present in this case, but the tidying up
4742 of the other signals was missing. I have made the two cases consistent.
4744 3. The ignore_target_hosts setting on a manualroute router was being ignored
4745 for hosts that were looked up using the /MX notation.
4747 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4750 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4751 operated on the sender address. After changing the $sender_address to <>
4752 for the sender address verify, Exim was re-instated it as the original
4753 (before rewriting) address, but remembering that it had rewritten it, so it
4754 wasn't rewriting it again. This bug also had the effect of breaking the
4755 sender address verification caching when the sender address was rewritten.
4757 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4758 This has been changed so that if the ip literal address matches
4759 ignore_target_hosts, the router declines.
4761 7. Added expansion conditions match_domain, match_address, and match_local_
4762 part (NOT match_host).
4764 8. The placeholder for the Received: header didn't have a length field set.
4766 9. Added code to Exim itself and to exim_lock to test for a specific race
4767 condition that could lead to file corruption when using MBX delivery. The
4768 issue is with the lockfile that is created in /tmp. If this file is removed
4769 after a process has opened it but before that process has acquired a lock,
4770 there is the potential for a second process to recreate the file and also
4771 acquire a lock. This could lead to two Exim processes writing to the file
4772 at the same time. The added code performs the same test as UW imapd; it
4773 checks after acquiring the lock that its file descriptor still refers to
4774 the same named file.
4776 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4777 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4780 11. Added the smtp_active_hostname option. If used, this will typically be made
4781 to depend on the incoming interface address. Because $interface_address is
4782 not set up until the daemon has forked a reception process, error responses
4783 that can happen earlier (such as "too many connections") no longer contain
4786 12. If an expansion in a condition on a "warn" statement fails because a lookup
4787 defers, the "warn" statement is abandoned, and the next ACL statement is
4788 processed. Previously this caused the whole ACL to be aborted.
4790 13. Added the iplsearch lookup type.
4792 14. Added ident_timeout as a log selector.
4794 15. Added tls_certificate_verified as a log selector.
4796 16. Added a global option tls_require_ciphers (compare the smtp transport
4797 option of the same name). This controls incoming TLS connections.
4799 17. I finally figured out how to make tls_require_ciphers do a similar thing
4800 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4801 before starting the TLS session.
4803 18. Tabs are now shown as \t in -bP output.
4805 19. If the log selector return_path_on_delivery was set, Exim crashed when
4806 bouncing a message because it had too many Received: header lines.
4808 20. If two routers both had headers_remove settings, and the first one included
4809 a superfluous trailing colon, the final name in the first list and the
4810 first name in the second list were incorrectly joined into one item (with a
4811 colon in the middle).
4817 1. Added -C and -D options to the exinext utility, mainly to make it easier
4818 to include in the automated testing, but these could be helpful when
4819 multiple configurations are in use.
4821 2. The exinext utility was not formatting the output nicely when there was
4822 an alternate port involved in the retry record key, nor when there was a
4823 message id as well (for retries that were specific to a specific message
4824 and a specific host). It was also confused by IPv6 addresses, because of
4825 the additional colons they contain. I have fixed the IPv4 problem, and
4826 patched it up to do a reasonable job for IPv6.
4828 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4829 delivery, the log line now contains "pipelined" if PIPELINING was used.
4831 4. An SMTP transport process used to panic and die if the bind() call to set
4832 an explicit outgoing interface failed. This has been changed; it is now
4833 treated in the same way as a connect() failure.
4835 5. A reference to $sender_host_name in the part of a conditional expansion
4836 that was being skipped was still causing a DNS lookup. This no longer
4839 6. The def: expansion condition was not recognizing references to header lines
4840 that used bh_ and bheader_.
4842 7. Added the _cache feature to named lists.
4844 8. The code for checking quota_filecount in the appendfile transport was
4845 allowing one more file than it should have been.
4847 9. For compatibility with Sendmail, the command line option
4855 and sets the incoming protocol and host name (for trusted callers). The
4856 host name and its colon can be omitted when only the protocol is to be set.
4857 Note the Exim already has two private options, -pd and -ps, that refer to
4858 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4859 "s", but I don't think that's a major issue.
4861 10. A number of refactoring changes to the code, none of which should affect
4864 (a) The number of logging options was getting close to filling up the
4865 32-bit word that was used as a bit map. I have split them into two classes:
4866 those that are passed in the argument to log_write(), and those that are
4867 only ever tested independently outside of that function. These are now in
4868 separate 32-bit words, so there is plenty of room for expansion again.
4869 There is no change in the user interface or the logging behaviour.
4871 (b) When building, for example, log lines, the code previously used a
4872 macro that called string_cat() twice, in order to add two strings. This is
4873 not really sufficiently general. Furthermore, there was one instance where
4874 it was actually wrong because one of the argument was used twice, and in
4875 one call a function was used. (As it happened, calling the function twice
4876 did not affect the overall behaviour.) The macro has been replaced by a
4877 function that can join an arbitrary number of extra strings onto a growing
4880 (c) The code for expansion conditions now uses a table and a binary chop
4881 instead of a serial search (which was left over from when there were very
4882 few conditions). Also, it now recognizes conditions like "pam" even when
4883 the relevant support is not compiled in: a suitably worded error message is
4884 given if an attempt is made to use such a condition.
4886 11. Added ${time_interval:xxxxx}.
4888 12. A bug was causing one of the ddress fields not to be passed back correctly
4889 from remote delivery subprocesses. The field in question was not being
4890 subsequently used, so this caused to problems in practice.
4892 13. Added new log selectors queue_time and deliver_time.
4894 14. Might have fixed a bug in maildirsizefile handling that threw up
4895 "unexpected character" debug warnings, and recalculated the data
4896 unnecessarily. In any case, I expanded the warning message to give more
4899 15. Added the message "Restricted characters in address" to the statements in
4900 the default ACL that block characters like @ and % in local parts.
4902 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4903 Three changes have been made:
4905 (a) There was a serious bug; a negative response to MAIL caused the whole
4906 recipient domain to be cached as invalid, thereby blocking all messages
4907 to all local parts at the same domain, from all senders. This bug has
4908 been fixed. The domain is no longer cached after a negative response to
4909 MAIL if the sender used is not empty.
4911 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4914 (c) A new callout option, "use_sender" has been added for people who want
4915 the modified behaviour.
4921 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4924 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4925 indeed breaks things for older releases.
4927 3. Added additional logging to the case where there is a problem reading data
4928 from a filter that is running in a subprocess using a pipe, in order to
4929 try to track down a specific problem.
4931 4. Testing facility fudge: when running in the test harness and attempting
4932 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4933 getting "No route to host". Convert this to a timeout.
4935 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4938 6. Some OS don't have socklen_t but use size_t instead. This affects the
4939 fifth argument of getsockopt() amongst other things. This is now
4940 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4941 can be set for individual OS. I have set it for SunOS5, OSF1, and
4942 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4943 some earlier ones do not.
4945 7. Change 4.30/15 was not doing the test caselessly.
4947 8. The standard form for an IPv6 address literal was being rejected by address
4948 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4949 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4950 this, as well as the form without the "IPv6" on the front (but only when
4951 address literals are enabled, of course).
4953 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4955 10. Exim crashed if a message with an empty sender address specified by -f
4956 encountered a router with an errors_to setting. This could be provoked only
4957 by a command such as
4961 where an empty string was supplied; "<>" did not hit this bug.
4963 11. Installed PCRE release 4.5.
4965 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4966 remained set. It is now erased.
4968 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4969 times from message ids (which are base 36 rather than the normal 62).
4971 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4972 were being counted as actual protocol errors, and logged if the log
4973 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4974 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4975 and DATA following a set of rejected RCPTs do not count as protocol errors.
4976 In other words, Exim assumes they were pipelined, though this may not
4977 actually be the case. Of course, in all cases the client gets an
4978 appropriate error code.
4980 15. If a lookup fails in an ACL condition, a message about the failure may
4981 be available; it is used if testing the ACL cannot continue, because most
4982 such messages specify what the cause of the deferral is. However, some
4983 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4984 that caused an old message to be retained and used if a later statement
4985 caused a defer, replacing the real cause of the deferral.
4987 16. If an IP address had so many PTR records that the DNS lookup buffer
4988 was not large enough to hold them, Exim could crash while trying to process
4989 the truncated data. It now detects and logs this case.
4991 17. Further to 4.21/58, another change has been made: if (and only if) the
4992 first line of a message (the first header line) ends with CRLF, a bare LF
4993 in a subsequent header line has a space inserted after it, so as not to
4994 terminate the header.
4996 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4997 unnecessarily, used atoi() instead of strtol(), and didn't check the
4998 termination when getting file sizes from file names by regex.
5000 19. Completely re-implemented the support for maildirsize files, in the light
5001 of a number of problems with the previous contributed implementation
5002 (4.30/29). In particular:
5004 . If the quota is zero, the maildirsize file is maintained, but no quota is
5007 . If the maildir directory does not exist, it is created before any attempt
5008 to write a maildirsize file.
5010 . The quota value in the file is just a cache; if the quota is changed in
5011 the transport, the new value overrides.
5013 . A regular expression is available for excluding directories from the
5016 20. The autoreply transport checks the characters in options that define the
5017 message's headers; it allows continued headers, but it was checking with
5018 isspace() after an embedded newline instead of explicitly looking for a
5021 21. If all the "regular" hosts to which an address was routed had passed their
5022 expiry times, and had not reached their retry times, the address was
5023 bounced, even if fallback hosts were defined. Now Exim should go on to try
5026 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5027 equivalent code in the SMTP transport. Some hosts send humungous responses
5028 to HELO/EHLO, more than 1024 it seems.
5030 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5031 gives compiler warnings in some environments. I've now done it "properly",
5034 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5035 (because of problems with the built-in one) was declared to return uschar *
5036 instead of char *, causing compiler failure.
5038 25. Fixed a file descriptor leak when processing alias/forward files.
5040 26. Fixed a minor format string issue in dbfn.c.
5042 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5044 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5045 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5048 29. When a sender address is verified, it is cached, to save repeating the test
5049 when there is more than one recipient in a message. However, when the
5050 verification involves a callout, it is possible for different callout
5051 options to be set for different recipients. It is too complicated to keep
5052 track of this in the cache, so now Exim always runs a verification when a
5053 callout is required, relying on the callout cache for the optimization.
5054 The overhead is duplication of the address routing, but this should not be
5057 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5058 to be verified with callout=postmaster, and the main callout worked but the
5059 postmaster check failed, the verification correctly failed. However, if a
5060 subsequent RCPT command asked for sender verification *without* the
5061 postmaster check, incorrect caching caused this verification also to fail,
5064 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5065 it was not caching the DNS options (qualify_single, search_parents) that
5066 were used when the lookup failed. A subsequent lookup with different
5067 options therefore always gave the same answer, though there were cases
5068 where it should not have. (Example: a "domains = !$mx_any" option on a
5069 dnslookup router: the "domains" option is always processed without any
5070 widening, but the router might have qualify_single set.) Now Exim uses the
5071 cached value only when the same options are set.
5073 32. Added John Jetmore's "exipick" utility to the distribution.
5075 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5076 than a timeout (e.g. a certificate is required, and is not provided), an
5077 Exim server now closes the connection immediately. Previously it waited for
5078 the client to close - but if the client is SSL, it seems that they each
5079 wait for each other, leading to a delay before one of them times out.
5081 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5082 maintained 0.8.x compatibility because I don't think many are using it, and
5083 it is clearly obsolete.
5085 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5088 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5089 client certificate was expired. A simple patch fixes this, though I don't
5090 understand the full logic of why the verify callback is called multiple
5093 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5094 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5095 which causes problems with some clients (such as the Certicom SSL Plus
5096 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5097 disables the coutermeasure allowing Eudora to connect."
5099 38. Exim was not checking that a write() to a log file succeeded. This could
5100 lead to Bad Things if a log got too big, in particular if it hit a file
5101 size limit. Exim now panics and dies if it cannot write to a log file, just
5102 as it does if it cannot open a log file.
5104 39. Modified OS/Makefile-Linux so that it now contains
5106 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5108 The two -D definitions ensure that Exim is compiled with large file
5109 support, which makes it possible to handle log files that are bigger than
5112 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5113 instance) a domain was checked against a named list that involved a lookup,
5114 causing $domain_data to be set, then another domain was checked against the
5115 same list, then the first domain was re-checked, the value of $domain_data
5116 after the final check could be wrong. In particular, if the second check
5117 failed, it could be set empty. This bug probably also applied to
5120 41. The strip_trailing_dot option was not being applied to the address given
5121 with the -f command-line option.
5123 42. The code for reading a message's header from the spool was incrementing
5124 $received_count, but never initializing it. This meant that the value was
5125 incorrect (doubled) while delivering a message in the same process in which
5126 it was received. In the most common configuration of Exim, this never
5127 happens - a fresh exec is done - but it can happen when
5128 deliver_drop_privilege is set.
5130 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5131 it now includes up to 150 characters of the unexpected data in the log
5134 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5135 and building data strings. The size of both of these buffers was 10 000
5136 bytes - far larger than anybody would *ever* want, thought I. Needless to
5137 say, somebody hit the limit. I have increased the maximum line length to
5138 20 000 and the maximum data length of concatenated lines to 100 000. I have
5139 also fixed two bugs, because there was no checking on these buffers. Tsk,
5140 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5141 buffer is too small.
5143 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5144 lsearch lookups. Now it does.
5146 46. When parsing a route_list item in a manualroute router, a fixed-length
5147 buffer was used for the list of hosts. I made this 1024 bytes long,
5148 thinking that nobody would ever have a list of hosts that long. Wrong.
5149 Somebody had a whole pile of complicated expansion conditions, and the
5150 string was silently truncated, leading to an expansion error. It turns out
5151 that it is easier to change to an unlimited length (owing to other changes
5152 that have happened since this code was originally written) than to build
5153 structure for giving a limitation error. The length of the item that
5154 expands into the list of hosts is now unlimited.
5156 47. The lsearch lookup could not handle data where the length of text line was
5157 more than 4095 characters. Such lines were truncated, leading to shortened
5158 data being returned. It should now handle lines of any length.
5160 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5161 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5164 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5165 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5166 no longer happens in comment lines. A list of replacements is now placed
5167 at the head of all of the source files, except those whose only change is
5168 to replace PERL_COMMAND in the very first #! line.
5170 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5171 messages on the queue, with a bottom-up merge sort, using code contributed
5172 by Michael Haardt. This should make operations like -bp somewhat faster on
5173 large queues. It won't affect queue runners, except when queue_run_in_order
5176 51. Installed eximstats 1.31 in the distribution.
5178 52. Added support for SRV lookups to the dnslookup router.
5180 53. If an ACL referred to $message_body or $message_body_end, the value was not
5181 reset for any messages that followed in the same SMTP session.
5183 54. The store-handling optimization for building very long strings was not
5184 differentiating between the different store pools. I don't think this
5185 actually made any difference in practice, but I've tidied it.
5187 55. While running the routers to verify a sender address, $sender_address
5188 was still set to the sender address. This is wrong, because when routing to
5189 send a bounce to the sender, it would be empty. Therefore, I have changed
5190 it so that, while verifying a sender address, $sender_address is set to <>.
5191 (There is no change to what happens when verifying a recipient address.)
5193 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5194 target A or AAAA records (if not already returned) without resetting the
5195 qualify_single or search_parents options of the DNS resolver. These are
5196 inappropriate in this case because the targets of MX and SRV records must
5197 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5198 target that, when qualified, matched something in the local domain. These
5199 two options are now turned off when doing these lookups.
5201 57. It seems that at least some releases of Reiserfs (which does not have the
5202 concept of a fixed number of inodes) returns zero and not -1 for the
5203 number of available inodes. This interacted badly with check_spool_inodes,
5204 which assumed that -1 was the "no such thing" setting. What I have done is
5205 to check that the total number of inodes is greater than zero before doing
5206 the test of how many are available.
5208 58. When a "warn" ACL statement has a log_message modifier, the message is
5209 remembered, and not repeated. This is to avoid a lot of repetition when a
5210 message has many recipients that cause the same warning to be written.
5211 Howewer, Exim was preserving the list of already written lines for an
5212 entire SMTP session, which doesn't seem right. The memory is now reset if a
5213 new message is started.
5215 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5216 debugging output unless log_rewrite was also set.
5218 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5219 of (int)(handle) when we know that handle contains (void *)(-1).
5221 61. The Exim daemon panic-logs an error return when it closes the incoming
5222 connection. However "connection reset by peer" seems to be common, and
5223 isn't really an error worthy of noting specially, so that particular error
5226 62. When Exim is trying to find all the local interfaces, it used to panic and
5227 die if the ioctl to get the interface flags failed. However, it seems that
5228 on at least one OS (Solaris 9) it is possible to have an interface that is
5229 included in the list of interfaces, but for which you get a failure error
5230 for this call. This happens when the interface is not "plumbed" into a
5231 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5232 failure of the "get flags" call assumes that the interface is down.
5234 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5235 makes life easier for people who are doing arithmetic on fields extracted
5236 from dates, where you often get leading zeros that should not be
5237 interpreted as octal.
5239 64. Added qualify_domain to the redirect router, to override the global
5242 65. If a pathologically long header line contained very many addresses (the
5243 report of this problem mentioned 10 000) and each of them was rewritten,
5244 Exim could use up a very large amount of memory. (It kept on making new
5245 copies of the header line as it rewrote, and never released the old ones.)
5246 At the expense of a bit more processing, the header rewriting function has
5247 been changed so that it no longer eats memory in this way.
5249 66. The generation of the Received: header has been moved from the time that a
5250 message starts to be received, to the time that it finishes. The timestamp
5251 in the Received: header should now be very close to that of the <= log
5252 line. There are two side-effects of this change:
5254 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5255 logged header lines no longer include the local Received: line, because
5256 it has not yet been created. The same applies to a copy of the message
5257 that is returned to a non-SMTP sender when a message is rejected.
5259 (b) When a filter file is tested using -bf, no additional Received: header
5260 is added to the test message. After some thought, I decided that this
5263 This change does not affect the value of $received_for. It is still set
5264 after address rewriting, but before local_scan() is called.
5266 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5268 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5269 gave an unhelpful panic error message, and a defer error. I have managed to
5270 change this behaviour so that it now rejects any supplied certificate,
5271 which seems right, as the list of acceptable certificates is empty.
5273 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5274 gave an unhelpful defer error. I have not managed to make this reject any
5275 supplied certificates, but the error message it gives is "no certificate
5276 supplied", which is not helpful.
5278 70. exigrep's output now also includes lines that are not associated with any
5279 message, but which match the given pattern. Implemented by a patch from
5280 Martin Sluka, which also tidied up the Perl a bit.
5282 71. Recipient callout verification, like sender verification, was using <> in
5283 the MAIL FROM command. This isn't really the right thing, since the actual
5284 sender may affect whether the remote host accepts the recipient or not. I
5285 have changed it to use the actual sender in the callout; this means that
5286 the cache record is now keyed on a recipient/sender pair, not just the
5287 recipient address. There doesn't seem to be a real danger of callout loops,
5288 since a callout by the remote host to check the sender would use <>.
5289 [SEE ABOVE: changed after hitting problems.]
5291 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5292 temporary errors. However, in the case of such a code being given after
5293 the end of a data transmission (i.e. after ".") Exim was failing to write
5294 a retry record for the message. (Yes, there was some broken host that was
5295 actually sending 8xx at this point.)
5297 73. An unknown lookup type in a host list could cause Exim to panic-die when
5298 the list was checked. (An example that provoked this was putting <; in the
5299 middle of a list instead of at the start.) If this happened during a DATA
5300 ACL check, a -D file could be left lying around. This kind of configuration
5301 error no longer causes Exim to die; instead it causes a defer errror. The
5302 incident is still logged to the main and panic logs.
5304 74. Buglet left over from Exim 3 conversion. The message "too many messages
5305 in one connection" was written to the rejectlog but not the mainlog, except
5306 when address rewriting (yes!) was being logged.
5308 75. Added write_rejectlog option.
5310 76. When a system filter was run not as root (that is, when system_filter_user
5311 was set), the values of the $n variables were not being returned to the
5312 main process; thus, they were not subsequently available in the $sn
5315 77. Added +return_path_on_delivery log selector.
5317 78. A connection timeout was being treated differently from recipients deferred
5318 when testing hosts_max_try with a message that was older than the host's
5319 retry timeout. (The host should not be counted, thus allowing all hosts to
5320 be tried at least once before bouncing.) This may have been the cause of an
5321 occasionally reported bug whereby a message would remain on the queue
5322 longer than the retry timeout, but would be bounced if a delivery was
5323 forced. I say "may" because I never totally pinned down the problem;
5324 setting up timeout/retry tests is difficult. See also the next item.
5326 79. The ultimate address timeout was not being applied to errors that involved
5327 a combination of host plus message (for example, a timeout on a MAIL
5328 command). When an address resolved to a number of possible hosts, and they
5329 were not all tried for each delivery (e.g. because of hosts_max_try), a
5330 message could remain on the queue longer than the retry timeout.
5332 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5335 81. Fixed an obscure SMTP outgoing bug which required at least the following
5336 conditions: (a) there was another message waiting for the same server;
5337 (b) the server returned 5xx to all RCPT commands in the first message so
5338 that the message was not completed; (c) the server dropped the connection
5339 or gave a negative response to the RSET that Exim sends to abort the
5340 transaction. The observed case was a dropped connection after DATA that had
5341 been sent in pipelining mode. That is, the server had advertised PIPELINING
5342 but was not implementing it correctly. The effect of the bug was incorrect
5343 behaviour, such as trying another host, and this could lead to a crash.
5349 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5350 and daemon.c were passed as pointers to ints; they should have been
5351 pointers to socklen_t variables (which are typically unsigned ints).
5353 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5356 3. Fixed a really odd bug that affected only the testing scheme; patching a
5357 certain fixed string in the binary changed the value of another string that
5358 happened to be identical to the end of the original first string.
5360 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5361 name", it returns that address as the IP address. On some operating
5362 systems (e.g. Solaris), it also passes back the IP address string as the
5363 "host name". However, on others (e.g. Linux), it passes back an empty
5364 string. Exim wasn't checking for this, and was changing the host name to an
5365 empty string, assuming it had been canonicized.
5367 5. Although rare, it is permitted to have more than one PTR record for a given
5368 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5369 all the names associated with an address, because they do in Solaris.
5370 However, it seems that they do not in Linux for data that comes from the
5371 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5372 I found this out when I moved to a new Linux workstation and tried to run
5373 the Exim test suite.
5375 To get round this problem I have changed the code so that it now does its
5376 own call to the DNS to look up PTR records when searching for a host name.
5377 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5378 addresses that are only in /etc/hosts are still found.
5380 This behaviour is, however, controlled by an option called host_lookup_
5381 order, which defaults to "bydns:byaddr". If people want to use the other
5382 order, or indeed, just use one or the other means of lookup, they can
5383 specify it in this variable.
5385 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5386 some operating systems, this comes back from gethostbyaddr() as an empty
5387 string, and this is what Exim used to test for. However, it seems that in
5388 other systems, "." is yielded. Exim now tests for this case too.
5390 7. The values of check_spool_space and check_log_space are now held internally
5391 as a number of kilobytes instead of an absolute number of bytes. If a
5392 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5393 kilobyte. This means that much larger values can be stored.
5395 8. Exim monitor: an attempt to get the action menu when not actually pointing
5396 at a message produces an empty menu entitled "No message selected". This
5397 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5398 no entries in it ("Shell widget menu has zero width and/or height"). So I
5399 have added a single, blank menu entry in this case.
5401 9. Added ${quote_local_part.
5403 10. MIME decoding is now applied to the contents of Subject: header lines when
5406 11. Now that a reference to $sender_host_address automatically causes a reverse
5407 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5408 host lookup before query-style lookups in lists that might use this
5409 variable. This has therefore been abolished, and the "net-" prefix is no
5410 longer necessary for query-style lookups.
5412 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5413 have been a typo for LFLAGS, so it has been changed.
5415 13. The install script calls Exim with "-C /dev/null" in order to find the
5416 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5417 to be output. Howeve, since Exim outputs its version number before the
5418 error, it didn't break the script. It just looked ugly. I fixed this by
5419 always allowing "-C /dev/null" if the caller is root.
5421 14. Ignore overlarge ACL variable number when reading spool file - insurance
5422 against a later release with more variables having written the file.
5424 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5425 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5426 this, as well as the form without the "IPv6" on the front.
5428 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5429 OS/Makefile-Darwin file.
5431 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5432 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5434 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5435 variable. However, some OS use ints for inodes. I've added cast to long int
5436 to get rid of the compiler warning.
5438 19. I had forgotten to lock out "/../" in configuration file names when
5439 ALT_CONFIG_PREFIX was set.
5441 20. Routers used for verification do not need to specify transports. However,
5442 if such a router generated a host list, and callout was configured, Exim
5443 crashed, because it could not find a port number from the (non-existent)
5444 transport. It now assumes port 25 in this circumstance.
5446 21. Added the -t option to exigrep.
5448 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5449 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5450 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5451 them is set, LOOKUP_LSEARCH is forced.
5453 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5455 24. Added sender and host information to the "rejected by local_scan()" log
5456 line; previously there was no indication of these.
5458 25. Added .include_if_exists.
5460 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5461 Linux. It turns out that not all file systems support this. Apparently some
5462 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5463 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5464 not supported on the file descriptor, is now ignored when Exim is trying to
5465 sync a directory. This applies only to Linux.
5467 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5469 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5470 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5473 29. Added support for maildirsize files from supplied patch (modified a bit).
5475 30. The use of :fail: followed by an empty string could lead Exim to respond to
5476 sender verification failures with (e.g.):
5478 550 Verification failed for <xxx>
5479 550 Sender verify failed
5481 where the first response line was missing the '-' that indicates it is not
5482 the final line of the response.
5484 31. The loop for finding the name of the user that called Exim had a hardwired
5485 limit of 10; it now uses the value of finduser_retries, which is used for
5486 all other user lookups.
5488 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5491 33. Exim was neglecting to zero errno before one call of strtol() when
5492 expanding a string and expecting an integer value. On some systems this
5493 resulted in spurious "integer overflow" errors. Also, it was casting the
5494 result into an int without checking.
5496 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5497 did not work. The code looks as if it has *never* worked, though it appears
5498 to have been documented since at least releast 1.62. I have made it work.
5500 35. The "timeout_DNS" error in retry rules, also documented since at least
5501 1.62, also never worked. As it isn't clear exactly what this means, and
5502 clearly it isn't a major issue, I have abolished the feature by treating it
5503 as "timeout", and writing a warning to the main and panic logs.
5505 36. The display of retry rules for -brt wasn't always showing the error code
5508 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5509 timeout_connect_A, timeout_connect_MX.
5511 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5512 to the empty sender.
5514 39. The daemon was not analysing the content of -oX till after it had closed
5515 stderr and disconnected from the controlling terminal. This meant that any
5516 syntax errors were only noted on the panic log, and the return code from
5517 the command was 0. By re-arranging the code a little, I've made the
5518 decoding happen first, so such errors now appear on stderr, and the return
5519 code is 1. However, the actual setting up of the sockets still happens in
5520 the disconnected process, so errors there are still only recorded on the
5523 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5524 connections (as happens on some IP stacks) was logged at start up time as
5525 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5526 it from "IPv6 and IPv4", which means that two separate sockets are being
5529 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5530 says whether AF_INET or AF_INET6 was passed as an argument.
5532 42. Exiwhat output was messed up when time zones were included in log
5535 43. Exiwhat now gives more information about the daemon's listening ports,
5536 and whether -tls-on-connect was used.
5538 44. The "port" option of the smtp transport is now expanded.
5540 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5541 silently ignored. Now an error message is written to the main and panic
5544 46. There's a new ACL modifier called "logwrite" which writes to a log file
5545 as soon as it is encountered.
5547 47. Added $local_user_uid and $local_user_gid at routing time.
5549 48. Exim crashed when trying to verify a sender address that was being
5552 49. Exim was recognizing only a space character after ".include". It now also
5553 recognizes a tab character.
5555 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5556 extracting the relevant information from the specification. The man page no
5557 longer contains scrambled data for the -d option, and I've added a section
5558 at the front about calling Exim under different names.
5560 51. Added "extra_headers" argument to the "mail" command in filter files.
5562 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5565 53. Installed eximstats 1.29.
5567 54. Added transport_filter_timeout as a generic transport option.
5569 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5570 Cc: header lines. This was required by RFC 822, but it not required by RFC
5573 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5574 incoming messages that did not have them. Now it does so only if the
5575 message originates locally, that is, if there is no associated remote host
5576 address. When Resent- header lines are present, this applies to the Resent-
5577 lines rather than the non-Resent- lines.
5579 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5580 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5582 58. Messages for configuration errors now include the name of the main
5583 configuration file - useful now that there may be more than one file in a
5584 list (.included file names were always shown).
5586 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5587 for those rare installations that do not start the daemon as root or run it
5588 setuid root. I've cut out the call to initgroups() if the daemon is not
5591 60. The Exim user and group can now be bound into the binary as text strings
5592 that are looked up at the start of Exim's processing.
5594 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5596 62. Added $mailstore_basename variable.
5598 63. Installed patch to sieve.c from Michael Haardt.
5600 64. When Exim failed to open the panic log after failing to open the main log,
5601 the original message it was trying to log was written to stderr and debug
5602 output, but if they were not available (the usual case in production), it
5603 was lost. Now it is written to syslog before the two lines that record the
5604 failures to open the logs.
5606 65. Users' Exim filters run in subprocesses under the user's uid. It is
5607 possible for a "deliver" command or an alias in a "personal" command to
5608 provoke an address rewrite. If logging of address rewriting is configured,
5609 this fails because the process is not running as root or exim. There may be
5610 a better way of dealing with this, but for the moment (because 4.30 needs
5611 to be released), I have disabled address rewrite logging when running a
5612 filter in a non-root, non-exim process.
5618 1. The buildconfig auxiliary program wasn't quoting the value set for
5619 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5620 not defined. This bug was masked in 4.22 by the effect that was fixed in
5623 2. Some messages that were rejected after a message id was allocated were
5624 shown as "incomplete" by exigrep. It no longer does this for messages that
5625 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5627 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5628 have allow_domain_literals set, the ID did not get logged in the <= line.
5629 Domain literals are now always recognized in Message-ID: header lines.
5631 4. The first argument for a ${extract expansion item is the key name or field
5632 number. Leading and trailing spaces in this item were not being ignored,
5633 causing some misleading effects.
5635 5. When deliver_drop_privilege was set, single queue runner processes started
5636 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5637 same command in the process it spins off) were not dropping privilege.
5639 6. When the daemon running as "exim" started a queue runner, it always
5640 re-executed Exim in the spun-off process. This is a waste of effort when
5641 deliver_drop_privilege is set. The new process now just calls the
5642 queue-runner function directly.
5648 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5651 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5652 ignored. Though the use of -oP was forcing the writing of a pid file, it
5653 was always written to the default place.
5655 3. If the message "no IP address found for host xxxx" is generated during
5656 incoming verification, it is now followed by identification of the incoming
5657 connection (so you can more easily find what provoked it).
5659 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5661 5. Added some features to "harden" Exim a bit more against certain attacks:
5663 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5664 be put in Local/Makefile. This is like the never_users runtime option,
5665 but it cannot be overridden. The default setting is "root".
5667 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5668 prefix string with which any file named in a -C command line option
5671 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5672 is retained for -C and -D only if the caller of Exim is root. Without
5673 it, the exim user may also use -C and -D and retain privilege.
5675 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5676 command line option is disabled.
5678 6. Macro names set by the -D option must start with an upper case letter, just
5679 like macro names defined in the configuration file.
5681 7. Added "dereference=" facility to LDAP.
5683 8. Two instances of the typo "uknown" in the source files are fixed.
5685 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5686 the Configure-Makefile script screwed up while processing it.
5688 10. Incorporated PCRE 4.4.
5690 11. The SMTP synchronization check was not operating right at the start of an
5691 SMTP session. For example, it could not catch a HELO sent before the client
5692 waited for the greeting. There is now a check for outstanding input at the
5693 point when the greeting is written. Because of the duplex, asynchronous
5694 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5695 way, but not yet received, when the check is performed.
5697 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5698 on TCP/IP sockets, because this apparently causes some broken clients to
5701 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5702 unchanged) from the Cygwin maintainer.
5704 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5705 was supported instead of testing for mbx. Effectively a typo.
5707 15. The spa authenticator server code was not checking that the input it
5708 received was valid base64.
5710 16. The debug output line for the "set" modifier in ACLs was not showing the
5711 name of the variable that was being set.
5713 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5715 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5716 lookup on its own. Something else was needed to trigger the lookup. For
5717 example, a match in host_lookup or the need for a host name in a host list.
5718 Now, if $sender_host_name is referenced and the host name has not yet been
5719 looked up, a lookup is performed. If the lookup fails, the variable remains
5720 empty, and $host_lookup_failed is set to "1".
5722 19. Added "eqi" as a case-independent comparison operator.
5724 20. The saslauthd authentication condition could segfault if neither service
5725 nor realm was specified.
5727 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5728 error message that was logged was misleading, and incoming SMTP
5729 connections were dropped. The message is now more accurate, and temporary
5730 errors are given to SMTP connections.
5732 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5733 (except RSET) with a 421 error, until QUIT is received. However, it was
5734 failing to send a response to QUIT.
5736 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5737 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5738 ACL is now run after the helo_try_verify_hosts code.
5740 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5741 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5742 case-independent, so other case variants are also recognized). Apparently
5743 some systems use these upper case variants.
5745 25. If more than two messages were waiting for the same host, and a transport
5746 filter was specified for the transport, Exim sent two messages over the
5747 same TCP/IP connection, and then failed with "socket operation on non-
5748 socket" when it tried to send the third.
5750 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5752 27. The extern definition of crypt16() in expand.c was not being excluded when
5753 the OS had its own crypt16() function.
5755 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5756 as a preferred synonym for return_size_limit, both as an option and as an
5759 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5761 30. Changed the default configuration ACL to relax the local part checking rule
5762 for addresses that are not in any local domains. For these addresses,
5763 slashes and pipe symbols are allowed within local parts, but the sequence
5764 /../ is explicitly forbidden.
5766 31. SPA server authentication was not clearing the challenge buffer before
5769 32. log_message in a "warn" ACL statement was writing to the reject log as
5770 well as to the main log, which contradicts the documentation and doesn't
5771 seem right (because no rejection is happening). So I have stopped it.
5773 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5774 However, I am unable to do any testing of this.
5776 34. Fixed an infelicity in the appendfile transport. When checking directories
5777 for a mailbox, to see if any needed to be created, it was accidentally
5778 using path names with one or more superfluous leading slashes; tracing
5779 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5781 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5782 contents are added to the log line that is written for every discarded
5783 recipient. (Previously a log_message setting was ignored.)
5785 36. The ${quote: operator now quotes the string if it is empty.
5787 37. The install script runs exim in order to find its version number. If for
5788 some reason other than non-existence or emptiness, which it checks, it
5789 could not run './exim', it was installing it with an empty version number,
5790 i.e. as "exim-". This error state is now caught, and the installation is
5793 38. An argument was missing from the function that creates an error message
5794 when Exim fails to connect to the socket for saslauthd authentication.
5795 This could cause Exim to crash, or give a corrupted message.
5797 39. Added isip, isip4, and isip6 to ${if conditions.
5799 40. The ACL variables $acl_xx are now saved with the message, and can be
5800 accessed later in routers, transports, and filters.
5802 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5803 strings in the file are not string-expanded.
5805 42. If a MAIL command specified a SIZE value that was too large to fit into an
5806 int variable, the check against message_size_limit failed. Such values are
5807 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5808 day this will have to be increased, but I don't think I want to be around
5809 when emails are that large.
5816 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5817 iconv() is not standard in FreeBSD.
5819 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5820 IPv6 enabled. The observed symptom was a segmentation fault on return from
5821 the function os_common_find_running_interfaces() in src/os.c.
5823 3. In the check_special_case() function in daemon.c I had used "errno" as an
5824 argument name, which causes warnings on some systems. This was basically a
5825 typo, since it was named "eno" in the comments!
5827 4. The code that waits for the clock to tick (at a resolution of some fraction
5828 of a second) so as to ensure message-id uniqueness was always waiting for
5829 at least one whole tick, when it could have waited for less. [This is
5830 almost certainly not relevant at current processor speeds, where it is
5831 unlikely to ever wait at all. But we try to future-proof.]
5833 5. The function that sleeps for a time interval that includes fractions of a
5834 second contained a race. It did not block SIGALRM between setting the
5835 timer, and suspending (a couple of lines later). If the interval was short
5836 and the sigsuspend() was delayed until after it had expired, the suspension
5837 never ended. On busy systems this could lead to processes getting stuck for
5840 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5841 process, before it forks any delivery processes. The open lookup caching
5842 mechanism meant that the open file or database connection was passed into
5843 the delivery process. The problem was that delivery processes always tidy
5844 up cached lookup data. This could cause a problem for the next delivery
5845 process started by the queue runner, because the external queue runner
5846 process does not know about the closure. So the next delivery process
5847 still has data in the lookup cache. In the case of a file lookup, there was
5848 no problem because closing a file descriptor in a subprocess doesn't affect
5849 the parent. However, if the lookup was caching a connection to a database,
5850 the connection was closed, and the second delivery process was likely to
5851 see errors such as "PGSQL: query failed: server closed the connection
5852 unexpectedly". The problem has been fixed by closing all cached lookups
5853 in a queue runner before running a delivery process.
5855 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5856 seem to have the "const" qualifier which it has on other OS. I've
5859 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5860 *of the same type* (client or server) with the same public name that an
5861 error should be diagnosed.
5863 9. When Exim looked up a host name for an IP address, but failed to find the
5864 original IP address when looking up the host name (a safety check), it
5865 output the message "<ip address> does not match any IP for NULL", which was
5866 confusing, to say the least. The bug was that the host name should have
5867 appeared instead of "NULL".
5869 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5870 user that is built into the binary, and the -C or -D options is used, root
5871 privilege is dropped before the configuration file is read. In addition,
5872 logging is switched to stderr instead of the normal log files. If the
5873 configuration then re-defines the Exim user, the unprivileged environment
5874 is probably not what is expected, so Exim logs a panic warning message (but
5877 However, if deliver_drop_privilege is set, the unprivileged state may well
5878 be exactly what is intended, so the warning has been cut out in that case,
5879 and Exim is allowed to try to write to its normal log files.
5885 1. smtp_return_error_details was not giving details for temporary sender
5886 or receiver verification errors.
5888 2. Diagnose a configuration error if two authenticators have the same public
5891 3. Exim used not to create the message log file for a message until the first
5892 delivery attempt. This could be confusing when incoming messages were held
5893 for policy or load reasons. The message log file is now created at the time
5894 the message is received, and an initial "Received" line is written to it.
5896 4. The automatically generated man page for command line options had a minor
5897 bug that caused no ill effects; however, a more serious problem was that
5898 the procedure for building the man page automatically didn't always
5899 operate. Consequently, release 4.20 contains an out-of-date version. This
5900 shouldn't happen again.
5902 5. When building Exim with embedded Perl support, the script that builds the
5903 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5904 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5906 6. The freeze_tell option was not being used for messages that were frozen on
5907 arrival, either by an ACL or by local_scan().
5909 7. Added the smtp_incomplete_transaction log selector.
5911 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5912 was accepting AUTH without a new EHLO.
5914 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5915 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5918 10. Logging of TCP/IP connections (when configured) now happens in the main
5919 daemon process instead of the child process, so that the TCP/IP connection
5920 count is more accurate (but it can never be perfect).
5922 11. The use of "drop" in a nested ACL was not being handled correctly in the
5923 outer ACL. Now, if condition failure induced by the nested "drop" causes
5924 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5925 or "require"), the connection is dropped.
5927 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5928 that yield "discard" can now be used with an "accept" or a "discard" verb,
5929 but an error is generated for any others (because I can't see a useful way
5930 to define what should happen).
5932 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5933 are now processed in the same way as lines in the Exim configuration file.
5934 In particular, continuation lines are supported.
5936 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5938 15. Added -ti meaning -t -i.
5940 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5941 domains, and warn by logging if others are found.
5943 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5944 returned from GIFCONF when Exim is trying to find the list of interfaces on
5945 a host. The code in os.c has been modified to copy each ifreq to an aligned
5946 structure in all cases.
5948 Also, in some cases, the returned ifreq's were being copied to a 'struct
5949 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5950 means the last couple of bytes of an IPv6 address could be chopped if the
5951 ifreq contained only a normal sockaddr (14 bytes storage).
5953 18. Named domain lists were not supported in the hosts_treat_as_local option.
5954 An entry such as +xxxx was not recognized, and was treated as a literal
5957 19. Ensure that header lines added by a DATA ACL are included in the reject log
5958 if the ACL subsequently rejects the message.
5960 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5961 MD5 (which is deprecated).
5963 21. When testing a filter file using -bf, Exim was writing a message when it
5964 took the sender from a "From " line in the message, but it was not doing so
5965 when it took $return_path from a Return-Path: header line. It now does.
5967 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5968 with a valid header line field name (a series of printing characters
5969 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5971 23. Changed "disc" in the source to "disk" to conform to the documentation and
5972 the book and for uniformity.
5974 24. Ignore Sendmail's -Ooption=value command line item.
5976 25. When execve() failed while trying to run a command in a pipe transport,
5977 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5978 could be confused with a return value of 69 from the command itself. This
5979 has been changed to 127, the value the shell returns if it is asked to run
5980 a non-existent command. The wording for the related log line suggests a
5981 non-existent command as the problem.
5983 26. If received_header_text expands to an empty string, do not add a Received:
5984 header line to the message. (Well, it adds a token one on the spool, but
5985 marks it "old" so that it doesn't get used or transmitted.)
5987 27. Installed eximstats 1.28 (addition of -nt option).
5989 28. There was no check for failure on the call to getsockname() in the daemon
5990 code. This can fail if there is a shortage of resources on the system, with
5991 ENOMEM, for example. A temporary error is now given on failure.
5993 29. Contrary to the C standard, it seems that in some environments, the
5994 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5995 program. Exim now does this explicitly; it affects the formatting of
5996 timestamps using strftime().
5998 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5999 complaints. I've now initialized all the variables, to avoid this.
6001 32. Header lines added by a system filter were not being "seen" during
6002 transport-time rewrites.
6004 33. The info_callback() function passed to OpenSSL is set up with type void
6005 (*)(SSL *, int, int), as described somewhere. However, when calling the
6006 function (actually a macro) that sets it up, the type void(*)() is
6007 expected. I've put in a cast to prevent warnings from picky compilers.
6009 34. If a DNS black list lookup found a CNAME record, but there were no A
6010 records associated with the domain it pointed at, Exim crashed.
6012 35. If a DNS black list lookup returned more than one A record, Exim ignored
6013 all but the first. It now scans all returned addresses if a particular IP
6014 value is being sought. In this situation, the contents of the
6015 $dnslist_value variable are a list of all the addresses, separated by a
6018 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6019 to accept a host name and all its aliases if the forward lookup for any of
6020 them yielded the IP address of the incoming connection. Now it accepts only
6021 those names whose forward lookup yields the correct IP address. Any other
6022 names are discarded. This closes a loophole whereby a rogue DNS
6023 administrator could create reverse DNS records to break through a
6024 wildcarded host restriction in an ACL.
6026 37. If a user filter or a system filter that ran in a subprocess used any of
6027 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6028 the wrong values were passed to the pipe command ($thisaddress had the
6029 value of $0, $0 had the value of $1, etc). This bug was introduced by
6030 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6032 38. Improved the line breaking for long SMTP error messages from ACLs.
6033 Previously, if there was no break point between 40 and 75 characters, Exim
6034 left the rest of the message alone. Two changes have been made: (a) I've
6035 reduced the minimum length to 35 characters; (b) if it can't find a break
6036 point between 35 and 75 characters, it looks ahead and uses the first one
6037 that it finds. This may give the occasional overlong line, but at least the
6038 remaining text gets split now.
6040 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6041 file descriptors might be low, and that setting 1000 would always raise it.
6042 It turns out that in some environments, the limit is already over 1000 and
6043 that lowering it causes trouble. So now Exim takes care not to decrease it.
6045 40. When delivering a message, the value of $return_path is set to $sender_
6046 address at the start of routing (routers may change the value). By an
6047 oversight, this default was not being set up when an address was tested by
6048 -bt or -bv, which affected the outcome if any router or filter referred to
6051 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6052 log only when "message" or "log_message" are set. However, if one of the
6053 conditions was an address verification, or a call to a nested ACL, the
6054 messages generated by the underlying test were being passed through. This
6055 no longer happens. The underlying message is available in $acl_verify_
6056 message for both "message" and "log_message" expansions, so it can be
6057 passed through if needed.
6059 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6060 new expansion $bh_ to give the encoded byte string without charset
6061 translation. Translation happens only if iconv() is available; HAVE_ICONV
6062 indicates this at build time. HEADERS_CHARSET gives the charset to
6063 translate to; headers_charset can change it in the configuration, and
6064 "headers charset" can change it in an individual filter file.
6066 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6067 that creates RFC 2047 encoded "words" labels them as that charset instead
6068 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6069 expansion operator; (ii) when Exim creates a From: line for a local
6070 message; (iii) when a header line is rewritten to include a "phrase" part.
6072 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6073 buggy, causing it to skip the first lines of messages whose message ID
6074 ended in 'D'. This would not have bitten before Exim release 4.14, because
6075 message IDs were unlikely to end in 'D' before then. The effect was to have
6076 incorrect size information for certain domains.
6078 45. #include "config.h" was missing at the start of the crypt16.c module. This
6079 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6082 46. If there was a timeout during a "random" callout check, Exim treated it as
6083 a failure of the random address, and carried on sending RSET and the real
6084 address. If the delay was just some slowness somewhere, the response to the
6085 original RCPT would be taken as a response to RSET and so on, causing
6086 mayhem of various kinds.
6088 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6089 when I implemented it. It didn't allow for the fact that some option values
6090 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6091 the right test for positive values.
6093 48. Domain names in DNS records are case-independent. Exim always looks them up
6094 in lower case. Some resolvers return domain names in exactly the case they
6095 appear in the zone file, that is, they may contain uppercase letters. Not
6096 all resolvers do this - some return always lower case. Exim was treating a
6097 change of case by a resolver as a change of domain, similar to a widening
6098 of a domain abbreviation. This triggered its re-routing code and so it was
6099 trying to route what was effectively the same domain again. This normally
6100 caused routing to fail (because the router wouldn't handle the domain
6101 twice). Now Exim checks for this case specially, and just changes the
6102 casing of the domain that it ultimately uses when it transmits the message
6105 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6108 50. If a filter generated a file delivery with a non-absolute name (possible if
6109 no home directory exists for the router), the forbid_file option was not
6112 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6113 the existing equality matching.
6115 52. Exim was using ints instead of ino_t variables in some places where it was
6116 dealing with inode numbers.
6118 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6119 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6120 called TMPDIR, and if it finds it is different, it changes its value.
6122 54. The smtp_printf() function is now made available to local_scan() so
6123 additional output lines can be written before returning. There is also an
6124 smtp_fflush() function to enable the detection of a dropped connection.
6125 The variables smtp_input and smtp_batched_input are exported to
6128 55. Changed the default runtime configuration: the message "Unknown user"
6129 has been removed from the ACL, and instead placed on the localuser router,
6130 using the cannot_route_message feature. This means that any verification
6131 failures that generate their own messages won't get overridden. Similarly,
6132 the "Unrouteable address" message that was in the ACL for unverifiable
6133 relay addresses has also been removed.
6135 56. Added hosts_avoid_esmtp to the smtp transport.
6137 57. The exicyclog script was not checking for the esoteric option
6138 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6139 will work only if exicyclog is run under the appropriate euid.
6141 58. Following a discussion on the list, the rules by which Exim recognises line
6142 endings on incoming messages have been changed. The -dropcr and drop_cr
6143 options are now no-ops, retained only for backwards compatibility. The
6144 following line terminators are recognized: LF CRLF CR. However, special
6145 processing applies to CR:
6147 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6148 nor a local message in the state where . is a terminator.
6150 (ii) If a bare CR is encountered in a header line, an extra space is added
6151 after the line terminator so as not to end the header. The reasoning
6152 behind this is that bare CRs in header lines are most likely either
6153 to be mistakes, or people trying to play silly games.
6155 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6156 was being incorrectly given as 18 bytes larger than it should have been.
6157 This is a VOB (very old bug).
6159 60. This may never have affected anything current, but just in case it has:
6160 When the local host is found other than at the start of a list of hosts,
6161 the local host, those with the same MX, and any that follow, are discarded.
6162 When the list in question was part of a longer list of hosts, the following
6163 hosts (not currently being processed) were also being discarded. This no
6164 longer happens. I'm not sure if this situation could ever has previously
6167 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6170 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6171 header. This is something that is recommended in a new Internet Draft, and
6172 is something that is documented as being done by Sendmail. There are two
6173 possible values. For messages generated by the autoreply transport, Exim
6176 Auto-Submitted: auto-replied
6178 whereas for all other generated messages (e.g. bounces) it adds
6180 Auto-Submitted: auto-generated
6182 63. The "personal" condition in filters now includes a test for the
6183 Auto-Submitted: header. If it contains the string "auto-" the message it
6184 not considered personal.
6186 64. Added rcpt_include_affixes as a generic transport option.
6188 65. Added queue_only_override (default true).
6190 66. Added the syslog_duplication option.
6192 67. If what should have been the first header line of a message consisted of
6193 a space followed by a colon, Exim was mis-interpreting it as a header line.
6194 It isn't of course - it is syntactically invalid and should therefore be
6195 treated as the start of the message body. The misbehaviour could have
6196 caused a number of strange effects, including loss of data in subsequent
6197 header lines, and spool format errors.
6199 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6200 client host had authenticated. This control can now be exercised by an ACL
6201 for more flexibility.
6203 69. By default, callouts do not happen when testing with -bh. There is now a
6204 variant, -bhc, which does actually run the callout code, including
6205 consulting and updating the callout cache.
6207 70. Added support for saslauthd authentication, courtesy of Alexander
6210 71. If statvfs() failed on the spool or log directories while checking their
6211 size for availability, Exim confusingly gave the error "space shortage".
6212 Furthermore, in debugging mode it crashed with a floating point exception.
6213 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6214 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6215 serious problem, Exim now writes to the main and panic logs when this
6216 happens, with details of the failure. It then refuses to accept the
6217 incoming message, giving the message "spool directory problem" or "log
6218 directory problem" with a 421 code for SMTP messages.
6220 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6221 0, 1, and 2 exist, because some OS complain for execs without them (see
6222 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6223 descriptors. However, the code omitted to check that the open succeeded,
6224 causing mysterious errors if for some reason the permissions on /dev/null
6225 got screwed. Now Exim writes a message to the main and panic logs, and
6226 bombs out if it can't open /dev/null.
6228 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6229 interact so that it is all more flexible. It is supposed to remain
6230 backwards compatible. Also added extra_local_interfaces.
6232 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6233 to bomb out with an assertion failure - to the client this appears as a
6234 connection drop. This problem occurs in the part of the code that was taken
6235 from the Samba project. Fortunately, the assertion is in a very simple
6236 function, so I have fixed this by reproducing the function inline in the
6237 one place where it is called, and arranging for authentication to fail
6238 instead of killing the process with assert().
6240 75. The SPA client code was not working when the server requested OEM rather
6241 than Unicode encoding.
6243 76. Added code to make require_files with a specific uid setting more usable in
6244 the case where statting the file as root fails - usually a non-root-mounted
6245 NFS file system. When this happens and the failure is EACCES, Exim now
6246 forks a subprocess and does the per-uid checking as the relevant uid.
6248 77. Added process_log_path.
6250 78. If log_file_path was not explicitly set, a setting of check_log_space or
6251 check_log_inodes was ignored.
6253 79. If a space check for the spool or log partitions fails, the incident is now
6254 logged. Of course, in the latter case the data may get lost...
6256 80. Added the %p formatting code to string_format() so that it can be used to
6257 print addresses in debug_print(). Adjusted all the address printing in the
6258 debugging in store.c to use %p rather than %d.
6260 81. There was a concern that a line of code in smtp_in.c could overflow a
6261 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6262 initially expressed, the concern was not well-founded, because trailing
6263 spaces are removed early. However, if the trailing spaces were followed by
6264 a NULL, they did not get removed, so the overflow was possible. Two fixes
6267 (a) I re-wrote the offending code in a cleaner fashion.
6268 (b) If an incoming SMTP command contains a NULL character, it is rejected
6271 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6272 runs initgroups(), so that if the Exim user is in any additional groups,
6273 they will be used during message reception.
6279 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob