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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
98 TL/01 When running the test suite, the README says that variables such as
99 no_msglog_check are global and can be placed anywhere in a specific
100 test's script, however it was observed that placement needed to be near
101 the beginning for it to behave that way. Changed the runtest perl
102 script to read through the entire script once to detect and set these
103 variables, reset to the beginning of the script, and then run through
104 the script parsing/test process like normal.
106 TL/02 The BSD's have an arc4random API. One of the functions to induce
107 adding randomness was arc4random_stir(), but it has been removed in
108 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
109 function when detected.
111 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
112 cause callback expansion.
114 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
115 syntax errors in an expansion can be treated as a string instead of
116 logging or causing an error, due to the internal use of bool_lax
117 instead of bool when processing it.
119 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
120 server certificates when making smtp deliveries.
122 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
124 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
126 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
128 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
129 Merged patch from Sebastian Wiedenroth.
131 JH/05 Fix results-pipe from transport process. Several recipients, combined
132 with certificate use, exposed issues where response data items split
133 over buffer boundaries were not parsed properly. This eventually
134 resulted in duplicates being sent. This issue only became common enough
135 to notice due to the introduction of conection certificate information,
136 the item size being so much larger. Found and fixed by Wolfgang Breyha.
138 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
139 size buffer was used, resulting in syntax errors when an expansion
142 JH/07 Add support for directories of certificates when compiled with a GnuTLS
143 version 3.3.6 or later.
145 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
146 is EXPERIMENTAL_EVENT, the main-configuration and transport options
147 both become "event_action", the variables become $event_name, $event_data
148 and $event_defer_errno. There is a new variable $verify_mode, usable in
149 routers, transports and related events. The tls:cert event is now also
150 raised for inbound connections, if the main configuration event_action
153 TL/06 In test suite, disable OCSP for old versions of openssl which contained
154 early OCSP support, but no stapling (appears to be less than 1.0.0).
156 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
157 server certificate names available under the smtp transport option
158 "tls_verify_cert_hostname" now do not permit multi-component wildcard
161 JH/10 Time-related extraction expansions from certificates now use the main
162 option "timezone" setting for output formatting, and are consistent
163 between OpenSSL and GnuTLS compilations. Bug 1541.
165 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
166 encoded parameter in the incoming message. Bug 1558.
168 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
169 include certificate info, eximon was claiming there were spoolfile
172 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
174 JH/14 Log delivery-related information more consistently, using the sequence
175 "H=<name> [<ip>]" wherever possible.
177 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
178 are problematic for Debian distribution, omit them from the release
181 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
183 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
186 JH/17 Fix a null-indirection in certextract expansions when a nondefault
187 output list separator was used.
192 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
193 checkers that were complaining about end of non-void function with no
196 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
197 This was a regression intruduced in 4.83 by another bugfix.
199 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
201 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
202 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
208 TF/01 Correctly close the server side of TLS when forking for delivery.
210 When a message was received over SMTP with TLS, Exim failed to clear up
211 the incoming connection properly after forking off the child process to
212 deliver the message. In some situations the subsequent outgoing
213 delivery connection happened to have the same fd number as the incoming
214 connection previously had. Exim would try to use TLS and fail, logging
215 a "Bad file descriptor" error.
217 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
218 utilities have not been installed.
220 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
221 temporary space as the ACL may create new global variables.
223 TL/01 LDAP support uses per connection or global context settings, depending
224 upon the detected version of the libraries at build time.
226 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
227 to extract and use the src ip:port in logging and expansions as if it
228 were a direct connection from the outside internet. PPv2 support was
229 updated based on HAProxy spec change in May 2014.
231 JH/02 Add ${listextract {number}{list}{success}{fail}}.
233 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
234 Properly escape header and check for NULL return.
236 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
239 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
241 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
242 characters in header names, implemented as a verify condition.
243 Contributed by Michael Fischer v. Mollard.
245 TL/05 Rename SPF condition results err_perm and err_temp to standardized
246 results permerror and temperror. Previous values are deprecated but
247 still accepted. In a future release, err_perm and err_temp will be
248 completely removed, which will be a backward incompatibility if the
249 ACL tests for either of these two old results. Patch contributed by
250 user bes-internal on the mailing list.
252 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
254 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
255 selectors, in both main and reject logs.
257 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
260 JH/07 Add malware type "sock" for talking to simple daemon.
262 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
264 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
265 routers/transports under cutthrough routing.
267 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
268 numbers. Touch up "bool" conditional to keep the same definition.
270 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
272 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
274 JH/12 Expand items in router/transport headers_add or headers_remove lists
275 individually rather than the list as a whole. Bug 1452.
277 Required for reasonable handling of multiple headers_ options when
278 they may be empty; requires that headers_remove items with embedded
279 colons must have them doubled (or the list-separator changed).
281 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
282 view the policy declared in the DMARC record. Currently, $dmarc_status
283 is a combined value of both the record presence and the result of the
286 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
288 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
289 dnslookup router and the smtp transport (applying to the forward
292 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
293 of ldap servers used for a specific lookup. Patch provided by Heiko
296 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
297 New variable $lookup_dnssec_authenticated for observability.
299 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
300 Patch submitted by Lars Timman.
302 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
304 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
305 Requires trusted mode and valid format message id, aborts otherwise.
306 Patch contributed by Heiko Schlichting.
308 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
309 certextract with support for various fields. Bug 1358.
311 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
312 is requested by default, modifiable by smtp transport option
315 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
316 operate on certificate variables to give certificate fingerprints
317 Also new ${sha256:cert_variable}.
319 JH/23 The PRDR feature is moved from being Experimental into the mainline.
321 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
322 Christian Aistleitner.
324 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
326 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
327 file. Patch from Wolfgang Breyha.
329 JH/25 Expand the coverage of the delivery $host and $host_address to
330 client authenticators run in verify callout. Bug 1476.
332 JH/26 Port service names are now accepted for tls_on_connect_ports, to
333 align with daemon_smtp_ports. Bug 72.
335 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
336 support and error reporting did not work properly.
338 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
339 and is readable. Patch from Andrew Colin Kissa.
341 TL/14 Enhance documentation of ${run expansion and how it parses the
342 commandline after expansion, particularly in the case when an
343 unquoted variable expansion results in an empty value.
345 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
347 PP/02 Fix internal collision of T_APL on systems which support RFC3123
348 by renaming away from it. Addresses GH issue 15, reported by
351 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
353 TL/15 SECURITY: prevent double expansion in math comparison functions
354 (can expand unsanitized data). Not remotely exploitable.
361 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
363 PP/02 Make -n do something, by making it not do something.
364 When combined with -bP, the name of an option is not output.
366 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
369 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
370 $sender_host_name and config options to manage this, and basic check
373 PP/05 DSCP support for outbound connections and control modifier for inbound.
375 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
376 (Only plugin which currently uses this is kerberos4, which nobody should
377 be using, but we should make it available and other future plugins might
378 conceivably use it, even though it would break NAT; stuff *should* be
379 using channel bindings instead).
381 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
382 name; added for Sendmail compatibility; requires admin caller.
383 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
384 just ignore it); requires trusted caller.
385 Also parse but ignore: -Ac -Am -X<logfile>
388 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
390 TL/02 Add +smtp_confirmation as a default logging option.
392 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
393 Patch by Magnus Holmgren from 2007-02-20.
395 TL/04 Bugzilla 1281 - Spec typo.
396 Bugzilla 1283 - Spec typo.
397 Bugzilla 1290 - Spec grammar fixes.
399 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
401 TL/06 Add Experimental DMARC support using libopendmarc libraries.
403 TL/07 Fix an out of order global option causing a segfault. Reported to dev
404 mailing list by by Dmitry Isaikin.
406 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
408 JH/02 Support "G" suffix to numbers in ${if comparisons.
410 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
412 NM/01 Bugzilla 1197 - Spec typo
413 Bugzilla 1196 - Spec examples corrections
415 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
417 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
418 gnutls_enable_pkcs11, but renamed to more accurately indicate its
421 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
422 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
424 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
425 "acl {{name}{arg}...}", and optional args on acl condition
428 JH/05 Permit multiple router/transport headers_add/remove lines.
430 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
432 JH/07 Avoid using a waiting database for a single-message-only transport.
433 Performance patch from Paul Fisher. Bugzilla 1262.
435 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
438 JH/09 Add $headers_added variable, with content from use of ACL modifier
439 add_header (but not yet added to the message). Bugzilla 199.
441 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
442 Pulled from Bugzilla 817 by Wolfgang Breyha.
444 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
446 (nb: this is the same fix as in Exim 4.80.1)
448 JH/11 Add A= logging on delivery lines, and a client_set_id option on
451 JH/12 Add optional authenticated_sender logging to A= and a log_selector
454 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
456 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
457 advertise SMTP AUTH mechanism to us, instead of a generic
458 protocol violation error. Also, make Exim more robust to bad
459 data from the Dovecot auth socket.
461 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
463 When a queue runner is handling a message, Exim first routes the
464 recipient addresses, during which it prunes them based on the retry
465 hints database. After that it attempts to deliver the message to
466 any remaining recipients. It then updates the hints database using
469 So if a recipient address works intermittently, it can get repeatedly
470 deferred at routing time. The retry hints record remains fresh so the
471 address never reaches the final cutoff time.
473 This is a fairly common occurrence when a user is bumping up against
474 their storage quota. Exim had some logic in its local delivery code
475 to deal with this. However it did not apply to per-recipient defers
476 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
478 This change adds a proper retry rule check during routing so that the
479 final cutoff time is checked against the message's age. We only do
480 this check if there is an address retry record and there is not a
481 domain retry record; this implies that previous attempts to handle
482 the address had the retry_use_local_parts option turned on. We use
483 this as an approximation for the destination being like a local
484 delivery, as in LMTP.
486 I suspect this new check makes the old local delivery cutoff check
487 redundant, but I have not verified this so I left the code in place.
489 TF/02 Correct gecos expansion when From: is a prefix of the username.
491 Test 0254 submits a message to Exim with the header
495 When I ran the test suite under the user fanf2, Exim expanded
496 the header to contain my full name, whereas it should have added
497 a Resent-Sender: header. It erroneously treats any prefix of the
498 username as equal to the username.
500 This change corrects that bug.
502 GF/01 DCC debug and logging tidyup
503 Error conditions log to paniclog rather than rejectlog.
504 Debug lines prefixed by "DCC: " to remove any ambiguity.
506 TF/03 Avoid unnecessary rebuilds of lookup-related code.
508 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
509 Bug spotted by Jeremy Harris; was flawed since initial commit.
510 Would have resulted in OCSP responses post-SNI triggering an Exim
511 NULL dereference and crash.
513 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
515 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
516 Bug detection, analysis and fix by Samuel Thibault.
517 Bugzilla 1331, Debian bug #698092.
519 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
521 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
522 Server implementation by Todd Lyons, client by JH.
523 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
524 config variable "prdr_enable" controls whether the server
525 advertises the facility. If the client requests PRDR a new
526 acl_data_smtp_prdr ACL is called once for each recipient, after
527 the body content is received and before the acl_smtp_data ACL.
528 The client is controlled by bolth of: a hosts_try_prdr option
529 on the smtp transport, and the server advertisement.
530 Default client logging of deliveries and rejections involving
531 PRDR are flagged with the string "PRDR".
533 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
534 fclose(). Diagnosis by Todd Lyons.
536 PP/17 Update configure.default to handle IPv6 localhost better.
537 Patch by Alain Williams (plus minor tweaks).
540 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
541 This is now consistent with GnuTLS, and is now documented: the
542 previous undocumented portable approach to treating the option as
543 unset was to force an expansion failure. That still works, and
544 an empty string is now equivalent.
546 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
547 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
548 not performing validation itself.
550 PP/20 Added force_command boolean option to pipe transport.
551 Patch from Nick Koston, of cPanel Inc.
553 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
556 TF/04 Added udpsend ACL modifer and hexquote expansion operator
558 PP/21 Fix eximon continuous updating with timestamped log-files.
559 Broken in a format-string cleanup in 4.80, missed when I repaired the
560 other false fix of the same issue.
561 Report and fix from Heiko Schlichting.
564 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
565 Report from Prashanth Katuri.
567 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
568 It's SecureTransport, so affects any MacOS clients which use the
569 system-integrated TLS libraries, including email clients.
571 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
572 using a MIME ACL for non-SMTP local injection.
573 Report and assistance in diagnosis by Warren Baker.
575 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
577 JH/16 Fix comparisons for 64b. Bugzilla 1385.
579 TL/09 Add expansion variable $authenticated_fail_id to keep track of
580 last id that failed so it may be referenced in subsequent ACL's.
582 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
585 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
586 ldap library initialization, allowing self-signed CA's to be
587 used. Also properly sets require_cert option later in code by
588 using NULL (global ldap config) instead of ldap handle (per
589 session). Bug diagnosis and testing by alxgomz.
591 TL/12 Enhanced documentation in the ratelimit.pl script provided in
592 the src/util/ subdirectory.
594 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
595 renamed to Transport Post Delivery Action by Jeremy Harris, as
598 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
599 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
600 redis_servers = needs to be configured which will be used by the redis
601 lookup. Patch from Warren Baker, of The Packet Hub.
603 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
605 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
606 hostname or reverse DNS when processing a host list. Used suggestions
607 from multiple comments on this bug.
609 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
611 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
612 Missed a few lines, added it to make the runtest require no keyboard
615 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
616 contains upper case chars. Make router use caseful_local_part.
618 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
619 support when GnuTLS has been built with p11-kit.
625 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
627 This, or similar/improved, will also be change PP/11 of 4.82.
633 PP/01 Handle short writes when writing local log-files.
634 In practice, only affects FreeBSD (8 onwards).
635 Bugzilla 1053, with thanks to Dmitry Isaikin.
637 NM/01 Bugzilla 949 - Documentation tweak
639 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
642 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
644 PP/02 Implemented gsasl authenticator.
646 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
648 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
649 `pkg-config foo` for cflags/libs.
651 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
652 with rest of GSASL and with heimdal_gssapi.
654 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
655 `pkg-config foo` for cflags/libs for the TLS implementation.
657 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
658 properties get this fed in as external SSF. A number of robustness
659 and debugging improvements to the cyrus_sasl authenticator.
661 PP/08 cyrus_sasl server now expands the server_realm option.
663 PP/09 Bugzilla 1214 - Log authentication information in reject log.
664 Patch by Jeremy Harris.
666 PP/10 Added dbmjz lookup type.
668 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
670 PP/12 MAIL args handles TAB as well as SP, for better interop with
671 non-compliant senders.
672 Analysis and variant patch by Todd Lyons.
674 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
675 Bug report from Lars Müller <lars@samba.org> (via SUSE),
676 Patch from Dirk Mueller <dmueller@suse.com>
678 PP/13 tls_peerdn now print-escaped for spool files.
679 Observed some $tls_peerdn in wild which contained \n, which resulted
680 in spool file corruption.
682 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
683 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
684 or write after TLS renegotiation, which otherwise led to messages
687 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
688 as a tracking header (ie: a signed header comes before the signature).
689 Patch from Wolfgang Breyha.
691 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
692 comma-sep list; embedded commas doubled.
694 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
696 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
698 Report and patch from Dmitry Banschikov.
700 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
701 Removed SSL_clear() after SSL_new() which led to protocol negotiation
702 failures. We appear to now support TLS1.1+ with Exim.
704 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
705 lets Exim select keys and certificates based upon TLS SNI from client.
706 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
707 before an outbound SMTP session. New log_selector, +tls_sni.
709 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
710 NULL dereference. Report and patch from Alun Jones.
712 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
713 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
714 Not seeing resolver debug output on NetBSD, but suspect this is a
715 resolver implementation change.
717 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
718 Left warnings. Added "eximon gdb" invocation mode.
720 PP/21 Defaulting "accept_8bitmime" to true, not false.
722 PP/22 Added -bw for inetd wait mode support.
724 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
725 locate the relevant includes and libraries. Made this the default.
727 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
728 Bugzilla 1246, report and most of solution from Tomasz Kusy.
730 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
731 This may cause build issues on older platforms.
733 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
734 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
735 gnutls_require_mac & gnutls_require_protocols (no longer supported).
736 Added SNI support via GnuTLS too.
737 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
739 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
741 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
742 Applied second patch from Janne, implementing suggestion to default
743 multiple-strings-in-record handling to match SPF spec.
745 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
747 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
748 read-only, out of scope).
749 Patch from Wolfgang Breyha, report from Stuart Northfield.
751 PP/29 Fix three issues highlighted by clang analyser static analysis.
752 Only crash-plausible issue would require the Cambridge-specific
753 iplookup router and a misconfiguration.
754 Report from Marcin Mirosław.
756 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
758 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
759 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
760 As part of this, removing so much warning spew let me fix some minor
761 real issues in debug logging.
763 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
764 assignment on my part. Fixed.
766 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
767 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
768 Janne Snabb (who went above and beyond: thank you).
770 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
771 string otherwise requires a connection and a bunch more work and it's
772 relatively easy to get wrong. Should also expose TLS library linkage
775 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
776 64-bit ${eval} (JH/03).
778 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
779 GNU libc to support some of the 64-bit stuff, should not lead to
780 conflicts. Defined before os.h is pulled in, so if a given platform
781 needs to override this, it can.
783 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
784 protection layer was required, which is not implemented.
785 Bugzilla 1254, patch from Wolfgang Breyha.
787 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
788 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
789 tls_dhparam take prime identifiers. Also unbreak combination of
790 OpenSSL+DH_params+TLSSNI.
792 PP/39 Disable SSLv2 by default in OpenSSL support.
798 PP/01 Solaris build fix for Oracle's LDAP libraries.
799 Bugzilla 1109, patch from Stephen Usher.
801 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
803 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
806 TF/02 Fix a couple more cases where we did not log the error message
807 when unlink() failed. See also change 4.74-TF/03.
809 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
810 lock up or crash if it happened to be inside a call to libc when it
811 got a SIGUSR1 from exiwhat.
813 The SIGUSR1 handler appends the current process status to the process
814 log which is later printed by exiwhat. It used to use the general
815 purpose logging code to do this, but several functions it calls are
816 not safe for signals.
818 The new output code in the SIGUSR1 handler is specific to the process
819 log, and simple enough that it's easy to inspect for signal safety.
820 Removing some special cases also simplifies the general logging code.
821 Removing the spurious timestamps from the process log simplifies
824 TF/04 Improved ratelimit ACL condition.
826 The /noupdate option has been deprecated in favour of /readonly which
827 has clearer semantics. The /leaky, /strict, and /readonly update modes
828 are mutually exclusive. The update mode is no longer included in the
829 database key; it just determines when the database is updated. (This
830 means that when you upgrde Exim will forget old rate measurements.)
832 Exim now checks that the per_* options are used with an update mode that
833 makes sense for the current ACL. For example, when Exim is processing a
834 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
835 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
836 must specify per_mail/readonly. If you omit the update mode it defaults to
837 /leaky where that makes sense (as before) or /readonly where required.
839 The /noupdate option is now undocumented but still supported for
840 backwards compatibility. It is equivalent to /readonly except that in
841 ACLs where /readonly is required you may specify /leaky/noupdate or
842 /strict/noupdate which are treated the same as /readonly.
844 A useful new feature is the /count= option. This is a generalization
845 of the per_byte option, so that you can measure the throughput of other
846 aggregate values. For example, the per_byte option is now equivalent
847 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
849 The per_rcpt option has been generalized using the /count= mechanism
850 (though it's more complicated than the per_byte equivalence). When it is
851 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
852 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
853 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
854 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
855 Note that using per_rcpt with a non-readonly update mode in more than
856 one ACL will cause the recipients to be double-counted. (The per_mail
857 and per_byte options don't have this problem.)
859 The handling of very low rates has changed slightly. If the computed rate
860 is less than the event's count (usually one) then this event is the first
861 after a long gap. In this case the rate is set to the same as this event's
862 count, so that the first message of a spam run is counted properly.
864 The major new feature is a mechanism for counting the rate of unique
865 events. The new per_addr option counts the number of different
866 recipients that someone has sent messages to in the last time period. It
867 behaves like per_rcpt if all the recipient addresses are different, but
868 duplicate recipient addresses do not increase the measured rate. Like
869 the /count= option this is a general mechanism, so the per_addr option
870 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
871 example, measure the rate that a client uses different sender addresses
872 with the options per_mail/unique=$sender_address. There are further
873 details in the main documentation.
875 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
877 TF/06 Removed a few PCRE remnants.
879 TF/07 Automatically extract Exim's version number from tags in the git
880 repository when doing development or release builds.
882 PP/02 Raise smtp_cmd_buffer_size to 16kB.
883 Bugzilla 879. Patch from Paul Fisher.
885 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
886 Heavily based on revision 40f9a89a from Simon Arlott's tree.
889 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
891 PP/05 Variable $av_failed, true if the AV scanner deferred.
892 Bugzilla 1078. Patch from John Horne.
894 PP/06 Stop make process more reliably on build failure.
895 Bugzilla 1087. Patch from Heiko Schlittermann.
897 PP/07 Make maildir_use_size_file an _expandable_ boolean.
898 Bugzilla 1089. Patch from Heiko Schlittermann.
900 PP/08 Handle ${run} returning more data than OS pipe buffer size.
901 Bugzilla 1131. Patch from Holger Weiß.
903 PP/09 Handle IPv6 addresses with SPF.
904 Bugzilla 860. Patch from Wolfgang Breyha.
906 PP/10 GnuTLS: support TLS 1.2 & 1.1.
908 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
911 PP/11 match_* no longer expand right-hand-side by default.
912 New compile-time build option, EXPAND_LISTMATCH_RHS.
913 New expansion conditions, "inlist", "inlisti".
915 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
917 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
919 PP/14 fix log_write() format string regression from TF/03.
920 Bugzilla 1152. Patch from Dmitry Isaikin.
926 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
928 PP/02 Harmonised TLS library version reporting; only show if debugging.
929 Layout now matches that introduced for other libraries in 4.74 PP/03.
931 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
933 PP/04 New "dns_use_edns0" global option.
935 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
938 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
939 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
941 TK/01 Updated PolarSSL code to 0.14.2.
942 Bugzilla 1097. Patch from Andreas Metzler.
944 PP/07 Catch divide-by-zero in ${eval:...}.
947 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
950 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
951 format-string attack -- SECURITY: remote arbitrary code execution.
953 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
954 time unintentionally subject to list matching rules, letting the header
955 cause arbitrary Exim lookups (of items which can occur in lists, *not*
956 arbitrary string expansion). This allowed for information disclosure.
958 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
959 INT_MIN/-1 -- value coerced to INT_MAX.
965 NM/01 Workround for PCRE version dependency in version reporting
968 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
969 This fixes portability to compilers other than gcc, notably
970 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
972 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
973 makefiles for portability to HP-UX and POSIX correctness.
975 PP/01 Permit LOOKUP_foo enabling on the make command-line.
976 Also via indented variable definition in the Makefile.
977 (Debugging by Oliver Heesakkers).
979 PP/02 Restore caching of spamd results with expanded spamd_address.
980 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
982 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
983 Improves build reliability. Fix from: Frank Elsner
985 NM/02 Fix wide character breakage in the rfc2047 coding
986 Fixes bug 1064. Patch from Andrey N. Oktyabrski
988 NM/03 Allow underscore in dnslist lookups
989 Fixes bug 1026. Patch from Graeme Fowler
991 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
992 Code patches from Adam Ciarcinski of NetBSD.
994 NM/04 Fixed exiqgrep to cope with mailq missing size issue
997 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
998 is logged, to avoid truncation. Patch from John Horne.
1000 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1001 Patch from Jakob Hirsch.
1003 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1004 SQL string expansion failure details.
1005 Patch from Andrey Oktyabrski.
1007 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1008 Patch from Simon Arlott.
1010 PP/09 New lookups functionality failed to compile on old gcc which rejects
1011 extern declarations in function scope.
1012 Patch from Oliver Fleischmann
1014 PP/10 Use sig_atomic_t for flags set from signal handlers.
1015 Check getgroups() return and improve debugging.
1016 Fixed developed for diagnosis in bug 927 (which turned out to be
1019 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1020 Patch from Mark Zealey.
1022 PP/12 Bugzilla 1056: Improved spamd server selection.
1023 Patch from Mark Zealey.
1025 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1026 Based on patch from Heiko Schlittermann.
1028 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1029 Patch from Uwe Doering, sign-off by Michael Haardt.
1031 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1032 variable declaration deep within a block. Bug and patch from
1035 PP/15 lookups-Makefile IRIX compatibilty coercion.
1037 PP/16 Make DISABLE_DKIM build knob functional.
1039 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1040 Patch by Simon Arlott
1042 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1043 variable argument macros. Our copy now differs from upstream.
1049 TF/01 Failure to get a lock on a hints database can have serious
1050 consequences so log it to the panic log.
1052 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1053 controlled using the smtp_confirmation log selector.
1055 TF/03 Include the error message when we fail to unlink a spool file.
1057 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1058 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1059 for maintaining out-of-tree patches for some time.
1061 PP/01 Bugzilla 139: Documentation and portability issues.
1062 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1063 Handle per-OS dynamic-module compilation flags.
1065 PP/02 Let /dev/null have normal permissions.
1066 The 4.73 fixes were a little too stringent and complained about the
1067 permissions on /dev/null. Exempt it from some checks.
1068 Reported by Andreas M. Kirchwitz.
1070 PP/03 Report version information for many libraries, including
1071 Exim version information for dynamically loaded libraries. Created
1072 version.h, now support a version extension string for distributors
1073 who patch heavily. Dynamic module ABI change.
1075 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1076 privilege escalation vulnerability whereby the Exim run-time user
1077 can cause root to append content of the attacker's choosing to
1080 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1083 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1084 If dropping privileges for untrusted macros, we disabled normal logging
1085 on the basis that it would fail; for the Exim run-time user, this is not
1086 the case, and it resulted in successful deliveries going unlogged.
1087 Fixed. Reported by Andreas Metzler.
1093 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1094 only prepend for the Resent-* case. Fixes regression introduced in
1095 Exim 4.70 by NM/22 for Bugzilla 607.
1097 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1098 increasing numbers of administrators be bitten by this.
1100 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1102 PP/03 Bugzilla 994: added openssl_options main configuration option.
1104 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1106 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1108 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1110 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1111 without a peer certificate, leading to a segfault because of an
1112 assumption that peers always have certificates. Be a little more
1113 paranoid. Problem reported by Martin Tscholak.
1115 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1116 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1117 NB: ClamAV planning to remove STREAM in "middle of 2010".
1118 CL also introduces -bmalware, various -d+acl logging additions and
1119 more caution in buffer sizes.
1121 PP/09 Implemented reverse_ip expansion operator.
1123 PP/10 Bugzilla 937: provide a "debug" ACL control.
1125 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1127 PP/12 Bugzilla 973: Implement --version.
1129 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1131 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1133 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1135 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1136 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1137 ignore trailing whitespace.
1139 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1141 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1144 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1145 Notification from Dr Andrew Aitchison.
1147 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1148 ExtendedDetectionInfo response format.
1149 Notification from John Horne.
1151 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1154 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1155 XSL and documented dependency on system catalogs, with examples of how
1158 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1161 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1162 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1163 configuration file which is writeable by the Exim user or group.
1165 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1166 of configuration files to cover files specified with the -C option if
1167 they are going to be used with root privileges, not just the default
1170 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1171 option (effectively making it always true).
1173 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1174 files to be used while preserving root privileges.
1176 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1177 that rogue child processes cannot use them.
1179 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1180 run-time user, instead of root.
1182 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1183 Exim run-time user without dropping privileges.
1185 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1186 result string, instead of calling string_vformat() twice with the same
1189 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1190 for other users. Others should always drop root privileges if they use
1191 -C on the command line, even for a whitelisted configure file.
1193 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1195 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1201 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1202 $data_path, and $header_path variables; fixed documentation bugs and
1205 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1206 exipick to access non-standard spools, including the "frozen" queue
1209 NM/01 Bugzilla 965: Support mysql stored procedures.
1210 Patch from Alain Williams
1212 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1214 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1215 Patch from Andreas Metzler
1217 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1218 Patch from Kirill Miazine
1220 NM/05 Bugzilla 671: Added umask to procmail example.
1222 JJ/03 installed exipick 20100323.0, fixing doc bug
1224 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1225 directory. Notification and patch from Dan Rosenberg.
1227 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1229 TK/02 Improve log output when DKIM signing operation fails.
1231 MH/01 Treat the transport option dkim_domain as a colon separated
1232 list, not as a single string, and sign the message with each element,
1233 omitting multiple occurences of the same signer.
1235 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1236 Bugzilla 985, 986. Patch by Simon Arlott
1238 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1239 Patch by Simon Arlott
1241 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1242 MBX locking. Notification from Dan Rosenberg.
1248 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1250 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1252 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1254 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1256 NM/04 Bugzilla 915: Flag broken perl installation during build.
1262 TK/01 Added patch by Johannes Berg that expands the main option
1263 "spamd_address" if it starts with a dollar sign.
1265 TK/02 Write list of recipients to X-Envelope-Sender header when building
1266 the mbox-format spool file for content scanning (suggested by Jakob
1269 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1270 (http://www.dcc-servers.net/) support via dccifd. Activated by
1271 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1273 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1274 by Mark Daniel Reidel <mr@df.eu>.
1276 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1277 When building exim an external PCRE library is now needed -
1278 PCRE is a system library on the majority of modern systems.
1279 See entry on PCRE_LIBS in EDITME file.
1281 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1282 conversation. Added nologin parameter to request.
1283 Patch contributed by Kirill Miazine.
1285 TF/01 Do not log submission mode rewrites if they do not change the address.
1287 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1289 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1290 log files in place. Contributed by Roberto Lima.
1292 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1294 TF/03 Bugzilla 615: When checking the local_parts router precondition
1295 after a local_part_suffix or local_part_prefix option, Exim now
1296 does not use the address's named list lookup cache, since this
1297 contains cached lookups for the whole local part.
1299 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1300 Robert Millan. Documentation is in experimental-spec.txt.
1302 TF/04 Bugzilla 668: Fix parallel build (make -j).
1304 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1306 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1307 Patch provided by Jan Srzednicki.
1309 TF/05 Leading white space used to be stripped from $spam_report which
1310 wrecked the formatting. Now it is preserved.
1312 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1313 that they are available at delivery time.
1315 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1317 TF/08 TLS error reporting now respects the incoming_interface and
1318 incoming_port log selectors.
1320 TF/09 Produce a more useful error message if an SMTP transport's hosts
1321 setting expands to an empty string.
1323 NM/06 Bugzilla 744: EXPN did not work under TLS.
1324 Patch provided by Phil Pennock.
1326 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1327 Patch provided by Richard Godbee.
1329 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1330 acl_smtp_notquit, added index entry.
1332 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1333 Patch provided by Eugene Bujak.
1335 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1336 accept(). Patch provided by Maxim Dounin.
1338 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1339 Patch provided by Phil Pennock.
1341 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1343 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1344 Patch provided by Brad "anomie" Jorsch.
1346 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1347 Patch provided by Dean Brooks.
1349 TK/05 Add native DKIM support (does not depend on external libraries).
1351 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1352 Patch provided by Graeme Fowler.
1354 NM/16 Bugzilla 851: Documentation example syntax fix.
1356 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1358 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1361 NM/19 Bugzilla 745: TLS version reporting.
1362 Patch provided by Phil Pennock.
1364 NM/20 Bugzilla 167: bool: condition support.
1365 Patch provided by Phil Pennock.
1367 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1368 clients. Patch provided by Phil Pennock.
1370 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1371 Patch provided by Brad "anomie" Jorsch.
1373 NM/23 Bugzilla 687: Fix misparses in eximstats.
1374 Patch provided by Heiko Schlittermann.
1376 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1377 Patch provided by Heiko Schlittermann.
1379 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1380 plus update to original patch.
1382 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1384 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1385 Patch provided by David Brownlee.
1387 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1389 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1391 NM/30 Bugzilla 888: TLS documentation bugfixes.
1393 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1395 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1396 Unlike the original bugzilla I have changed all shell scripts in src tree.
1398 NM/33 Bugzilla 898: Transport filter timeout fix.
1399 Patch by Todd Rinaldo.
1401 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1402 Patch by Serge Demonchaux.
1404 NM/35 Bugzilla 39: Base64 decode bug fixes.
1405 Patch by Jakob Hirsch.
1407 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1409 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1411 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1413 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1419 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1420 ALT-N's libdkim that I have put here:
1421 http://duncanthrax.net/exim-experimental/
1423 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1424 were called 'true' and it seems that C99 defines that as a reserved
1425 keyword to be used with 'bool' variable types. That means you could
1426 not include C99-style headers which use bools without triggering
1427 build errors in sieve.c.
1429 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1430 as mailq or other aliases. Changed the --help handling significantly
1431 to do whats expected. exim_usage() emits usage/help information.
1433 SC/01 Added the -bylocaldomain option to eximstats.
1435 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1437 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1439 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1445 PH/01 Another patch from the Sieve maintainer.
1447 PH/02 When an IPv6 address is converted to a string for single-key lookup
1448 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1449 dots are used instead of colons so that keys in lsearch files need not
1450 contain colons. This was done some time before quoting was made available
1451 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1452 (notated using the quote facility) so as to distinguish them from IPv4
1453 keys. This meant that lookups for IP addresses in host lists did not work
1454 for iplsearch lookups.
1456 This has been fixed by arranging for IPv6 addresses to be expressed with
1457 colons if the lookup type is iplsearch. This is not incompatible, because
1458 previously such lookups could never work.
1460 The situation is now rather anomolous, since one *can* have colons in
1461 ordinary lsearch keys. However, making the change in all cases is
1462 incompatible and would probably break a number of configurations.
1464 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1467 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1468 conversion specification without a maximum field width, thereby enabling
1469 a rogue spamd server to cause a buffer overflow. While nobody in their
1470 right mind would setup Exim to query an untrusted spamd server, an
1471 attacker that gains access to a server running spamd could potentially
1472 exploit this vulnerability to run arbitrary code as the Exim user.
1474 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1475 $primary_hostname instead of what libspf2 thinks the hosts name is.
1477 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1478 a directory entry by the name of the lookup key. Previously, if a
1479 symlink pointed to a non-existing file or a file in a directory that
1480 Exim lacked permissions to read, a lookup for a key matching that
1481 symlink would fail. Now it is enough that a matching directory entry
1482 exists, symlink or not. (Bugzilla 503.)
1484 PH/03 The body_linecount and body_zerocount variables are now exported in the
1487 PH/04 Added the $dnslist_matched variable.
1489 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1490 This means they are set thereafter only if the connection becomes
1493 PH/06 Added the client_condition to authenticators so that some can be skipped
1494 by clients under certain conditions.
1496 PH/07 The error message for a badly-placed control=no_multiline_responses left
1497 "_responses" off the end of the name.
1499 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1501 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1502 (without spaces) instead of just copying the configuration text.
1504 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1506 PH/11 Added $max_received_linelength.
1508 PH/12 Added +ignore_defer and +include_defer to host lists.
1510 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1511 way in which PCRE > 7.0 is built.
1513 PH/14 Implemented queue_only_load_latch.
1515 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1516 MAIL command. The effect was to mangle the value on 64-bit systems.
1518 PH/16 Another patch from the Sieve maintainer.
1520 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1522 PH/18 If a system quota error occurred while trying to create the file for
1523 a maildir delivery, the message "Mailbox is full" was not appended to the
1524 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1525 only to a quota excession during the actual writing of the file.
1527 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1528 characters?) which causes problems in log lines. The DN values are now
1529 passed through string_printing() before being added to log lines.
1531 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1532 and InterBase are left for another time.)
1534 PH/21 Added message_body_newlines option.
1536 PH/22 Guard against possible overflow in moan_check_errorcopy().
1538 PH/23 POSIX allows open() to be a macro; guard against that.
1540 PH/24 If the recipient of an error message contained an @ in the local part
1541 (suitably quoted, of course), incorrect values were put in $domain and
1542 $local_part during the evaluation of errors_copy.
1548 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1549 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1552 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1553 issue a MAIL command.
1555 PH/02 In an ACL statement such as
1557 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1559 if a client was not listed at all, or was listed with a value other than
1560 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1561 the condition was not true (as it should be), so access was not denied.
1562 The bug was that the ! inversion was incorrectly passed on to the second
1563 item. This has been fixed.
1565 PH/03 Added additional dnslists conditions == and =& which are different from
1566 = and & when the dns lookup returns more than one IP address.
1568 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1569 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1571 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1572 FSYNC, which compiles an option called disable_fsync that allows for
1573 bypassing fsync(). The documentation is heavily laced with warnings.
1575 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1577 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1578 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1579 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1580 including adding "make clean"; (3) Added -fPIC when compiling the test
1581 dynamically loaded module, to get rid of a warning.
1583 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1584 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1585 = 0s. The bug is otherwise harmless.
1587 PH/07 There was a bug in the dovecot authenticator such that the value of
1588 $auth1 could be overwritten, and so not correctly preserved, after a
1589 successful authentication. This usually meant that the value preserved by
1590 the server_setid option was incorrect.
1592 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1594 PH/09 Installed PCRE release 7.0.
1596 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1597 run for batched SMTP input. It is now run at the start of every message
1598 in the batch. While fixing this I discovered that the process information
1599 (output by running exiwhat) was not always getting set for -bs and -bS
1600 input. This is fixed, and it now also says "batched" for BSMTP.
1602 PH/11 Added control=no_pipelining.
1604 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1605 patch, slightly modified), and move the expansion of helo_data till after
1606 the connection is made in the smtp transport (so it can use these
1609 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1611 PH/14 Added log_selector = +pid.
1613 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1615 PH/16 Add ${if forany and ${if forall.
1617 PH/17 Added dsn_from option to vary the From: line in DSNs.
1619 PH/18 Flush SMTP output before performing a callout, unless control =
1620 no_callout_flush is set.
1622 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1623 was true (the default) a successful delivery failed to delete the retry
1624 item, thus causing premature timeout of the address. The bug is now
1627 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1629 PH/21 Long custom messages for fakedefer and fakereject are now split up
1630 into multiline reponses in the same way that messages for "deny" and
1631 other ACL rejections are.
1633 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1634 with slight modification.
1636 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1637 draft, changing the syntax and factoring some duplicate code".
1639 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1640 for deliveries of the second and subsequent messages over the same SMTP
1643 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1644 ${reduce, with only minor "tidies".
1646 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1648 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1649 expansion side effects.
1651 PH/27 When a message times out after an over-quota error from an Exim-imposed
1652 quota, the bounce message says "mailbox is full". This message was not
1653 being given when it was a system quota that was exceeded. It now should
1656 MH/03 Made $recipients available in local_scan(). local_scan() already has
1657 better access to the recipient list through recipients_list[], but
1658 $recipients can be useful in postmaster-provided expansion strings.
1660 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1661 in the case of a MAIL command with additional options following the
1662 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1663 were accidentally chopped off.
1665 PH/29 SMTP synchronization checks are implemented when a command is read -
1666 there is a check that no more input is waiting when there shouldn't be
1667 any. However, for some commands, a delay in an ACL can mean that it is
1668 some time before the response is written. In this time, more input might
1669 arrive, invalidly. So now there are extra checks after an ACL has run for
1670 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1671 pipelining has not been advertised.
1673 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1675 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1676 returned in the SMTP response when smtp_return_error_details was false.
1677 This has been fixed.
1679 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1680 instead of the C I/O that was originally supplied, because problems were
1681 reported on Solaris.
1683 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1684 Exim which did not show up earlier: it was assuming that a call to
1685 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1686 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1687 was a macro that became an assignment, so it seemed to work. This has
1688 changed to a proper function call with a void return, hence the compile
1689 error. Exim's code has been fixed.
1691 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1694 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1696 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1698 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1699 because it turns out to be too big or there is a timeout) there is no
1700 "Completed" line in the log. When some messages of this type were
1701 selected by exigrep, they were listed as "not completed". Others were
1702 picked up by some special patterns. I have improved the selection
1703 criteria to be more general.
1705 PH/38 The host_find_failed option in the manualroute router can now be set
1706 to "ignore", to completely ignore a host whose IP address cannot be
1707 found. If all hosts are ignored, the behaviour is controlled by the new
1708 host_all_ignored option.
1710 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1711 homing or because of multiple MX records with /mx) generated more than
1712 one IP address, and the following item turned out to be the local host,
1713 all the secondary addresses of the first item were incorrectly removed
1714 from the list, along with the local host and any following hosts (which
1715 is what is supposed to happen).
1717 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1718 whoever called Exim into the -H file. In the case of the daemon it was
1719 behaving confusingly. When first started, it used values for whoever
1720 started the daemon, but after a SIGHUP it used the Exim user (because it
1721 calls itself on a restart). I have changed the code so that it now always
1724 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1725 message are rejected with the same error (e.g. no authentication or bad
1726 sender address), and a DATA command is nevertheless sent (as can happen
1727 with PIPELINING or a stupid MUA), the error message that was given to the
1728 RCPT commands is included in the rejection of the DATA command. This is
1729 intended to be helpful for MUAs that show only the final error to their
1732 PH/42 Another patch from the Sieve maintainer.
1734 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1735 Eximstats - Fixed some broken HTML links and added missing column headers
1737 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1738 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1740 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1742 PH/43 Yet another patch from the Sieve maintainer.
1744 PH/44 I found a way to check for a TCP/IP connection going away before sending
1745 the response to the final '.' that terminates a message, but only in the
1746 case where the client has not sent further data following the '.'
1747 (unfortunately, this is allowed). However, in many cases there won't be
1748 any further data because there won't be any more messages to send. A call
1749 to select() can be used: if it shows that the input is "ready", there is
1750 either input waiting, or the socket has been closed. An attempt to read
1751 the next input character can distinguish the two cases. Previously, Exim
1752 would have sent an OK response which the client would never have see.
1753 This could lead to message repetition. This fix should cure that, at
1754 least in a lot of common cases.
1756 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1757 advertised in response to EHLO.
1763 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1764 fixed by 4.65/MH/01 (is this a record?) are fixed:
1766 (i) An empty string was always treated as zero by the numeric comparison
1767 operators. This behaviour has been restored.
1769 (ii) It is documented that the numeric comparison operators always treat
1770 their arguments as decimal numbers. This was broken in that numbers
1771 starting with 0 were being interpreted as octal.
1773 While fixing these problems I realized that there was another issue that
1774 hadn't been noticed. Values of message_size_limit (both the global option
1775 and the transport option) were treated as octal if they started with 0.
1776 The documentation was vague. These values are now always treated as
1777 decimal, and I will make that clear in the documentation.
1783 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1784 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1787 MH/01 Don't check that the operands of numeric comparison operators are
1788 integers when their expansion is in "skipping" mode (fixes bug
1789 introduced by 4.64-PH/07).
1791 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1792 child addresses, Exim now panics and dies. Previously, because the count
1793 is held in a short int, deliveries were likely to be lost. As such a
1794 large number of recipients for a single message is ridiculous
1795 (performance will be very, very poor), I have chosen to impose a limit
1796 rather than extend the field.
1802 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1803 leftover -K file (the existence of which was triggered by #402).
1804 While we were at it, introduced process PID as part of the -K
1805 filename. This should rule out race conditions when creating
1808 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1809 processing considerably. Previous code took too long for large mails,
1810 triggering a timeout which in turn triggers #401.
1812 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1813 in the DK code in transports.c. sendfile() is not really portable,
1814 hence the _LINUX specificness.
1816 TF/01 In the add_headers option to the mail command in an Exim filter,
1817 there was a bug that Exim would claim a syntax error in any
1818 header after the first one which had an odd number of characters
1821 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1822 callout verification, Exim cached a "reject" for the entire domain. This
1823 is correct for most verifications, but it is not correct for a recipient
1824 verification with use_sender or use_postmaster set, because in that case
1825 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1826 case of MAIL FROM:<> rejection from other early rejections (e.g.
1827 rejection of HELO). When verifying a recipient using a non-null MAIL
1828 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1829 Whatever the result of the callout, the value of the domain cache is
1830 left unchanged (for any other kind of callout, getting as far as trying
1831 RCPT means that the domain itself is ok).
1833 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1836 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1837 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1838 session, the daemon ignores ECONNECT errors and logs others; it now
1839 ignores EPIPE as well.
1841 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1842 (quoted-printable decoding).
1844 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1845 later the small subsequent patch to fix an introduced bug.
1847 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1849 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1851 PH/08 An error is now given if message_size_limit is specified negative.
1853 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1854 to be given (somewhat) arbitrary names.
1856 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1859 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1860 miscellaneous code fixes
1862 PH/10 Added the log_reject_target ACL modifier to specify where to log
1865 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1866 hostname. This is wrong, because it relates to the incoming message (and
1867 probably the interface on which it is arriving) and not to the outgoing
1868 callout (which could be using a different interface). This has been
1869 changed to use the value of the helo_data option from the smtp transport
1870 instead - this is what is used when a message is actually being sent. If
1871 there is no remote transport (possible with a router that sets up host
1872 addresses), $smtp_active_hostname is used.
1874 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1875 tweaks were necessary in order to get it to work (see also 21 below):
1876 (a) The code assumed that strncpy() returns a negative number on buffer
1877 overflow, which isn't the case. Replaced with Exim's string_format()
1879 (b) There were several signed/unsigned issues. I just did the minimum
1880 hacking in of casts. There is scope for a larger refactoring.
1881 (c) The code used strcasecmp() which is not a standard C function.
1882 Replaced with Exim's strcmpic() function.
1883 (d) The code set only $1; it now sets $auth1 as well.
1884 (e) A simple test gave the error "authentication client didn't specify
1885 service in request". It would seem that Dovecot has changed its
1886 interface. Fortunately there's a specification; I followed it and
1887 changed what the client sends and it appears to be working now.
1889 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1892 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1893 address A is aliased to B and C, where B exists and C does not. Without
1894 -v the output is "A verified" because verification stops after a
1895 successful redirection if more than one address is generated. However,
1896 with -v the child addresses are also verified. Exim was outputting "A
1897 failed to verify" and then showing the successful verification for C,
1898 with its parentage. It now outputs "B failed to verify", showing B's
1899 parentage before showing the successful verification of C.
1901 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1902 look up a TXT record in a specific list after matching in a combined
1905 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1906 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1907 they consult the DNS. I had assumed they would set it the way they
1908 wanted; and indeed my experiments on Linux seem to show that in some
1909 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1910 To be on the safe side, however, I have now made the interface to
1911 host_find_byname() similar to host_find_bydns(), with an argument
1912 containing the DNS resolver options. The host_find_byname() function now
1913 sets these options at its start, just as host_find_bydns() does. The smtp
1914 transport options dns_qualify_single and dns_search_parents are passed to
1915 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1916 of host_find_byname() use the default settings of RES_DEFNAMES
1917 (qualify_single) but not RES_DNSRCH (search_parents).
1919 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1920 spool_read_header() do less string testing, by means of a preliminary
1921 switch on the second character of optional "-foo" lines. (This is
1922 overdue, caused by the large number of possibilities that now exist.
1923 Originally there were few.) While I was there, I also converted the
1924 str(n)cmp tests so they don't re-test the leading "-" and the first
1925 character, in the hope this might squeeze out yet more improvement.
1927 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1928 flag allowing group syntax was set by the header_syntax check but not
1929 turned off, possible causing trouble later; (2) The flag was not being
1930 set at all for the header_verify test, causing "group"-style headers to
1931 be rejected. I have now set it in this case, and also caused header_
1932 verify to ignore an empty address taken from a group. While doing this, I
1933 came across some other cases where the code for allowing group syntax
1934 while scanning a header line wasn't quite right (mostly, not resetting
1935 the flag correctly in the right place). These bugs could have caused
1936 trouble for malformed header lines. I hope it is now all correct.
1938 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1939 with the "reply" argument non-NULL. The code, however (which originally
1940 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1941 but it didn't always do it. This confused somebody who was copying the
1942 code for some other use. I have removed all the tests.
1944 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1945 feature that was used to support insecure browsers during the U.S. crypto
1946 embargo. It requires special client support, and Exim is probably the
1947 only MTA that supported it -- and would never use it because real RSA is
1948 always available. This code has been removed, because it had the bad
1949 effect of slowing Exim down by computing (never used) parameters for the
1950 RSA_EXPORT functionality.
1952 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1953 authenticator to fail if there's a tab character in the incoming data
1954 (there should never be unless someone is messing about, as it's supposed
1955 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1956 if the connection is using TLS or if the remote IP is the same as the
1957 local IP, and the "valid-client-cert option" if a client certificate has
1960 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1961 authenticators. This can be used for authorization after authentication
1962 succeeds. (In the case of plaintext, it servers for both authentication
1965 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1966 if any retry times were supplied.
1968 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1969 connection, where there is no client IP address to check. In this
1970 situation, the verify now always succeeds.
1972 PH/25 Applied John Jetmore's -Mset patch.
1974 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1976 PH/27 In a string expansion for a processed (not raw) header when multiple
1977 headers of the same name were present, leading whitespace was being
1978 removed from all of them, but trailing whitespace was being removed only
1979 from the last one. Now trailing whitespace is removed from each header
1980 before concatenation. Completely empty headers in a concatenation (as
1981 before) are ignored.
1983 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1984 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1986 PH/29 [Removed. This was a change that I later backed out, and forgot to
1987 correct the ChangeLog entry (that I had efficiently created) before
1988 committing the later change.]
1990 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1991 address errors (4xx response to RCPT) over the same connection as other
1992 messages routed to the same hosts. Such deliveries are always "forced",
1993 so retry times are not inspected. This resulted in far too many retries
1994 for the affected addresses. The effect occurred only when there were more
1995 hosts than the hosts_max_try setting in the smtp transport when it had
1996 the 4xx errors. Those hosts that it had tried were not added to the list
1997 of hosts for which the message was waiting, so if all were tried, there
1998 was no problem. Two fixes have been applied:
2000 (i) If there are any address or message errors in an SMTP delivery, none
2001 of the hosts (tried or untried) are now added to the list of hosts
2002 for which the message is waiting, so the message should not be a
2003 candidate for sending over the same connection that was used for a
2004 successful delivery of some other message. This seems entirely
2005 reasonable: after all the message is NOT "waiting for some host".
2006 This is so "obvious" that I'm not sure why it wasn't done
2007 previously. Hope I haven't missed anything, but it can't do any
2008 harm, as the worst effect is to miss an optimization.
2010 (ii) If, despite (i), such a delivery is accidentally attempted, the
2011 routing retry time is respected, so at least it doesn't keep
2012 hammering the server.
2014 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2015 in ${readsocket because some servers need this prod.
2017 PH/32 Added some extra debug output when updating a wait-xxx database.
2019 PH/33 The hint "could be header name not terminated by colon", which has been
2020 given for certain expansion errors for a long time, was not being given
2021 for the ${if def:h_colon_omitted{... case.
2023 PH/34 The spec says: "With one important exception, whenever a domain list is
2024 being scanned, $domain contains the subject domain." There was at least
2025 one case where this was not true.
2027 PH/35 The error "getsockname() failed: connection reset by peer" was being
2028 written to the panic log as well as the main log, but it isn't really
2029 panic-worthy as it just means the connection died rather early on. I have
2030 removed the panic log writing for the ECONNRESET error when getsockname()
2033 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2034 runs only) independently of the message's sender address. This meant
2035 that, if the 4xx error was in fact related to the sender, a different
2036 message to the same recipient with a different sender could confuse
2037 things. In particualar, this can happen when sending to a greylisting
2038 server, but other circumstances could also provoke similar problems.
2039 I have changed the default so that the retry time for these errors is now
2040 based a combination of the sender and recipient addresses. This change
2041 can be overridden by setting address_retry_include_sender=false in the
2044 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2045 remote server are returned as part of bounce messages. This was not
2046 happening for LMTP over a pipe (the lmtp transport), but now it is the
2047 same for both kinds of LMTP.
2049 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2050 in header lines that were in fact CNAMEs. This is no longer the case.
2052 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2053 and queue runs started by the daemon processed all messages. This has
2054 been fixed so that -R and -S can now usefully be given with -q<time>.
2056 PH/40 Import PCRE release 6.7 (fixes some bugs).
2058 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2060 PH/42 Give an error if -q is specified more than once.
2062 PH/43 Renamed the variables $interface_address and $interface_port as
2063 $received_ip_address and $received_port, to make it clear that these
2064 values apply to message reception, and not to the outgoing interface when
2065 a message is delivered. (The old names remain recognized, of course.)
2067 PH/44 There was no timeout on the connect() call when using a Unix domain
2068 socket in the ${readsocket expansion. There now is.
2070 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2071 be meaningful with "accept".
2073 SC/01 Eximstats V1.43
2074 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2076 SC/02 Eximstats V1.44
2077 Use a glob alias rather than an array ref in the generated
2078 parser. This improves both readability and performance.
2080 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2081 Collect SpamAssassin and rejection statistics.
2082 Don't display local sender or destination tables unless
2083 there is data to show.
2084 Added average volumes into the top table text output.
2086 SC/04 Eximstats V1.46
2087 Collect data on the number of addresses (recipients)
2088 as well as the number of messages.
2090 SC/05 Eximstats V1.47
2091 Added 'Message too big' to the list of mail rejection
2092 reasons (thanks to Marco Gaiarin).
2094 SC/06 Eximstats V1.48
2095 Mainlog lines which have GMT offsets and are too short to
2096 have a flag are now skipped.
2098 SC/07 Eximstats V1.49 (Alain Williams)
2099 Added the -emptyok flag.
2101 SC/08 Eximstats V1.50
2102 Fixes for obtaining the IP address from reject messages.
2104 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2105 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2106 whitesspace changes from 4.64-PH/27
2108 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2111 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2112 are found, allow negative numbers in numeric criteria)
2114 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2116 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2117 to match changes made in 4.64-PH/43
2119 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2121 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2122 "domain" is actually an IP address, and force a failure. This locks out
2123 those revolvers/nameservers that support "A-for-A" lookups, in
2124 contravention of the specifications.
2126 PH/48 When a host name was looked up from an IP address, and the subsequent
2127 forward lookup of the name timed out, the host name was left in
2128 $sender_host_name, contrary to the specification.
2130 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2131 restricted to single-key lookups, Exim was not diagnosing an error if
2132 * or *@ was used with a query-style lookup.
2134 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2136 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2137 long ago, but noone interested enough thought of it. Let's just say that
2138 the "1.1" means that there are some new functions that weren't there at
2139 some point in the past.
2141 PH/51 Error processing for expansion failure of helo_data from an smtp
2142 transport during callout processing was broken.
2144 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2145 tested/used via the -bh/-bhc/-bs options.
2147 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2148 bug, fixed in subsequent PCRE releases).
2150 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2151 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2153 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2159 SC/01 Use a glob alias rather than an array ref in eximstats generated
2160 parser. This improves both readability and performance.
2162 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2163 Don't display local sender or destination tables in eximstats unless
2164 there is data to show.
2165 Added average volumes into the eximstats top table text output.
2167 SC/03 Collect data on the number of addresses (recipients) as well
2168 as the number of messages in eximstats.
2170 TF/01 Correct an error in the documentation for the redirect router. Exim
2171 does (usually) call initgroups() when daemonizing.
2173 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2174 with consistent privilege compared to when running as a daemon.
2176 TF/03 Note in the spec that $authenticated_id is not set for local
2177 submissions from trusted users.
2179 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2180 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2182 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2183 by adding some example configuration directives to the default
2184 configuration file. A little bit of work is required to uncomment the
2185 directives and define how usernames and passwords are checked, but
2186 there is now a framework to start from.
2188 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2189 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2190 without this. I don't know how relevant this is to other LDAP libraries.
2192 PH/02 Add the verb name to the "unknown ACL verb" error.
2194 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2196 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2198 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2199 directory not expanded when it should be if an expanded home directory
2200 was set for the address (which is overridden by the transport).
2202 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2205 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2206 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2207 because it is too late at that time, and has no effect.
2209 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2210 security issue with \' (bugzilla #107). I could not use the
2211 PQescapeStringConn() function, because it needs a PGconn value as one of
2214 PH/08 When testing addresses using -bt, indicate those final addresses that
2215 are duplicates that would not cause an additional delivery. At least one
2216 person was confused, thinking that -bt output corresponded to deliveries.
2217 (Suppressing duplicates isn't a good idea as you lose the information
2218 about possibly different redirections that led to the duplicates.)
2220 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2221 systems where poll() doesn't work, in particular OS X.
2223 PH/10 Added more information to debugging output for retry time not reached.
2225 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2226 operations in malware.c.
2228 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2231 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2232 syslog_duplication set false, log lines that would normally be written
2233 both the the main log and to the reject log were not written to syslog at
2236 PH/14 In the default configuration, change the use of "message" in ACL warn
2237 statements to "add_header".
2239 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2240 not followed by a command (e.g. "seen endif").
2242 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2243 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2246 PH/17 Added extra conditions to the default value of delay_warning_condition
2250 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2251 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2252 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2255 The Auto-Submitted: and various List- headers are standardised, whereas I
2256 don't think Precedence: ever was.
2258 PH/18 Refactored debugging code in route_finduser() to show more information,
2259 in particular, the error code if getpwnam() issues one.
2261 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2262 This is apparently needed in addition to the PH/07 change above to avoid
2263 any possible encoding problems.
2265 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2266 but not after initializing Perl.
2268 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2269 output them only if debugging. By default they are written stderr,
2270 apparently, which is not desirable.
2272 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2275 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2278 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2280 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2281 authenticated or an ident call has been made. Suppress the default
2282 values for $authenticated_id and $authenticated_sender (but permit -oMai
2283 and -oMas) when testing with -bh.
2285 PH/24 Re-jigged the order of the tests in the default configuration so that the
2286 tests for valid domains and recipients precede the DNS black list and CSA
2287 tests, on the grounds that those ones are more expensive.
2289 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2290 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2291 This bug exists in every version of Exim that I still have, right back to
2294 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2295 However, an attempt to turn on case-sensitivity in a regex key by
2296 including (?-i) didn't work because the subject string was already
2297 lowercased, and the effects were non-intuitive. It turns out that a
2298 one-line patch can be used to allow (?-i) to work as expected.
2304 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2305 other effects) broke the use of negated acl sub-conditions.
2307 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2310 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2311 "Deny" causes Exim to reject the incoming connection with a 554 error.
2312 Unfortunately, if there is a major crisis, such as a disk failure,
2313 tcp-wrappers gives "deny", whereas what one would like would be some
2314 kind of temporary error. A kludge has been added to help with this.
2315 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2316 554 error is used if errno is still zero or contains ENOENT (which occurs
2317 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2320 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2322 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2323 errors. Otherwise a message that provokes a temporary error (when other
2324 messages do not) can cause a whole host to time out.
2326 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2327 addresses were routed directly to files or pipes from a redirect router.
2328 File deliveries just didn't batch; pipe deliveries might have suffered
2331 PH/06 A failure to get a lock for a hints database would erroneously always say
2332 "Failed to get write lock", even when it was really a read lock.
2334 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2335 of 0600. This has been changed to use the value of the lockfile_mode
2336 option (which defaults to 0600).
2338 PH/08 Applied small patch from the Sieve maintainer.
2340 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2341 folder from quota calculations, a direct delivery into this folder messed
2342 up the contents of the maildirsize file. This was because the regex was
2343 used only to exclude .Trash (or whatever) when the size of the mailbox
2344 was calculated. There was no check that a delivery was happening into an
2345 excluded directory. This bug has been fixed by ignoring all quota
2346 processing for deliveries into excluded directories.
2348 PH/10 Added the maildirfolder_create_regex option to appendfile.
2354 PH/01 The code for finding all the local interface addresses on a FreeBSD
2355 system running IPv6 was broken. This may well have applied to all BSD
2356 systems, as well as to others that have similar system calls. The broken
2357 code found IPv4 interfaces correctly, but gave incorrect values for the
2358 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2359 that it would not match correctly against @[] and not recognize the IPv6
2362 PH/02 The ipliteral router was not recognizing addresses of the form user@
2363 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2365 PH/03 Added disable_ipv6.
2367 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2368 decoded form, because it is most often used to construct To: headers
2369 lines in autoreplies, and the decoded form may well be syntactically
2370 invalid. However, $reply_address has leading white space removed, and all
2371 newlines turned into spaces so that the autoreply transport does not
2374 PH/05 If group was specified without a user on a router, and no group or user
2375 was specified on a transport, the group from the router was ignored.
2377 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2378 for visible compile-time settings that can be used to change these
2379 numbers, for those that want even more. Backwards compatibility with old
2380 spool files has been maintained. However, going back to a previous Exim
2381 release will lost any variables that are in spool files.
2383 PH/07 Two small changes when running in the test harness: increase delay when
2384 passing a TCP/IP connection to a new process, in case the original
2385 process has to generate a bounce, and remove special handling of
2386 127.0.0.2 (sic), which is no longer necessary.
2388 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2389 be the same on different OS.
2391 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2394 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2395 whether --show-vars was specified or not
2397 JJ/02 exipick: Added support for new ACL variable spool format introduced
2400 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2401 syntactically invalid From: or Reply-to: line, and a filter used this to
2402 generate an autoreply, and therefore failed to obtain an address for the
2403 autoreply, Exim could try to deliver to a non-existent relative file
2404 name, causing unrelated and misleading errors. What now happens is that
2405 it logs this as a hard delivery error, but does not attempt to create a
2408 PH/11 The exinext utility has a -C option for testing purposes, but although
2409 the given file was scanned by exinext itself; it wasn't being passed on
2410 when Exim was called.
2412 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2413 an end-of-file indication when reading a command response.
2415 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2416 compiled. In many other places in Exim, IPv6 addresses are always
2417 recognized, so I have changed this. It also means that IPv4 domain
2418 literals of the form [IPV4:n.n.n.n] are now always recognized.
2420 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2421 used if the router is not running as root, for example, when verifying at
2422 ACL time, or when using -bh. The debugging output from this situation was
2423 non-existent - all you got was a failure to exec. I have made two
2426 (a) Failures to set uid/gid, the current directory, or a process leader
2427 in a subprocess such as that created by queryprogram now generate
2428 suitable debugging ouput when -d is set.
2430 (b) The queryprogram router detects when it is not running as root,
2431 outputs suitable debugging information if -d is set, and then runs
2432 the subprocess without attempting to change uid/gid.
2434 PH/15 Minor change to Makefile for building test_host (undocumented testing
2437 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2438 additional section of a DNS packet that returns MX or SRV records.
2439 Instead, it always explicitly searches for A/AAAA records. This avoids
2440 major problems that occur when a DNS server includes only records of one
2441 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2442 fixed another bug: if SRV records were looked up and the corresponding
2443 address records were *not* found in the additional section, the port
2444 values from the SRV records were lost.
2446 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2447 using the correct key (the original address) when searching the retry
2448 rules in order to find which one to use for generating the retry hint.
2450 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2451 adding the default one. Similarly, if it contains a Reply-To: header, the
2452 errors_reply_to option, if set, is not used.
2454 PH/19 When calculating a retry time, Exim used to measure the "time since
2455 failure" by looking at the "first failed" field in the retry record. Now
2456 it does not use this if it is later than than the arrival time of the
2457 message. Instead it uses the arrival time. This makes for better
2458 behaviour in cases where some deliveries succeed, thus re-setting the
2459 "first failed" field. An example is a quota failure for a huge message
2460 when small messages continue to be delivered. Without this change, the
2461 "time since failure" will always be short, possible causing more frequent
2462 delivery attempts for the huge message than are intended.
2463 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2465 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2466 $1, $2, $3) because the numerical variables can be reset during some
2467 expansion items (e.g. "match"), thereby losing the authentication data.
2469 PH/21 Make -bV show the size of off_t variables so that the test suite can
2470 decide whether to run tests for quotas > 2G.
2472 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2473 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2474 filecount value is greater than 2G or if a quota value is greater than 2G
2475 on a system where the size of off_t is not greater than 4, a panic error
2478 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2479 never match. The debug and -bh output now contains an explicit error
2480 message indicating a malformed IPv4 address or mask.
2482 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2483 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2484 PH/23 above applies.
2486 PH/25 Do not write to syslog when running in the test harness. The only
2487 occasion when this arises is a failure to open the main or panic logs
2488 (for which there is an explicit test).
2490 PH/26 Added the /no_tell option to "control=freeze".
2492 PH/27 If a host name lookup failed very early in a connection, for example, if
2493 the IP address matched host_lookup and the reverse lookup yielded a name
2494 that did not have a forward lookup, an error message of the form "no IP
2495 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2496 could be logged. Now it outputs the IP address instead of "NULL".
2498 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2499 allows the sender and the authenticated sender to be set when
2500 submitting a message from within Exim. Since child_open_exim() is
2501 documented for local_scan(), the new function should be too.
2503 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2504 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2505 results in an empty string is now treated as unset.
2507 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2509 PH/31 Added sender_verify_fail logging option.
2511 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2512 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2513 tidied the source and removed it altogether.
2515 PH/33 When a queue run was abandoned because the load average was too high, a
2516 log line was always written; now it is written only if the queue_run log
2517 selector is set. In addition, the log line for abandonment now contains
2518 information about the queue run such as the pid. This is always present
2519 in "start" and "stop" lines but was omitted from the "abandon" line.
2521 PH/34 Omit spaces between a header name and the colon in the error message that
2522 is given when verify = headers_syntax fails (if there are lots of them,
2523 the message gets confusing).
2525 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2526 names, as there are now some PTR records that contain slashes. This check
2527 is only to protect against broken name servers that fall over on strange
2528 characters, so the fact that it applies to all lookups doesn't matter.
2530 PH/36 Now that the new test suite is complete, we can remove some of the
2531 special code in Exim that was needed for the old test suite. For example,
2532 sorting DNS records because real resolvers return them in an arbitrary
2533 order. The new test suite's fake resolver always returns records in the
2536 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2537 bounces) except when queue_only is set, to avoid logging races between
2538 the different processes.
2540 PH/38 Panic-die if .include specifies a non-absolute path.
2542 PH/39 A tweak to the "H" retry rule from its user.
2544 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2545 a label. They prevented compilation on older perls.
2547 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2548 a warning to be raised on newish perls.
2550 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2551 on queue. Changes to match documented behaviour of showing count of
2552 messages matching specified criteria.
2554 PH/40 Changed the default ident timeout from 30s to 5s.
2556 PH/41 Added support for the use of login_cap features, on those BSD systems
2557 that have them, for controlling the resources used by pipe deliveries.
2559 PH/42 The content-scanning code uses fopen() to create files in which to put
2560 message data. Previously it was not paying any attention to the mode of
2561 the files. Exim runs with umask(0) because the rest of the code creates
2562 files with open(), and sets the required mode explicitly. Thus, these
2563 files were ending up world-writeable. This was not a big issue, because,
2564 being within the spool directory, they were not world-accessible. I have
2565 created a function called modefopen, which takes an additional mode
2566 argument. It sets umask(777), creates the file, chmods it to the required
2567 mode, then resets the umask. All the relevant calls to fopen() in the
2568 content scanning code have been changed to use this function.
2570 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2571 to 24 hours. This avoids potential overflow problems when processing G
2572 and H retry rules. I suspect nobody ever tinkers with this value.
2574 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2576 PH/45 When the plaintext authenticator is running as a client, the server's
2577 challenges are checked to ensure they are valid base64 strings. By
2578 default, the authentication attempt is cancelled if an invalid string is
2579 received. Setting client_ignore_invalid_base64 true ignores these errors.
2580 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2581 they are received. Thus, the responses can be made to depend on the
2582 challenges. If an invalid string is ignored, an empty string is placed in
2585 PH/46 Messages that are created by the autoreply transport now contains a
2586 References: header, in accordance with RFCs 2822 and 3834.
2588 PH/47 Added authenticated_sender_force to the smtp transport.
2590 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2592 PH/49 Installed latest patch from the Sieve maintainer.
2594 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2595 was also set, the appendfile transport was unnecessarily scanning a
2596 directory of message files (e.g. for maildir delivery) to find the count
2597 of files (along with the size), even though it did not need this
2598 information. It now does the scan only if it needs to find either the
2599 size of the count of files.
2601 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2603 PH/52 Two bugs concerned with error handling when the smtp transport is
2606 (i) Exim was not creating retry information for temporary errors given
2607 for individual recipients after the DATA command when the smtp transport
2608 was used in LMTP mode. This meant that they could be retried too
2609 frequently, and not timed out correctly.
2611 (ii) Exim was setting the flag that allows error details to be returned
2612 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2613 recipients that were returned after the DATA command.
2615 PH/53 This is related to PH/52, but is more general: for any failing address,
2616 when detailed error information was permitted to be returned to the
2617 sender, but the error was temporary, then after the final timeout, only
2618 "retry timeout exceeded" was returned. Now it returns the full error as
2619 well as "retry timeout exceeded".
2621 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2622 do this, and (what is worse) MTAs that accept it.
2624 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2625 will now be deprecated.
2627 PH/56 New os.c-cygwin from the Cygwin maintainer.
2629 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2630 formats (previously only available in exim formats via -bpr, -bpru,
2631 and -bpra. Now also available in native and exiqgrep formats)
2633 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2634 with very large, slow to parse queues
2636 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2638 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2640 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2641 responses to authentication challenges, though it was showing the
2642 challenges; (ii) I've removed the CR characters from the debug output for
2645 PH/58 Allow for the insertion of a newline as well as a space when a string
2646 is turned into more than one encoded-word during RFC 2047 encoding. The
2647 Sieve code now uses this.
2649 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2650 data_4xx, lost_connection, tls_required.
2652 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2653 message was being sent as an SMTP response.
2655 PH/61 Add -l and -k options to exicyclog.
2657 PH/62 When verifying, if an address was redirected to one new address, so that
2658 verification continued, and the new address failed or deferred after
2659 having set something in $address_data, the value of $address_data was not
2660 passed back to the ACL. This was different to the case when no
2661 redirection occurred. The value is now passed back in both cases.
2663 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2664 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2665 use login_cap.h, so on its own it isn't the distinguishing feature. The
2666 new name refers directly to the setclassresources() function.
2668 PH/65 Added configuration files for NetBSD3.
2670 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2672 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2673 is preferred over IPv4.
2675 PH/68 The bounce_return_message and bounce_return_body options were not being
2676 honoured for bounces generated during the reception of non-SMTP messages.
2677 In particular, this applied to messages rejected by the ACL. This bug has
2678 been fixed. However, if bounce_return_message is true and bounce_return_
2679 body is false, the headers that are returned for a non-SMTP message
2680 include only those that have been read before the error was detected.
2681 (In the case of an ACL rejection, they have all been read.)
2683 PH/69 The HTML version of the specification is now built in a directory called
2684 spec_html instead of spec.html, because the latter looks like a path with
2685 a MIME-type, and this confuses some software.
2687 PH/70 Catch two compiler warnings in sieve.c.
2689 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2690 function verify_get_ident() calls ip_connect() to connect a socket, but
2691 if the "connect()" function timed out, ip_connect() used to close the
2692 socket. However, verify_get_ident() also closes the socket later, and in
2693 between Exim writes to the log, which may get opened at this point. When
2694 the socket was closed in ip_connect(), the log could get the same file
2695 descriptor number as the socket. This naturally causes chaos. The fix is
2696 not to close the socket in ip_connect(); the socket should be closed by
2697 the function that creates it. There was only one place in the code where
2698 this was missing, in the iplookup router, which I don't think anybody now
2699 uses, but I've fixed it anyway.
2701 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2702 well as to direct DNS lookups. Otherwise the handling of names in host
2703 lists is inconsistent and therefore confusing.
2709 PH/01 Two changes to the default runtime configuration:
2711 (1) Move the checks for relay_from_hosts and authenticated clients from
2712 after to before the (commented out) DNS black list checks.
2714 (2) Add control=submission to the relay_from_hosts and authenticated
2715 clients checks, on the grounds that messages accepted by these
2716 statements are most likely to be submissions.
2718 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2720 (1) Generate an error if the third argument for the ${prvs expansion is
2723 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2726 (3) Reset the variables that are obtained from the first argument of
2727 ${prvscheck and used in the second argument before leaving the code,
2728 because their memory is reclaimed, so using them afterwards may do
2731 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2732 one (it's much easier than Tom thought :-).
2734 (5) Because of (4), we can now allow for the use of $prvscheck_result
2735 inside the third argument.
2737 PH/03 For some reason, the default setting of PATH when running a command from
2738 a pipe transport was just "/usr/bin". I have changed it to
2741 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2742 anything to be listed in the output from -bV.
2744 PH/05 When a filter generated an autoreply, the entire To: header line was
2745 quoted in the delivery log line, like this:
2747 => >A.N.Other <ano@some.domain> <original@ddress> ...
2749 This has been changed so that it extracts the operative address. There
2750 may be more than one such address. If so, they are comma-separated, like
2753 => >ano@some.domain,ona@other.domain <original@ddress> ...
2755 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2756 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2757 not being looked up in the rDNS to get a host name, Exim was showing the
2758 IP address twice in Received: lines, even though the IP addresses were
2759 identical. For example:
2761 Received: from [1.2.3.4] (helo=[1.2.3.4])
2763 However, if the real host name was known, it was omitting the HELO data
2764 if it matched the actual IP address. This has been tidied up so that it
2765 doesn't show the same IP address twice.
2767 PH/07 When both +timestamp and +memory debugging was on, the value given by
2768 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2769 called by the debug printing, thereby overwriting the timestamp buffer.
2770 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2772 PH/08 When the original message was included in an autoreply transport, it
2773 always said "this is a copy of the message, including all the headers",
2774 even if body_only or headers_only was set. It now gives an appropriate
2777 PH/09 Applied a patch from the Sieve maintainer which:
2779 o fixes some comments
2780 o adds the (disabled) notify extension core
2781 o adds some debug output for the result of if/elsif tests
2782 o points to the current vacation draft in the documentation
2783 and documents the missing references header update
2787 o fixes a bug in processing the envelope test (when testing
2788 multiple envelope elements, the last element determinted the
2791 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2792 Electronic Mail") by including:
2794 Auto-submitted: auto-generated
2796 in the messages that it generates (bounce messages and others, such as
2797 warnings). In the case of bounce messages for non-SMTP mesages, there was
2798 also a typo: it was using "Auto_submitted" (underscore instead of
2799 hyphen). Since every message generated by Exim is necessarily in response
2800 to another message, thes have all been changed to:
2802 Auto-Submitted: auto-replied
2804 in accordance with these statements in the RFC:
2806 The auto-replied keyword:
2808 - SHOULD be used on messages sent in direct response to another
2809 message by an automatic process,
2811 - MUST NOT be used on manually-generated messages,
2813 - MAY be used on Delivery Status Notifications (DSNs) and Message
2814 Disposition Notifications (MDNs),
2816 - MUST NOT be used on messages generated by automatic or periodic
2817 processes, except for messages which are automatic responses to
2820 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2821 to the default Received: header definition.
2823 PH/12 Added log selector acl_warn_skipped (default on).
2825 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2826 variables because (a) they are in the wrong storage pool and (b) even if
2827 they were copied, it wouldn't work properly because of the caching.
2829 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2830 checking when decoding. Apparently there are clients that generate
2831 overlong encoded strings. Why am I not surprised?
2833 PH/15 If the first argument of "${if match_address" was not empty, but did not
2834 contain an "@" character, Exim crashed. Now it writes a panic log message
2835 and treats the condition as false.
2837 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2839 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2840 extension "envelope-auth". The code is finished and in agreement with
2841 other implementations, but there is no documentation so far and in fact,
2842 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2843 not changing the active code.
2845 Print executed "if" and "elsif" statements when debugging is used. This
2846 helps a great deal to understand what a filter does.
2848 Document more things not specified clearly in RFC3028. I had all this
2849 sorted out, when out of a sudden new issues came to my mind. Oops."
2851 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2854 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2855 canonical form (as documented). However, after a host name lookup from
2856 the IP address, check_host() was doing a simple string comparison with
2857 addresses acquired from the DNS when checking that the found name did
2858 have the original IP as one of its addresses. Since any found IPv6
2859 addresses are likely to be in abbreviated form, the comparison could
2860 fail. Luckily, there already exists a function for doing the comparison
2861 by converting both addresses to binary, so now that is used instead of
2862 the text comparison.
2864 PH/20 There was another similar case to PH/19, when a complete host name was
2865 given in a host list; looking up its IP address could give an abbreviated
2866 form, whereas the current host's name might or might not be abbreviated.
2867 The same fix has been applied.
2873 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2874 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2877 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2878 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2880 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2882 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2883 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2884 Also, for "Auto-Submitted", treat anything other than "no" as
2885 non-personal, in accordance with RFC 3834. (Previously it treated
2886 anything starting "auto-" as non-personal.)
2888 TF/01 The control=submission/name=... option had a problem with syntax
2889 errors if the name included a slash character. The /name= option
2890 now slurps the rest of the string, so it can include any characters
2891 but it must come last in the list of options (after /sender_retain
2894 PH/05 Some modifications to the interface to the fake nameserver for the new
2902 TK/01 Added the "success_on_redirect" address verification option. See
2903 NewStuff for rationale and an example.
2905 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2907 PH/02 Patch to exigrep to allow it to work on syslog lines.
2909 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2910 fread() to skip over the body file's header line, because in Cygwin the
2911 header line is locked and is inaccessible.
2913 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2914 co-exist for some time) to make it clear that it is the Exim ID that is
2915 referenced, not the Message-ID: header line.
2917 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2918 string_format() function, because snprintf() does not exist on all
2921 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2922 ${stat: expansion item.
2924 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2925 protocol synchronization error", to keep the pedants happy.
2927 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2928 well as for IRIX systems, when gcc is being used. See the host.c source
2931 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2933 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2936 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2937 transport and to the smtp transport in LMTP mode.
2939 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2941 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2942 run a filter in a subprocess. This could lead to confusion in subsequent
2943 lookups in the parent process. There should also be a search_tidyup() at
2944 the end of the subprocess.
2946 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2947 only if the host matched helo_try_verify_hosts, which caused the
2948 verification to occur when the EHLO/HELO command was issued. The ACL just
2949 tested the remembered result. Now, if a previous verification attempt has
2950 not happened, "verify = helo" does it there and then.
2952 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2954 TK/03 Fix log output including CR from clamd.
2956 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2957 exist provoked a memory error which could cause a segfault.
2959 PH/15 Installed PCRE 6.2
2961 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2963 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2964 of the problem. Specifically, suggested +O2 rather than +O1 for the
2967 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2969 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2970 record, the port was not correctly logged when the outgoing_port log
2971 selector was set (it logged the transort's default port).
2973 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2974 fallback_hosts, and "hosts" in the smtp transport.
2976 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2977 host errors such as "Connection refused".
2979 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2980 authentication with radiusclient 0.4.9:
2982 - Error returned from rc_read_config was caught wrongly
2983 - Username/password not passed on to radius server due to wrong length.
2985 The presumption is that some radiusclient API changes for 4.51/PH/17
2986 were not taken care of correctly. The code is still untested by me (my
2987 Linux distribution still has 0.3.2 of radiusclient), but it was
2988 contributed by a Radius user.
2990 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2991 expanding the "port" option of the smtp transport.
2993 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2994 while reading a MIME header. Thanks to Tom Hughes for a patch.
2996 PH/24 Include config.h inside local_scan.h so that configuration settings are
2999 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3000 that in an ACL for RCPT (for example), you can examine exactly what was
3003 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3004 commands, but it was not correctly comparing the address with the actual
3005 client host address. Thus, it would show the EHLO address in Received:
3006 header lines when this was not necessary.
3008 PH/27 Added the % operator to ${eval:}.
3010 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3011 it should be ignoring failures (because with -C, for example, it has lost
3012 privilege). It wasn't ignoring creation failures other than "already
3015 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3018 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3019 because an input line was too long, either on its own, or by virtue of
3020 too many continuations, the temporary file was not being removed, and the
3021 return code was incorrect.
3023 PH/31 Missing "BOOL" in function definition in filtertest.c.
3025 PH/32 Applied Sieve patches from the maintainer.
3027 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3029 PH/33 Added "verify = not_blind".
3031 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3032 Local/Makefile (with some defaults set). These are used in built scripts
3033 such as exicyclog, but they have never been used in the exim_install
3034 script (though there are many overriding facilities there). I have
3035 arranged that the exim_install script now takes note of these two
3038 PH/35 Installed configuration files for Dragonfly.
3040 PH/36 When a locally submitted message by a trusted user did not contain a
3041 From: header, and the sender address was obtained from -f or from an SMTP
3042 MAIL command, and the trusted user did not use -F to supply a sender
3043 name, $originator_name was incorrectly used when constructing a From:
3044 header. Furthermore, $originator_name was used for submission mode
3045 messages from external hosts without From: headers in a similar way,
3046 which is clearly wrong.
3048 PH/37 Added control=suppress_local_fixups.
3050 PH/38 When log_selector = +received_sender was set, and the addition of the
3051 sender made the log line's construction buffer exactly full, or one byte
3052 less than full, an overflow happened when the terminating "\n" was
3055 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3056 when the result of a list match is failure because a DNS lookup failed.
3058 PH/40 RM_COMMAND is now used in the building process.
3060 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3061 the "build-* directories that it finds.
3063 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3064 address in a domain literal was a prefix of an interface address.
3066 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3067 when verifying a sender address, unless rewrite_headers is false.
3069 PH/44 Wrote a long comment about why errors_to addresses are verified as
3070 recipients, not senders.
3072 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3073 the ratelimit ACL was added.
3075 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3077 PH/46 Added extra information about PostgreSQL errors to the error string.
3079 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3080 avoiding the need to install special zones in a real server. This is
3081 backwards compatible; if it can't find the fake resolver, it drops back.
3082 Thus, both old and new test suites can be run.
3084 TF/02 Added util/ratelimit.pl
3086 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3087 clock is set back in time.
3089 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3090 Candler <B.Candler@pobox.com>.
3092 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3093 OK for addresses that are the result of redirections.
3095 PH/48 A number of further additions for the benefit of the new test suite,
3096 including a fake gethostbyname() that interfaces to the fake DNS resolver
3099 TF/06 The fix for widen_domains has also been applied to qualify_single and
3100 search_parents which are the other dnslookup options that can cause
3103 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3106 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3108 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3109 service" if it encountered an MX record with an empty target hostname.
3110 The message is now "an MX or SRV record indicated no SMTP service".
3112 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3113 if the DNS of the sending site is misconfigured. This is quite a
3114 common situation. This change restores the behaviour of treating a
3115 helo verification defer as a failure.
3117 PH/51 If self=fail was set on a router, the bounce message did not include the
3118 actual error message.
3124 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3126 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3127 command itself ended in error, the underlying message about the transport
3128 filter timeout was being overwritten with the pipe command error. Now the
3129 underlying error message should be appended to the second error message.
3131 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3133 PH/02 Reduce the amount of output that "make" produces by default. Full output
3134 can still be requested.
3136 PH/03 The warning log line about a condition test deferring for a "warn" verb
3137 was being output only once per connection, rather than after each
3138 occurrence (because it was using the same function as for successful
3139 "warn" verbs). This seems wrong, so I have changed it.
3141 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3142 it should not have, which might have caused a crash in the right
3143 circumstances, but probably never did.
3145 PH/04 Installed a modified version of Tony Finch's patch to make submission
3146 mode fix the return path as well as the Sender: header line, and to
3147 add a /name= option so that you can make the user's friendly name appear
3150 TF/03 Added the control = fakedefer ACL modifier.
3152 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3153 Mark Lowes for thorough testing.
3155 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3157 TK/03 Merged latest SRS patch from Miles Wilton.
3159 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3160 with the definition in sysexits.h (which is #included earlier).
3161 Fortunately, Exim does not actually use EX_OK. The code used to try to
3162 preserve the sysexits.h value, by assumimg that macro definitions were
3163 scanned for macro replacements. I have been disabused of this notion,
3164 so now the code just undefines EX_OK before #including unistd.h.
3166 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3167 in the smtp transport. When a block could not be written in a single
3168 write() function, the timeout was being re-applied to each part-write.
3169 This seems wrong - if the receiver was accepting one byte at a time it
3170 would take for ever. The timeout is now adjusted when this happens. It
3171 doesn't have to be particularly precise.
3173 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3174 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3176 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3177 without a domain if the check to <postmaster@domain> fails.
3179 SC/01 Eximstats: added -xls and the ability to specify output files
3180 (patch written by Frank Heydlauf).
3182 SC/02 Eximstats: use FileHandles for outputing results.
3184 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3186 SC/04 Eximstats: fixed display of large numbers with -nvr option
3188 SC/05 Eximstats: fixed merging of reports with empty tables.
3190 SC/06 Eximstats: added the -include_original_destination flag
3192 SC/07 Eximstats: removed tabs and trailing whitespace.
3194 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3196 TK/06 MBOX spool code: Add real "From " MBOX separator line
3197 so the .eml file is really in mbox format (even though
3198 most programs do not really care). Patch from Alex Miller.
3200 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3201 The latter is generated from $received_to and is only set if the
3202 message has one envelope recipient. SA can use these headers,
3203 obviously out-of-the-box. Patch from Alex Miller.
3205 PH/08 The ${def test on a variable was returning false if the variable's
3206 value was "0", contrary to what the specification has always said!
3207 The result should be true unless the variable is empty.
3209 PH/09 The syntax error of a character other than { following "${if
3210 def:variable_name" (after optional whitespace) was not being diagnosed.
3211 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3212 accidental colon was present, for example, could give incorrect results.
3214 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3215 result is used (not including appendfile, where other changes are about
3218 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3219 This involved changing a lot of size variables from int to off_t. It
3220 should work with maildirs and everything.
3222 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3223 spamd dying while we are connected to it.
3225 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3228 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3229 function for BDB 4.3.
3231 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3233 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3234 the sources for PCRE have been re-arranged and more files are now
3237 PH/15 The code I had for printing potentially long long variables in PH/11
3238 above was not the best (it lost precision). The length of off_t variables
3239 is now inspected at build time, and an appropriate printing format (%ld
3240 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3241 to be "long long int" or "long int". This is needed for the internal
3242 formatting function string_vformat().
3244 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3245 the configuration file to be ":syslog", then the script "guesses" where
3246 the logs files are, rather than using the compiled in default. In our
3247 case the guess is not the same as the compiled default, so the script
3248 suddenly stopped working when I started to use syslog. The patch checks
3249 to see if log_file_path is "". If so, it attempts to read it from exim
3250 with no configuration file to get the compiled in version, before it
3251 falls back to the previous guessing code."
3253 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3254 implementing BATV in an Exim configuration. See NewStuff for the gory
3257 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3258 Makefile that are specific to HP-UX.
3260 PH/18 If the "use_postmaster" option was set for a recipient callout together
3261 with the "random" option, the postmaster address was used as the MAIL
3262 FROM address for the random test, but not for the subsequent recipient
3263 test. It is now used for both.
3265 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3266 patch removes a few documentation additions to RFC 3028, because the
3267 latest draft now contains them. It adds the new en;ascii-case comparator
3268 and a new error check for 8bit text in MIME parts. Comparator and
3269 require names are now matched exactly. I enabled the subaddress
3270 extension, but it is not well tested yet (read: it works for me)."
3272 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3273 rework some of the code of TK/09 above to avoid the hardwired use of
3274 "%lld" and "long long". Replaced the call to snprintf() with a call to
3277 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3278 records point to non-existent hosts", "retry timeout exceeded", and
3279 "retry time not reached for any host after a long failure period".
3281 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3282 experimental DomainKeys support:
3284 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3285 (2) On an error such as an illegally used "control", the wrong name for
3286 the control was given.
3288 These problems did NOT occur unless DomainKeys support was compiled.
3290 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3292 PH/24 Added ${if match_ip condition.
3294 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3295 databases so that it will be absolutely obvious if a crash occurs in the
3296 DB library. This is a regular occurrence (often caused by mis-matched
3299 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3300 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3301 picked up on a user's system that detects such things. There doesn't seem
3302 to be a gcc warning option for this - only an attribute that has to be
3303 put on the function's prototype. It seems that in Fedora Core 4 they have
3304 set this on a number of new functions. No doubt there will be more in due
3307 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3308 specify a transport. However, if an address that was verified by such a
3309 router was the subject of a callout, Exim crashed because it tried to
3310 read the rcpt_include_affixes from the non-existent transport. Now it
3311 just assumes that the setting of that option is false. This bug was
3312 introduced by 4.51/PH/31.
3314 PH/28 Changed -d+all to exclude +memory, because that information is very
3315 rarely of interest, but it makes the output a lot bigger. People tend to
3316 do -d+all out of habit.
3318 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3319 code in os-type was giving problems when libc.so lives in lib64, like on
3322 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3323 aren't the modern standard, and it seems that some systems' include files
3324 don't always have them. Exim was already checking for some of the newer
3325 ones like T_AAAA, and defining it itself. I've added checks for all the
3326 record types that Exim uses.
3328 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3329 not automatically generating a new one, as it is supposed to. This
3330 prevented TLS from working. If the file did exist, but contained invalid
3331 data, a new version was generated, as expected. It was only the case of a
3332 non-existent file that was broken.
3334 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3335 with a change in libdomainkeys > 0.64.
3337 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3338 from DNS. If the selector record carries the flag, it now has
3339 precedence over the domain-wide flag.
3341 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3343 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3344 the use of a port name that isn't defined in /etc/services) occurred, the
3345 message was deferred as in a normal delivery, and thus remained on the
3346 spool, instead of being failed because of the mua_wrapper setting. This
3347 is now fixed, and I tidied up some of the mua_wrapper messages at the
3350 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3351 the messages in a hash of arrays rather than using individual hashes.
3352 This is a bit cleaner and results in dramatic memory savings, albeit
3353 at a slight CPU cost.
3355 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3356 as requested by Marc Sherman.
3358 SC/10 Eximstats: added histograms for user specified patterns as requested
3361 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3363 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3364 fopen() in the content-scanning modules that did not already have it.
3370 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3371 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3373 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3375 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3377 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3378 to test Sieve filters that use "vacation".
3380 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3381 that changes the way the GnuTLS parameters are stored in the cache file.
3382 The new format can be generated externally. For backward compatibility,
3383 if the data in the cache doesn't make sense, Exim assumes it has read an
3384 old-format file, and it generates new data and writes a new file. This
3385 means that you can't go back to an older release without removing the
3388 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3389 work if there are any delivery delays because "one_time" forces the
3390 parent to be marked "delivered", so its unseen clone is never tried
3391 again. For this reason, Exim now forbids the simultaneous setting of
3394 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3395 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3396 entry if you want to know the details. The fix, however, neglected to
3397 consider the case when local delivery batching is involved. The test for
3398 "previously delivered" was not happening when checking to see if an
3399 address could be batched with a previous (undelivered) one; under
3400 certain circumstances this could lead to multiple deliveries to the same
3403 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3404 in its include files, and this causes problems building Exim.
3406 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3407 header_syntax) but Exim was just ignoring anything given after a slash.
3408 In particular, this caused confusion with an attempt to use "verify =
3409 reverse_host_lookup/defer_ok". An error is now given when options are
3410 supplied for verify items that do not have them. (Maybe reverse_host_
3411 lookup should have a defer_ok option, but that's a different point.)
3413 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3414 defined by RFC 821) to 2048, because there were problems with some AUTH
3415 commands, and RFC 1869 says the size should be increased for extended
3416 SMTP commands that take arguments.
3418 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3421 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3422 "unknown" error; now it says that the functionality isn't in the binary.
3424 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3425 an address' error message when a string expansion fails (syntax or
3426 whatever). Otherwise the password may appear in the log. Following change
3427 PH/42 below, there is no longer a chance of it appearing in a bounce
3430 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3432 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3433 of its addresses was ever tried. (Bugzilla bug #2.)
3435 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3436 the result incorrectly in the debug output. (It correctly added a newline
3437 to what was transported.)
3439 TF/01 Added $received_time.
3441 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3442 commented out examples of how to interface to a virus scanner and to
3443 SpamAssassin. Also added commented examples of av_scanner and
3444 spamd_address settings.
3446 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3447 and controls are allowed in which ACLs. There were a couple of minor
3448 errors. Some of the entries in the conditions table (which is a table of
3449 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3450 negation of where the condition IS allowed.
3452 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3454 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3455 header file does not have a version number, so I've had to invent a new
3456 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3457 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3458 radiusclient), but it was contributed by a Radius user.
3460 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3461 files or directories, for OpenSSL.
3463 PH/19 When an Exim process that is running as root has to create an Exim log
3464 file, it does so in a subprocess that runs as exim:exim so as to get the
3465 ownership right at creation (otherwise, other Exim processes might see
3466 the file with the wrong ownership). There was no test for failure of this
3467 fork() call, which would lead to the process getting stuck as it waited
3468 for a non-existent subprocess. Forks do occasionally fail when resources
3469 run out. I reviewed all the other calls to fork(); they all seem to check
3472 PH/20 When checking for unexpected SMTP input at connect time (before writing
3473 the banner), Exim was not dealing correctly with a non-positive return
3474 from the read() function. If the client had disconnected by this time,
3475 the result was a log entry for a synchronization error with an empty
3476 string after "input=" when read() returned zero. If read() returned -1
3477 (an event I could not check), uninitialized data bytes were printed.
3478 There were reports of junk text (parts of files, etc) appearing after
3481 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3483 PH/22 Added support for macro redefinition, and (re)definition in between
3484 driver and ACL definitions.
3486 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3487 forgetting to use the resulting value; it was using the unexpanded value.
3489 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3490 hadn't been configured. The fix is from Juergen Kreileder, who
3491 understands it better than I do:
3493 "Here's what I see happening with three configured cyrus_sasl
3494 authenticators configured (plain, login, cram-md5):
3496 On startup auth_cyrus_sasl_init() gets called for each of these.
3497 This means three calls to sasl_listmech() without a specified mech_list.
3498 => SASL tests which mechs of all available mechs actually work
3499 => three warnings about OTP not working
3500 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3502 With the patch, sasl_listmech() also gets called three times. But now
3503 SASL's mech_list option is set to the server_mech specified in the the
3504 authenticator. Or in other words, the answer from sasl_listmech()
3505 gets limited to just the mech you're testing for (which is different
3507 => the return list contains just 'plain' or 'login', 'cram-md5' or
3508 nothing depending on the value of ob->server_mech.
3510 I've just tested the patch: Authentication still works fine,
3511 unavailable mechs specified in the exim configuration are still
3512 caught, and the auth.log warnings about OTP are gone."
3514 PH/25 When debugging is enabled, the contents of the command line are added
3515 to the debugging output, even when log_selector=+arguments is not
3518 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3519 answer is "GNU", and only if the return is "GNU/something" is the answer
3522 PH/27 $acl_verify_message is now set immediately after the failure of a
3523 verification in an ACL, and so is available in subsequent modifiers. In
3524 particular, the message can be preserved by coding like this:
3526 warn !verify = sender
3527 set acl_m0 = $acl_verify_message
3529 Previously, $acl_verify_message was set only while expanding "message"
3530 and "log_message" when a very denied access.
3532 PH/28 Modified OS/os.c-Linux with
3534 -#ifndef OS_LOAD_AVERAGE
3535 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3537 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3538 nomenclature these days.)
3540 PH/29 Installed patch from the Sieve maintainer that adds the options
3541 sieve_useraddress and sieve_subaddress to the redirect router.
3543 PH/30 In these circumstances:
3544 . Two addresses routed to the same list of hosts;
3545 . First host does not offer TLS;
3546 . First host accepts first address;
3547 . First host gives temporary error to second address;
3548 . Second host offers TLS and a TLS session is established;
3549 . Second host accepts second address.
3550 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3551 and peerdn, if requested) that were in fact used only for the second
3554 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3555 attention to any local part prefix or suffix that was matched by the
3556 router that accepted the address. It now behaves in the same way as it
3557 does for delivery: the affixes are removed from the local part unless
3558 rcpt_include_affixes is set on the transport.
3560 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3561 timeout during the DATA phase of an incoming message.
3563 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3564 applied a patch sanctioned by the Sieve maintainer.
3566 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3567 the uid or gid is negative. A case of a negative gid caused this to be
3568 noticed. The fix allows for either to be negative.
3570 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3571 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3574 PH/36 The change PH/12 above was broken. Fixed it.
3576 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3577 the grounds that routing the same address twice would always produce the
3578 same answer. This might have been true once, but it is certainly no
3579 longer true now. Routing a child address may depend on the previous
3580 routing that produced that child. Some complicated redirection strategies
3581 went wrong when messages had multiple recipients, and made Exim's
3582 behaviour dependent on the order in which the addresses were given.
3584 I have moved the duplicate checking until after the routing is complete.
3585 Exim scans the addresses that are assigned to local and remote
3586 transports, and removes any duplicates. This means that more work will be
3587 done, as duplicates will always all be routed, but duplicates are
3588 presumably rare, so I don't expect this is of any significance.
3590 For deliveries to pipes, files, and autoreplies, the duplicate checking
3591 still happens during the routing process, since they are not going to be
3594 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3595 It corrects a timeout issue with spamd. This is Ian's comment: "The
3596 background is that sometimes spamd either never reads data from a
3597 connection it has accepted, or it never writes response data. The exiscan
3598 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3599 blindly assumes that writes won't block so it may never time out."
3601 PH/39 Allow G after quota size as well as K and M.
3603 PH/40 The value set for $authenticated_id in an authenticator may not contain
3604 binary zeroes or newlines because the value is written to log lines and
3605 to spool files. There was no check on this. Now the value is run through
3606 the string_printing() function so that such characters are converted to
3607 printable escape sequences.
3609 PH/41 $message_linecount is a new variable that contains the total number of
3610 lines in the message. Compare $body_linecount, which is the count for the
3613 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3614 bounce and delay warning messages, except in certain special cases, which
3617 (a) An SMTP error message from a remote host;
3618 (b) A message specified in a :fail: redirection;
3619 (c) A message specified in a "fail" command in a system filter;
3620 (d) A message specified in a FAIL return from the queryprogram router;
3621 (e) A message specified by the cannot_route_message router option.
3623 In these cases only, Exim does include the error details in bounce and
3624 warning messages. There are also a few cases where bland messages such
3625 as "unrouteable address" or "local delivery error" are given.
3627 PH/43 $value is now also set for the "else" part of a ${run expansion.
3629 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3630 being worked on, but at least Exim now implements the latest version to
3633 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3634 process to complete was treated as a delivery failure, a timeout while
3635 writing the message to the pipe was logged, but erroneously treated as a
3636 successful delivery. Such timeouts include transport filter timeouts. For
3637 consistency with the overall process timeout, these timeouts are now
3638 treated as errors, giving rise to delivery failures by default. However,
3639 there is now a new Boolean option for the pipe transport called
3640 timeout_defer, which, if set TRUE, converts the failures into defers for
3641 both kinds of timeout. A transport filter timeout is now identified in
3644 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3645 systems where "make" and "gmake" are different, calling "gmake" at top
3646 level broke things. I've arranged for the value of $(MAKE) to be passed
3647 from the Makefile to this script so that it can call the same version of
3651 A note about Exim versions 4.44 and 4.50
3652 ----------------------------------------
3654 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3655 changes of various kinds. As a consequence, a big documentation update was
3656 needed. This delayed the release for rather longer than seemed good, especially
3657 in the light of a couple of (minor) security issues. Therefore, the changes
3658 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3659 release. So 4.44 and 4.50 are in effect two different branches that both start
3662 I have left the 4.50 change log unchanged; it contains all the changes since
3663 4.43. The change log for 4.44 is below; many of its items are identical to
3664 those for 4.50. This seems to be the most sensible way to preserve the
3665 historical information.
3671 1. Minor wording change to the doc/README.SIEVE file.
3673 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3674 computation of the current number of files was incorrect.
3676 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3677 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3680 4. Give more explanation in the error message when the command for a transport
3681 filter fails to execute.
3683 5. There are several places where Exim runs a non-Exim command in a
3684 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3685 was being done only for the command run by the queryprogram router. It is
3686 now done for all such subprocesses. The other cases are: ${run, transport
3687 filters, and the commands run by the lmtp and pipe transports.
3689 6. Added CONFIGURE_GROUP build-time option.
3691 7. Some older OS have a limit of 256 on the maximum number of file
3692 descriptors. Exim was using setrlimit() to set 1000 as a large value
3693 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3694 systems. I've change it so that if it can't get 1000, it tries for 256.
3696 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3697 was an oversight, and furthermore, ever since the addition of extra
3698 controls (e.g. 4.43/32), the checks on when to allow different forms of
3699 "control" were broken. There should now be diagnostics for all cases when a
3700 control that does not make sense is encountered.
3702 9. Added the /retain_sender option to "control=submission".
3704 10. $recipients is now available in the predata ACL (oversight).
3706 11. Tidy the search cache before the fork to do a delivery from a message
3707 received from the command line. Otherwise the child will trigger a lookup
3708 failure and thereby defer the delivery if it tries to use (for example) a
3709 cached ldap connection that the parent has called unbind on.
3711 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3712 of $address_data from the recipient verification was clobbered by the
3713 sender verification.
3715 13. The value of address_data from a sender verification is now available in
3716 $sender_address_data in subsequent conditions in the ACL statement.
3718 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3720 15. Added a new option "connect=<time>" to callout options, to set a different
3723 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3724 was its contents. (It was OK if the option was not defined at all.)
3726 17. A "Completed" log line is now written for messages that are removed from
3727 the spool by the -Mrm option.
3729 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3730 information about exactly what failed.
3732 19. Added -dd to debug only the daemon process.
3734 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3735 handles timeouts, both on the server side and network timeouts. Renamed the
3736 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3738 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3739 It is now set to "smtps".
3741 22. $host_address is now set to the target address during the checking of
3742 ignore_target_hosts.
3744 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3745 being passed; this would have caused $sender_host_name to have been used if
3746 matching the list had actually called for a host name (not very likely,
3747 since this list is usually IP addresses). A host name is now passed as
3750 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3751 code that specifies a non-restarting handler (typically sigaction() in
3752 modern systems) in an attempt to fix a rare and obscure crash bug.
3754 25. Narrowed the window for a race in the daemon that could cause it to ignore
3755 SIGCHLD signals. This is not a major problem, because they are used only to
3756 wake it up if nothing else does.
3758 26. A malformed maildirsize file could cause Exim to calculate negative values
3759 for the mailbox size or file count. Odd effects could occur as a result.
3760 The maildirsize information is now recalculated if the size or filecount
3763 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3764 support for a long time. Removed HAVE_SYS_VFS_H.
3766 28. Installed the latest version of exipick from John Jetmore.
3768 29. In an address list, if the pattern was not a regular expression, an empty
3769 subject address (from a bounce message) matched only if the pattern was an
3770 empty string. Non-empty patterns were not even tested. This was the wrong
3771 because it is perfectly reasonable to use an empty address as part of a
3772 database query. An empty address is now tested by patterns that are
3773 lookups. However, all the other forms of pattern expect the subject to
3774 contain a local part and a domain, and therefore, for them, an empty
3775 address still always fails if the pattern is not itself empty.
3777 30. Exim went into a mad DNS loop when attempting to do a callout where the
3778 host was specified on an smtp transport, and looking it up yielded more
3779 than one IP address.
3781 31. Re-factored the code for checking spool and log partition space into a
3782 function that finds that data and another that does the check. The former
3783 is then used to implement four new variables: $spool_space, $log_space,
3784 $spool_inodes, and $log_inodes.
3786 32. The RFC2047 encoding function was originally intended for short strings
3787 such as real names; it was not keeping to the 75-character limit for
3788 encoded words that the RFC imposes. It now respects the limit, and
3789 generates multiple encoded words if necessary. To be on the safe side, I
3790 have increased the buffer size for the ${rfc2047: expansion operator from
3793 33. It is now permitted to omit both strings after an "if" condition; if the
3794 condition is true, the result is "true". As before, when the second string
3795 is omitted, a false condition yields an empty string. This makes it less
3796 cumbersome to write custom ACL and router conditions.
3798 34. Failure to deliver a bounce message always caused it to be frozen, even if
3799 there was an errors_to setting on the router. The errors_to setting is now
3802 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3803 canonical form (fully expanded) before being placed in
3804 $sender_host_address.
3806 36. The table in the code that translates DNS record types into text (T_A to
3807 "A" for instance) was missing entries for NS and CNAME. It is just possible
3808 that this could have caused confusion if both these types were looked up
3809 for the same domain, because the text type is used as part of Exim's
3810 per-process caching. But the chance of anyone hitting this buglet seems
3813 37. The dnsdb lookup has been extended in a number of ways.
3815 (1) There is a new type, "zns", which walks up the domain tree until it
3816 finds some nameserver records. It should be used with care.
3818 (2) There is a new type, "mxh", which is like "mx" except that it returns
3819 just the host names, not the priorities.
3821 (3) It is now possible to give a list of domains (or IP addresses) to be
3822 looked up. The behaviour when one of the lookups defers can be
3823 controlled by a keyword.
3825 (4) It is now possible to specify the separator character for use when
3826 multiple records are returned.
3828 38. The dnslists ACL condition has been extended: it is now possible to supply
3829 a list of IP addresses and/or domains to be looked up in a particular DNS
3832 39. Added log_selector=+queue_time_overall.
3834 40. When running the queue in the test harness, wait just a tad after forking a
3835 delivery process, to get repeatability of debugging output.
3837 41. Include certificate and key file names in error message when GnuTLS fails
3838 to set them up, because the GnuTLS error message doesn't include the name
3839 of the failing file when there is a problem reading it.
3841 42. Allow both -bf and -bF in the same test run.
3843 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3845 44. The "Exiscan patch" is now merged into the mainline Exim source.
3847 45. Sometimes the final signoff response after QUIT could fail to get
3848 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3849 < 0 before doing a fflush(). This bug looks as though it goes back to the
3850 introduction of TLS in release 3.20, but "sometimes" must have been rare
3851 because the tests only now provoked it.
3853 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3854 (this can affect the format of dates).
3856 47. exim_tidydb, when checking for the continued existence of a message for
3857 which it has found a message-specific retry record, was not finding
3858 messages that were in split spool directories. Consequently, it was
3859 deleting retry records that should have stayed in existence.
3861 48. Steve fixed some bugs in eximstats.
3863 49. The SPA authentication driver was not abandoning authentication and moving
3864 on to the next authenticator when an expansion was forced to fail,
3865 contradicting the general specification for all authenticators. Instead it
3866 was generating a temporary error. It now behaves as specified.
3868 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3869 (the order specifies the preference for clients). The order is now AES256,
3870 AES128, 3DES, ARCFOUR128.
3872 51. Small patch to Sieve code - explicitly set From: when generating an
3875 52. Exim crashed if a remote delivery caused a very long error message to be
3876 recorded - for instance if somebody sent an entire SpamAssassin report back
3877 as a large number of 550 error lines. This bug was coincidentally fixed by
3878 increasing the size of one of Exim's internal buffers (big_buffer) that
3879 happened as part of the Exiscan merge. However, to be on the safe side, I
3880 have made the code more robust (and fixed the comments that describe what
3883 53. Now that there can be additional text after "Completed" in log lines (if
3884 the queue_time_overall log selector is set), a one-byte patch to exigrep
3885 was needed to allow it to recognize "Completed" as not the last thing in
3888 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3889 patch that reportedly fixes this has been added. I am not expert enough to
3890 create a test for it. This is what the patch creator wrote:
3892 "I found a little strange behaviour of ldap code when working with
3893 Windows 2003 AD Domain, where users was placed in more than one
3894 Organization Units. When I tried to give exim partial DN, the exit code
3895 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3896 But simultaneously result of request was absolutely normal ldap result,
3897 so I produce this patch..."
3899 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3900 REFERENCE, so I have modified the code to exclude the patch when that macro
3903 55. Some experimental protocols are using DNS PTR records for new purposes. The
3904 keys for these records are domain names, not reversed IP addresses. The
3905 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3906 leaves it alone. Component reversal etc. now happens only for IP addresses.
3909 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3911 57. Double the size of the debug message buffer (to 2048) so that more of very
3912 long debug lines gets shown.
3914 58. The exicyclog utility now does better if the number of log files to keep
3915 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3917 59. Two changes related to the smtp_active_hostname option:
3919 (1) $smtp_active_hostname is now available as a variable.
3920 (2) The default for smtp_banner uses $smtp_active_hostname instead
3921 of $primary_hostname.
3923 60. The host_aton() function is supposed to be passed a string that is known
3924 to be a valid IP address. However, in the case of IPv6 addresses, it was
3925 not checking this. This is a hostage to fortune. Exim now panics and dies
3926 if the condition is not met. A case was found where this could be provoked
3927 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3928 components; fortuitously, this particular loophole had already been fixed
3929 by change 4.50/55 above.
3931 If there are any other similar loopholes, the new check in host_aton()
3932 itself should stop them being exploited. The report I received stated that
3933 data on the command line could provoke the exploit when Exim was running as
3934 exim, but did not say which command line option was involved. All I could
3935 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3936 running as the user.
3939 61. There was a buffer overflow vulnerability in the SPA authentication code
3940 (which came originally from the Samba project). I have added a test to the
3941 spa_base64_to_bits() function which I hope fixes it.
3944 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3945 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3947 63. The daemon start-up calls getloadavg() while still root for those OS that
3948 need the first call to be done as root, but it missed one case: when
3949 deliver_queue_load_max is set with deliver_drop_privilege. This is
3950 necessary for the benefit of the queue runner, because there is no re-exec
3951 when deliver_drop_privilege is set.
3953 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3954 This has been fixed.
3956 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3957 was in use, was not putting the data itself into the right store pool;
3958 consequently, it could be overwritten for a subsequent message in the same
3959 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3962 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3964 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3965 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3966 the function were treating the return as a boolean value, which happened to
3967 work because 0=false and not-0=true, but is not correct code.
3969 68. The host_aton() function was not handling scoped IPv6 addresses (those
3970 with, for example, "%eth0" on the end) correctly.
3972 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3973 negated items (that is, ~something) in unsigned ints. Some compilers
3974 apparently mutter when there is no cast.
3976 70. If an address verification called from an ACL failed, and did not produce a
3977 user-specific message (i.e. there was only a "system" message), nothing was
3978 put in $acl_verify_message. In this situation, it now puts the system
3981 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3982 session; change 4.31/43 added the unwanted input to the log line - except
3983 that it did not do this in the start of session case. It now does.
3985 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3986 This is wrong and can cause the other end to generate a synchronization
3987 error if it is another Exim or anything else that does the synchronization
3988 check. A QUIT command is no longer sent after a timeout.
3990 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3991 during host lookups.
3993 74. The defer_ok option of callout verification was not working if it was used
3994 when verifying addresses in header lines, that is, for this case:
3996 verify = header_sender/callout=defer_ok
3998 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3999 those file descriptors could be used for SMTP connections. If anything
4000 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4001 could be sent to the SMTP client, causing chaos. The daemon now opens
4002 stdin, stdout, and stderr to /dev/null when it puts itself into the
4005 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4006 log by default. The user can override this with suitable Perl magic.
4008 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4009 the log message when discard triggers, was not working for the DATA ACL or
4010 for the non-SMTP ACL.
4012 78. Error message wording change in sieve.c.
4014 79. If smtp_accept_max_per_host was set, the number of connections could be
4015 restricted to fewer than expected, because the daemon was trying to set up
4016 a new connection before checking whether the processes handling previous
4017 connections had finished. The check for completed processes is now done
4018 earlier. On busy systems, this bug wouldn't be noticed because something
4019 else would have woken the daemon, and it would have reaped the completed
4022 80. If a message was submitted locally by a user whose login name contained one
4023 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4024 It caused a spool format error. I have fixed the spool reading code. A
4025 related problem was that the "from" clause in the Received: line became
4026 illegal because of the space(s). It is now covered by ${quote_local_part.
4028 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4031 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4033 83. Patch from Sieve maintainer to fix unterminated string problem in
4034 "vacation" handling.
4036 84. Some minor changes to the Linux configuration files to help with other
4037 OS variants using glibc.
4039 85. One more patch for Sieve to update vacation handling to latest spec.
4042 ----------------------------------------------------
4043 See the note above about the 4.44 and 4.50 releases.
4044 ----------------------------------------------------
4050 1. Change 4.43/35 introduced a bug that caused file counts to be
4051 incorrectly computed when quota_filecount was set in an appendfile
4054 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4055 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4058 3. Give more explanation in the error message when the command for a transport
4059 filter fails to execute.
4061 4. There are several places where Exim runs a non-Exim command in a
4062 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4063 was being done only for the command run by the queryprogram router. It is
4064 now done for all such subprocesses. The other cases are: ${run, transport
4065 filters, and the commands run by the lmtp and pipe transports.
4067 5. Some older OS have a limit of 256 on the maximum number of file
4068 descriptors. Exim was using setrlimit() to set 1000 as a large value
4069 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4070 systems. I've change it so that if it can't get 1000, it tries for 256.
4072 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4073 was an oversight, and furthermore, ever since the addition of extra
4074 controls (e.g. 4.43/32), the checks on when to allow different forms of
4075 "control" were broken. There should now be diagnostics for all cases when a
4076 control that does not make sense is encountered.
4078 7. $recipients is now available in the predata ACL (oversight).
4080 8. Tidy the search cache before the fork to do a delivery from a message
4081 received from the command line. Otherwise the child will trigger a lookup
4082 failure and thereby defer the delivery if it tries to use (for example) a
4083 cached ldap connection that the parent has called unbind on.
4085 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4086 of $address_data from the recipient verification was clobbered by the
4087 sender verification.
4089 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4090 was its contents. (It was OK if the option was not defined at all.)
4092 11. A "Completed" log line is now written for messages that are removed from
4093 the spool by the -Mrm option.
4095 12. $host_address is now set to the target address during the checking of
4096 ignore_target_hosts.
4098 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4099 being passed; this would have caused $sender_host_name to have been used if
4100 matching the list had actually called for a host name (not very likely,
4101 since this list is usually IP addresses). A host name is now passed as
4104 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4105 code that specifies a non-restarting handler (typically sigaction() in
4106 modern systems) in an attempt to fix a rare and obscure crash bug.
4108 15. Narrowed the window for a race in the daemon that could cause it to ignore
4109 SIGCHLD signals. This is not a major problem, because they are used only to
4110 wake it up if nothing else does.
4112 16. A malformed maildirsize file could cause Exim to calculate negative values
4113 for the mailbox size or file count. Odd effects could occur as a result.
4114 The maildirsize information is now recalculated if the size or filecount
4117 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4118 support for a long time. Removed HAVE_SYS_VFS_H.
4120 18. Updated exipick to current release from John Jetmore.
4122 19. Allow an empty sender to be matched against a lookup in an address list.
4123 Previously the only cases considered were a regular expression, or an
4126 20. Exim went into a mad DNS lookup loop when doing a callout where the
4127 host was specified on the transport, if the DNS lookup yielded more than
4130 21. The RFC2047 encoding function was originally intended for short strings
4131 such as real names; it was not keeping to the 75-character limit for
4132 encoded words that the RFC imposes. It now respects the limit, and
4133 generates multiple encoded words if necessary. To be on the safe side, I
4134 have increased the buffer size for the ${rfc2047: expansion operator from
4137 22. Failure to deliver a bounce message always caused it to be frozen, even if
4138 there was an errors_to setting on the router. The errors_to setting is now
4141 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4142 canonical form (fully expanded) before being placed in
4143 $sender_host_address.
4145 24. Updated eximstats to version 1.33
4147 25. Include certificate and key file names in error message when GnuTLS fails
4148 to set them up, because the GnuTLS error message doesn't include the name
4149 of the failing file when there is a problem reading it.
4151 26. Expand error message when OpenSSL has problems setting up cert/key files.
4154 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4155 (this can affect the format of dates).
4157 28. exim_tidydb, when checking for the continued existence of a message for
4158 which it has found a message-specific retry record, was not finding
4159 messages that were in split spool directories. Consequently, it was
4160 deleting retry records that should have stayed in existence.
4162 29. eximstats updated to version 1.35
4163 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4164 1.35 - bugfix such that pie charts by volume are generated correctly
4166 30. The SPA authentication driver was not abandoning authentication and moving
4167 on to the next authenticator when an expansion was forced to fail,
4168 contradicting the general specification for all authenticators. Instead it
4169 was generating a temporary error. It now behaves as specified.
4171 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4172 (the order specifies the preference for clients). The order is now AES256,
4173 AES128, 3DES, ARCFOUR128.
4175 31. Small patch to Sieve code - explicitly set From: when generating an
4178 32. Exim crashed if a remote delivery caused a very long error message to be
4179 recorded - for instance if somebody sent an entire SpamAssassin report back
4180 as a large number of 550 error lines. This bug was coincidentally fixed by
4181 increasing the size of one of Exim's internal buffers (big_buffer) that
4182 happened as part of the Exiscan merge. However, to be on the safe side, I
4183 have made the code more robust (and fixed the comments that describe what
4186 33. Some experimental protocols are using DNS PTR records for new purposes. The
4187 keys for these records are domain names, not reversed IP addresses. The
4188 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4189 leaves it alone. Component reversal etc. now happens only for IP addresses.
4192 34. The host_aton() function is supposed to be passed a string that is known
4193 to be a valid IP address. However, in the case of IPv6 addresses, it was
4194 not checking this. This is a hostage to fortune. Exim now panics and dies
4195 if the condition is not met. A case was found where this could be provoked
4196 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4197 components; fortuitously, this particular loophole had already been fixed
4198 by change 4.50/55 or 4.44/33 above.
4200 If there are any other similar loopholes, the new check in host_aton()
4201 itself should stop them being exploited. The report I received stated that
4202 data on the command line could provoke the exploit when Exim was running as
4203 exim, but did not say which command line option was involved. All I could
4204 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4205 running as the user.
4208 35. There was a buffer overflow vulnerability in the SPA authentication code
4209 (which came originally from the Samba project). I have added a test to the
4210 spa_base64_to_bits() function which I hope fixes it.
4213 36. The daemon start-up calls getloadavg() while still root for those OS that
4214 need the first call to be done as root, but it missed one case: when
4215 deliver_queue_load_max is set with deliver_drop_privilege. This is
4216 necessary for the benefit of the queue runner, because there is no re-exec
4217 when deliver_drop_privilege is set.
4219 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4220 was in use, was not putting the data itself into the right store pool;
4221 consequently, it could be overwritten for a subsequent message in the same
4222 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4225 38. Sometimes the final signoff response after QUIT could fail to get
4226 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4227 < 0 before doing a fflush(). This bug looks as though it goes back to the
4228 introduction of TLS in release 3.20, but "sometimes" must have been rare
4229 because the tests only now provoked it.
4235 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4236 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4237 Now it returns an integer. A number of calls were still expecting a T/F
4238 return. Fortuitously, in all cases, the tests worked in OK situations,
4239 which is the norm. However, things would have gone wrong on any write
4240 failures on the smtp file descriptor. This function is used when sending
4241 messages over SMTP and also when doing verify callouts.
4243 2. When Exim is called to do synchronous delivery of a locally submitted
4244 message (the -odf or -odi options), it no longer closes stderr before doing
4247 3. Implemented the mua_wrapper option.
4249 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4251 5. Implemented the functions header_remove(), header_testname(),
4252 header_add_at_position(), and receive_remove_recipient(), and exported them
4255 6. If an ACL "warn" statement specified the addition of headers, Exim already
4256 inserted X-ACL-Warn: at the start if there was no header name. However, it
4257 was not making this test for the second and subsequent header lines if
4258 there were newlines in the string. This meant that an invalid header could
4259 be inserted if Exim was badly configured.
4261 7. Allow an ACL "warn" statement to add header lines at the start or after all
4262 the Received: headers, as well as at the end.
4264 8. Added the rcpt_4xx retry error code.
4266 9. Added postmaster_mailfrom=xxx to callout verification option.
4268 10. Added mailfrom=xxxx to the callout verification option, for verify=
4271 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4272 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4274 12. Inserted some casts to stop certain compilers warning when using pointer
4275 differences as field lengths or precisions in printf-type calls (mostly
4276 affecting debugging statements).
4278 13. Added optional readline() support for -be (dynamically loaded).
4280 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4281 same clock tick as a message's arrival, so that its received time was the
4282 same as the "first fail" time on the retry record, and that message
4283 remained on the queue past the ultimate address timeout, every queue runner
4284 would try a delivery (because it was past the ultimate address timeout) but
4285 after another failure, the ultimate address timeout, which should have then
4286 bounced the address, did not kick in. This was a "< instead of <=" error;
4287 in most cases the first failure would have been in the next clock tick
4288 after the received time, and all would be well.
4290 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4291 being recognized when the domain list was tested by the match_domain
4292 condition in an expansion string.
4294 16. Added the ${str2b64: operator.
4296 17. Exim was always calling setrlimit() to set a large limit for the number of
4297 processes, without checking whether the existing limit was already
4298 adequate. (It did check for the limit on file descriptors.) Furthermore,
4299 errors from getrlimit() and setrlimit() were being ignored. Now they are
4300 logged to the main and panic logs, but Exim does carry on, to try to do its
4301 job under whatever limits there are.
4303 18. Imported PCRE 5.0.
4305 19. Trivial typo in log message " temporarily refused connection" (the leading
4308 20. If the log selector return_path_on_delivery was set and an address was
4309 redirected to /dev/null, the delivery process crashed because it assumed
4310 that a return path would always be set for a "successful" delivery. In this
4311 case, the whole delivery is bypassed as an optimization, and therefore no
4314 21. Internal re-arrangement: the function for sending a challenge and reading
4315 a response while authentication was assuming a zero-terminated challenge
4316 string. It's now changed to take a pointer and a length, to allow for
4317 binary data in such strings.
4319 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4321 23. Exim was not respecting finduser_retries when seeking the login of the
4322 uid under which it was called; it was always trying 10 times. (The default
4323 setting of finduser_retries is zero.) Also, it was sleeping after the final
4324 failure, which is pointless.
4326 24. Implemented tls_on_connect_ports.
4328 25. Implemented acl_smtp_predata.
4330 26. If the domain in control=submission is set empty, Exim assumes that the
4331 authenticated id is a complete email address when it generates From: or
4332 Sender: header lines.
4334 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4335 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4336 chown and chgrp in /bin and hostname in /usr/bin.
4338 28. Exim was keeping the "process log" file open after each use, just as it
4339 does for the main log. This opens the possibility of it remaining open for
4340 long periods when the USR1 signal hits a daemon. Occasional processlog
4341 errors were reported, that could have been caused by this. Anyway, it seems
4342 much more sensible not to leave this file open at all, so that is what now
4345 29. The long-running daemon process does not normally write to the log once it
4346 has entered its main loop, and it closes the log before doing so. This is
4347 so that log files can straightforwardly be renamed and moved. However,
4348 there are a couple of unusual error situations where the daemon does write
4349 log entries, and I had neglected to close the log afterwards.
4351 30. The text of an SMTP error response that was received during a remote
4352 delivery was being truncated at 512 bytes. This is too short for some of
4353 the long messages that one sometimes sees. I've increased the limit to
4356 31. It is now possible to make retry rules that apply only when a message has a
4357 specific sender, in particular, an empty sender.
4359 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4360 it possible to be selective about when SMTP synchronization is enforced.
4362 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4364 32. Implemented hosts_connection_nolog.
4366 33. Added an ACL for QUIT.
4368 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4371 35. Added mailbox_size and mailbox_filecount to appendfile.
4373 36. Added control = no_multiline_responses to ACLs.
4375 37. There was a bug in the logic of the code that waits for the clock to tick
4376 in the case where the clock went backwards by a substantial amount such
4377 that the microsecond fraction of "now" was more than the microsecond
4378 fraction of "then" (but the whole seconds number was less).
4380 38. Added support for the libradius Radius client library this is found on
4381 FreeBSD (previously only the radiusclient library was supported).
4387 1. When certain lookups returned multiple values in the form name=value, the
4388 quoting of the values was not always being done properly. Specifically:
4389 (a) If the value started with a double quote, but contained no whitespace,
4391 (b) If the value contained whitespace other than a space character (i.e.
4392 tabs or newlines or carriage returns) it was not quoted.
4393 This fix has been applied to the mysql and pgsql lookups by writing a
4394 separate quoting function and calling it from the lookup code. The fix
4395 should probably also be applied to nisplus, ibase and oracle lookups, but
4396 since I cannot test any of those, I have not disturbed their existing code.
4398 2. A hit in the callout cache for a specific address caused a log line with no
4399 reason for rejecting RCPT. Now it says "Previous (cached) callout
4400 verification failure".
4402 3. There was an off-by-one bug in the queryprogram router. An over-long
4403 return line was truncated at 256 instead of 255 characters, thereby
4404 overflowing its buffer with the terminating zero. As well as fixing this, I
4405 have increased the buffer size to 1024 (and made a note to document this).
4407 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4408 when Exim is waiting for an SMTP response from a remote server, Exim
4409 restarts its select() call on the socket, thereby resetting its timeout.
4410 This is not a problem when such interrupts are rare. Somebody set up a cron
4411 job to run exiwhat every 2 minutes, which is less than the normal select()
4412 timeout (5 or 10 minutes). This meant that the select() timeout never
4413 kicked in because it was always reset. I have fixed this by comparing the
4414 time when an interrupt arrives with the time at the start of the first call
4415 to select(). If more time than the timeout has elapsed, the interrupt is
4416 treated as a timeout.
4418 5. Some internal re-factoring in preparation for the addition of Sieve
4419 extensions (by MH). In particular, the "personal" test is moved to a
4420 separate function, and given an option for scanning Cc: and Bcc: (which is
4421 not set for Exim filters).
4423 6. When Exim created an email address using the login of the caller as the
4424 local part (e.g. when creating a From: or Sender: header line), it was not
4425 quoting the local part when it contained special characters such as @.
4427 7. Installed new OpenBSD configuration files.
4429 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4430 try to make them clearer.
4432 9. Callout options, other than the timeout value, were being ignored when
4433 verifying sender addresses in header lines. For example, when using
4435 verify = header_sender/callout=no_cache
4437 the cache was (incorrectly) being used.
4439 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4440 only the Cygwin environment.
4442 11. When return_path_on_delivery was set as a log selector, if different remote
4443 addresses in the same message used different return paths and parallel
4444 remote delivery occurred, the wrong values would sometimes be logged.
4445 (Whenever a remote delivery process finished, the return path value from
4446 the most recently started remote delivery process was logged.)
4448 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4449 lines when AUTH and/or TLS are in use. This is the "received protocol"
4450 field. Exim used to use "asmtp" for authenticated SMTP, without any
4451 indication (in the protocol name) for TLS use. Now it follows the RFC and
4452 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4453 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4454 names appear in log lines as well as in Received: header lines.
4456 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4457 extensions, and comparison tests, and to fix some bugs.
4459 14. Changes to the "personal" filter test:
4461 (1) The test was buggy in that it was just doing the equivalent of
4462 "contains" tests on header lines. For example, if a user's address was
4463 anne@some.where, the "personal" test would incorrectly be true for
4465 To: susanne@some.where
4467 This test is now done by extracting each address from the header in turn,
4468 and checking the entire address. Other tests that are part of "personal"
4469 are now done using regular expressions (for example, to check local parts
4470 of addresses in From: header lines).
4472 (2) The list of non-personal local parts in From: addresses has been
4473 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4474 taken from the Sieve specification recommendations.
4476 (3) If the message contains any header line starting with "List-" it is
4477 treated as non-personal.
4479 (4) The test for "circular" in the Subject: header line has been removed
4480 because it now seems ill-conceived.
4482 15. Minor typos in src/EDITME comments corrected.
4484 16. Installed latest exipick from John Jetmore.
4486 17. If headers_add on a router specified a text string that was too long for
4487 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4488 of string_sprintf() is now avoided.
4490 18. $message_body_size was not set (it was always zero) when running the DATA
4491 ACL and the local_scan() function.
4493 19. For the "mail" command in an Exim filter, no default was being set for
4494 the once_repeat time, causing a random time value to be used if "once" was
4495 specified. (If the value happened to be <= 0, no repeat happened.) The
4496 default is now 0s, meaning "never repeat". The "vacation" command was OK
4497 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4498 (I found it when inspecting the code).
4500 20. There is now an overall timeout for performing a callout verification. It
4501 defaults to 4 times the callout timeout, which applies to individual SMTP
4502 commands during the callout. The overall timeout applies when there is more
4503 than one host that can be tried. The timeout is checked before trying the
4504 next host. This prevents very long delays if there are a large number of
4505 hosts and all are timing out (e.g. when the network connections are timing
4506 out). The value of the overall timeout can be changed by specifying an
4507 additional sub-option for "callout", called "maxwait". For example:
4509 verify = sender/callout=5s,maxwait=20s
4511 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4512 to the end before writing, but this should make it even safer).
4514 22. Exim was forgetting that it had advertised PIPELINING for the second and
4515 subsequent messages on an SMTP connection. It was also not resetting its
4516 memory on STARTTLS and an internal HELO.
4518 23. When Exim logs an SMTP synchronization error within a session, it now
4519 records whether PIPELINING has been advertised or not.
4521 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4522 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4523 rather than long int.
4525 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4527 26. Added the never_mail option to autoreply.
4533 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4534 crash if the getsockname() call failed; this can happen if a connection is
4535 closed very soon after it is established. The problem was simply in the
4536 order in which certain operations were done, causing Exim to try to write
4537 to the SMTP stream before it had set up the file descriptor. The bug has
4538 been fixed by making things happen in the correct order.
4544 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4545 before the connection was closed, thus losing the rejection response.
4547 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4548 some early Solaris releases, but causes trouble in current releases where
4549 socklen_t is defined.
4551 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4554 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4557 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4558 that releasing the top of it at the end releases what was used for sub-
4559 expansions (unless the block got too big). However, discard this block if
4560 the first thing is a variable or header, so that we can use its block when
4561 it is dynamic (useful for very large $message_headers, for example).
4563 6. Lookups now cache *every* query, not just the most recent. A new, separate
4564 store pool is used for this. It can be recovered when all lookup caches are
4565 flushed. Lookups now release memory at the end of their result strings.
4566 This has involved some general refactoring of the lookup sources.
4568 7. Some code has been added to the store_xxx() functions to reduce the amount
4569 of flapping under certain conditions.
4571 8. log_incoming_interface used to affect only the <= reception log lines. Now
4572 it causes the local interface and port to be added to several more SMTP log
4573 lines, for example "SMTP connection from", and rejection lines.
4575 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4577 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4579 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4580 could be overwritten at the end of the current message (or the start of a
4581 new message if it was set in a HELO ACL). The value is now preserved for
4582 the duration of the SMTP connection.
4584 12. If a transport had a headers_rewrite setting, and a matching header line
4585 contained an unqualified address, that address was qualified, even if it
4586 did not match any rewriting rules. The underlying bug was that the values
4587 of the flags that permit the existence of unqualified sender and recipient
4588 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4589 non-local messages, and by -bnq for local messages) were not being
4590 preserved with the message after it was received.
4592 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4593 "next input=" as part of the text comprising the host identity instead of
4594 the correct text. The code was using the same buffer for two different
4595 strings. However, depending on which order the printing function evaluated
4596 its arguments, the bug did not always show up. Under Linux, for example, my
4597 test suite worked just fine.
4599 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4600 doesn't work with some older versions of Perl. It has been changed to "my",
4601 which in any case is probably the better facility to use.
4603 15. A really picky compiler found some instances of statements for creating
4604 error messages that either had too many or two few arguments for the format
4607 16. The size of the buffer for calls to the DNS resolver has been increased
4608 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4609 for addresses that have a lot of PTR records. This alleviates a problem; it
4610 does not fully solve it.
4612 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4613 buffer now truncates the list and logs the incident, which is the same
4614 action as happens when Exim is looking up a host name and its aliases.
4615 Previously in this situation something unpredictable would happen;
4616 sometimes it was "internal error: store_reset failed".
4618 18. If a server dropped the connection unexpectedly when an Exim client was
4619 using GnuTLS and trying to read a response, the client delivery process
4620 crashed while trying to generate an error log message.
4622 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4623 string, for example:
4625 warn message = H1: something\nH2: something
4627 the text was added as a single header line from Exim's point of view
4628 though it ended up OK in the delivered message. However, searching for the
4629 second and subsequent header lines using $h_h2: did not work. This has been
4630 fixed. Similarly, if a system filter added multiple headers in this way,
4631 the routers could not see them.
4633 20. Expanded the error message when iplsearch is called with an invalid key to
4634 suggest using net-iplsearch in a host list.
4636 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4637 ACLs are no longer actually imposed (and a message to that effect is
4640 22. If a "gecos" field in a passwd entry contained escaped characters, in
4641 particular, if it contained a \" sequence, Exim got it wrong when building
4642 a From: or a Sender: header from that name. A second bug also caused
4643 incorrect handling when an unquoted " was present following a character
4644 that needed quoting.
4646 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4647 was not being matched caselessly.
4649 24. Arranged for all hyphens in the exim.8 source to be escaped with
4652 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4653 properly. Recipient callout cache records were still being keyed to include
4654 the sender, even when use_sender was set false. This led to far more
4655 callouts that were necessary. The sender is no longer included in the key
4656 when use_sender is false.
4658 26. Added "control = submission" modifier to ACLs.
4660 27. Added the ${base62d: operator to decode base 62 numbers.
4662 28. dnsdb lookups can now access SRV records.
4664 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4665 the configuration file.
4667 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4668 (-v) mode. This makes the output for a verbose queue run more intelligible.
4670 31. Added a use_postmaster feature to recipient callouts.
4672 32. Added the $body_zerocount variable, containing the number of binary zero
4673 bytes in the message body.
4675 33. The time of last modification of the "new" subdirectory is now used as the
4676 "mailbox time last read" when there is a quota error for a maildir
4679 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4681 35. Added +ignore_unknown as a special item in host lists.
4683 36. Code for decoding IPv6 addresses in host lists is now included, even if
4684 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4685 address was recognized as an IP address, but was then not correctly decoded
4686 into binary, causing unexpected and incorrect effects when compared with
4693 1. Very minor rewording of debugging text in manualroute to say "list of
4694 hosts" instead of "hostlist".
4696 2. If verify=header_syntax was set, and a header line with an unqualified
4697 address (no domain) and a large number of spaces between the end of the
4698 name and the colon was received, the reception process suffered a buffer
4699 overflow, and (when I tested it) crashed. This was caused by some obsolete
4700 code that should have been removed. The fix is to remove it!
4702 3. When running in the test harness, delay a bit after writing a bounce
4703 message to get a bit more predictability in the log output.
4705 4. Added a call to search_tidyup() just before forking a reception process. In
4706 theory, someone could use a lookup in the expansion of smtp_accept_max_
4707 per_host which, without the tidyup, could leave open a database connection.
4709 5. Added the variables $recipient_data and $sender_data which get set from a
4710 lookup success in an ACL "recipients" or "senders" condition, or a router
4711 "senders" option, similar to $domain_data and $local_part_data.
4713 6. Moved the writing of debug_print from before to after the "senders" test
4716 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4717 problems for message scanning, either using a data ACL, or using
4718 local_scan() because the Received: header was not generated till after they
4719 were called (in order to set the time as the time of reception completion).
4720 I have revised the way this works. The header is now generated after the
4721 body is received, but before the ACL or local_scan() are called. After they
4722 are run, the timestamp in the header is updated.
4728 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4729 before starting a queue runner without re-exec. This happened only when
4730 deliver_drop_privilege was set or when the Exim user was set to root. The
4731 effect of the bug was that timeouts during subsequent deliveries caused
4732 crashes instead of being properly handled. The handler is now left at its
4733 default (and expected) setting.
4735 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4736 message, again when deliver_drop_privilege is set or Exim is run as root.
4737 The bug described in (1) was not present in this case, but the tidying up
4738 of the other signals was missing. I have made the two cases consistent.
4740 3. The ignore_target_hosts setting on a manualroute router was being ignored
4741 for hosts that were looked up using the /MX notation.
4743 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4746 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4747 operated on the sender address. After changing the $sender_address to <>
4748 for the sender address verify, Exim was re-instated it as the original
4749 (before rewriting) address, but remembering that it had rewritten it, so it
4750 wasn't rewriting it again. This bug also had the effect of breaking the
4751 sender address verification caching when the sender address was rewritten.
4753 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4754 This has been changed so that if the ip literal address matches
4755 ignore_target_hosts, the router declines.
4757 7. Added expansion conditions match_domain, match_address, and match_local_
4758 part (NOT match_host).
4760 8. The placeholder for the Received: header didn't have a length field set.
4762 9. Added code to Exim itself and to exim_lock to test for a specific race
4763 condition that could lead to file corruption when using MBX delivery. The
4764 issue is with the lockfile that is created in /tmp. If this file is removed
4765 after a process has opened it but before that process has acquired a lock,
4766 there is the potential for a second process to recreate the file and also
4767 acquire a lock. This could lead to two Exim processes writing to the file
4768 at the same time. The added code performs the same test as UW imapd; it
4769 checks after acquiring the lock that its file descriptor still refers to
4770 the same named file.
4772 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4773 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4776 11. Added the smtp_active_hostname option. If used, this will typically be made
4777 to depend on the incoming interface address. Because $interface_address is
4778 not set up until the daemon has forked a reception process, error responses
4779 that can happen earlier (such as "too many connections") no longer contain
4782 12. If an expansion in a condition on a "warn" statement fails because a lookup
4783 defers, the "warn" statement is abandoned, and the next ACL statement is
4784 processed. Previously this caused the whole ACL to be aborted.
4786 13. Added the iplsearch lookup type.
4788 14. Added ident_timeout as a log selector.
4790 15. Added tls_certificate_verified as a log selector.
4792 16. Added a global option tls_require_ciphers (compare the smtp transport
4793 option of the same name). This controls incoming TLS connections.
4795 17. I finally figured out how to make tls_require_ciphers do a similar thing
4796 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4797 before starting the TLS session.
4799 18. Tabs are now shown as \t in -bP output.
4801 19. If the log selector return_path_on_delivery was set, Exim crashed when
4802 bouncing a message because it had too many Received: header lines.
4804 20. If two routers both had headers_remove settings, and the first one included
4805 a superfluous trailing colon, the final name in the first list and the
4806 first name in the second list were incorrectly joined into one item (with a
4807 colon in the middle).
4813 1. Added -C and -D options to the exinext utility, mainly to make it easier
4814 to include in the automated testing, but these could be helpful when
4815 multiple configurations are in use.
4817 2. The exinext utility was not formatting the output nicely when there was
4818 an alternate port involved in the retry record key, nor when there was a
4819 message id as well (for retries that were specific to a specific message
4820 and a specific host). It was also confused by IPv6 addresses, because of
4821 the additional colons they contain. I have fixed the IPv4 problem, and
4822 patched it up to do a reasonable job for IPv6.
4824 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4825 delivery, the log line now contains "pipelined" if PIPELINING was used.
4827 4. An SMTP transport process used to panic and die if the bind() call to set
4828 an explicit outgoing interface failed. This has been changed; it is now
4829 treated in the same way as a connect() failure.
4831 5. A reference to $sender_host_name in the part of a conditional expansion
4832 that was being skipped was still causing a DNS lookup. This no longer
4835 6. The def: expansion condition was not recognizing references to header lines
4836 that used bh_ and bheader_.
4838 7. Added the _cache feature to named lists.
4840 8. The code for checking quota_filecount in the appendfile transport was
4841 allowing one more file than it should have been.
4843 9. For compatibility with Sendmail, the command line option
4851 and sets the incoming protocol and host name (for trusted callers). The
4852 host name and its colon can be omitted when only the protocol is to be set.
4853 Note the Exim already has two private options, -pd and -ps, that refer to
4854 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4855 "s", but I don't think that's a major issue.
4857 10. A number of refactoring changes to the code, none of which should affect
4860 (a) The number of logging options was getting close to filling up the
4861 32-bit word that was used as a bit map. I have split them into two classes:
4862 those that are passed in the argument to log_write(), and those that are
4863 only ever tested independently outside of that function. These are now in
4864 separate 32-bit words, so there is plenty of room for expansion again.
4865 There is no change in the user interface or the logging behaviour.
4867 (b) When building, for example, log lines, the code previously used a
4868 macro that called string_cat() twice, in order to add two strings. This is
4869 not really sufficiently general. Furthermore, there was one instance where
4870 it was actually wrong because one of the argument was used twice, and in
4871 one call a function was used. (As it happened, calling the function twice
4872 did not affect the overall behaviour.) The macro has been replaced by a
4873 function that can join an arbitrary number of extra strings onto a growing
4876 (c) The code for expansion conditions now uses a table and a binary chop
4877 instead of a serial search (which was left over from when there were very
4878 few conditions). Also, it now recognizes conditions like "pam" even when
4879 the relevant support is not compiled in: a suitably worded error message is
4880 given if an attempt is made to use such a condition.
4882 11. Added ${time_interval:xxxxx}.
4884 12. A bug was causing one of the ddress fields not to be passed back correctly
4885 from remote delivery subprocesses. The field in question was not being
4886 subsequently used, so this caused to problems in practice.
4888 13. Added new log selectors queue_time and deliver_time.
4890 14. Might have fixed a bug in maildirsizefile handling that threw up
4891 "unexpected character" debug warnings, and recalculated the data
4892 unnecessarily. In any case, I expanded the warning message to give more
4895 15. Added the message "Restricted characters in address" to the statements in
4896 the default ACL that block characters like @ and % in local parts.
4898 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4899 Three changes have been made:
4901 (a) There was a serious bug; a negative response to MAIL caused the whole
4902 recipient domain to be cached as invalid, thereby blocking all messages
4903 to all local parts at the same domain, from all senders. This bug has
4904 been fixed. The domain is no longer cached after a negative response to
4905 MAIL if the sender used is not empty.
4907 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4910 (c) A new callout option, "use_sender" has been added for people who want
4911 the modified behaviour.
4917 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4920 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4921 indeed breaks things for older releases.
4923 3. Added additional logging to the case where there is a problem reading data
4924 from a filter that is running in a subprocess using a pipe, in order to
4925 try to track down a specific problem.
4927 4. Testing facility fudge: when running in the test harness and attempting
4928 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4929 getting "No route to host". Convert this to a timeout.
4931 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4934 6. Some OS don't have socklen_t but use size_t instead. This affects the
4935 fifth argument of getsockopt() amongst other things. This is now
4936 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4937 can be set for individual OS. I have set it for SunOS5, OSF1, and
4938 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4939 some earlier ones do not.
4941 7. Change 4.30/15 was not doing the test caselessly.
4943 8. The standard form for an IPv6 address literal was being rejected by address
4944 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4945 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4946 this, as well as the form without the "IPv6" on the front (but only when
4947 address literals are enabled, of course).
4949 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4951 10. Exim crashed if a message with an empty sender address specified by -f
4952 encountered a router with an errors_to setting. This could be provoked only
4953 by a command such as
4957 where an empty string was supplied; "<>" did not hit this bug.
4959 11. Installed PCRE release 4.5.
4961 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4962 remained set. It is now erased.
4964 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4965 times from message ids (which are base 36 rather than the normal 62).
4967 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4968 were being counted as actual protocol errors, and logged if the log
4969 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4970 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4971 and DATA following a set of rejected RCPTs do not count as protocol errors.
4972 In other words, Exim assumes they were pipelined, though this may not
4973 actually be the case. Of course, in all cases the client gets an
4974 appropriate error code.
4976 15. If a lookup fails in an ACL condition, a message about the failure may
4977 be available; it is used if testing the ACL cannot continue, because most
4978 such messages specify what the cause of the deferral is. However, some
4979 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4980 that caused an old message to be retained and used if a later statement
4981 caused a defer, replacing the real cause of the deferral.
4983 16. If an IP address had so many PTR records that the DNS lookup buffer
4984 was not large enough to hold them, Exim could crash while trying to process
4985 the truncated data. It now detects and logs this case.
4987 17. Further to 4.21/58, another change has been made: if (and only if) the
4988 first line of a message (the first header line) ends with CRLF, a bare LF
4989 in a subsequent header line has a space inserted after it, so as not to
4990 terminate the header.
4992 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4993 unnecessarily, used atoi() instead of strtol(), and didn't check the
4994 termination when getting file sizes from file names by regex.
4996 19. Completely re-implemented the support for maildirsize files, in the light
4997 of a number of problems with the previous contributed implementation
4998 (4.30/29). In particular:
5000 . If the quota is zero, the maildirsize file is maintained, but no quota is
5003 . If the maildir directory does not exist, it is created before any attempt
5004 to write a maildirsize file.
5006 . The quota value in the file is just a cache; if the quota is changed in
5007 the transport, the new value overrides.
5009 . A regular expression is available for excluding directories from the
5012 20. The autoreply transport checks the characters in options that define the
5013 message's headers; it allows continued headers, but it was checking with
5014 isspace() after an embedded newline instead of explicitly looking for a
5017 21. If all the "regular" hosts to which an address was routed had passed their
5018 expiry times, and had not reached their retry times, the address was
5019 bounced, even if fallback hosts were defined. Now Exim should go on to try
5022 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5023 equivalent code in the SMTP transport. Some hosts send humungous responses
5024 to HELO/EHLO, more than 1024 it seems.
5026 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5027 gives compiler warnings in some environments. I've now done it "properly",
5030 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5031 (because of problems with the built-in one) was declared to return uschar *
5032 instead of char *, causing compiler failure.
5034 25. Fixed a file descriptor leak when processing alias/forward files.
5036 26. Fixed a minor format string issue in dbfn.c.
5038 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5040 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5041 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5044 29. When a sender address is verified, it is cached, to save repeating the test
5045 when there is more than one recipient in a message. However, when the
5046 verification involves a callout, it is possible for different callout
5047 options to be set for different recipients. It is too complicated to keep
5048 track of this in the cache, so now Exim always runs a verification when a
5049 callout is required, relying on the callout cache for the optimization.
5050 The overhead is duplication of the address routing, but this should not be
5053 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5054 to be verified with callout=postmaster, and the main callout worked but the
5055 postmaster check failed, the verification correctly failed. However, if a
5056 subsequent RCPT command asked for sender verification *without* the
5057 postmaster check, incorrect caching caused this verification also to fail,
5060 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5061 it was not caching the DNS options (qualify_single, search_parents) that
5062 were used when the lookup failed. A subsequent lookup with different
5063 options therefore always gave the same answer, though there were cases
5064 where it should not have. (Example: a "domains = !$mx_any" option on a
5065 dnslookup router: the "domains" option is always processed without any
5066 widening, but the router might have qualify_single set.) Now Exim uses the
5067 cached value only when the same options are set.
5069 32. Added John Jetmore's "exipick" utility to the distribution.
5071 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5072 than a timeout (e.g. a certificate is required, and is not provided), an
5073 Exim server now closes the connection immediately. Previously it waited for
5074 the client to close - but if the client is SSL, it seems that they each
5075 wait for each other, leading to a delay before one of them times out.
5077 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5078 maintained 0.8.x compatibility because I don't think many are using it, and
5079 it is clearly obsolete.
5081 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5084 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5085 client certificate was expired. A simple patch fixes this, though I don't
5086 understand the full logic of why the verify callback is called multiple
5089 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5090 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5091 which causes problems with some clients (such as the Certicom SSL Plus
5092 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5093 disables the coutermeasure allowing Eudora to connect."
5095 38. Exim was not checking that a write() to a log file succeeded. This could
5096 lead to Bad Things if a log got too big, in particular if it hit a file
5097 size limit. Exim now panics and dies if it cannot write to a log file, just
5098 as it does if it cannot open a log file.
5100 39. Modified OS/Makefile-Linux so that it now contains
5102 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5104 The two -D definitions ensure that Exim is compiled with large file
5105 support, which makes it possible to handle log files that are bigger than
5108 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5109 instance) a domain was checked against a named list that involved a lookup,
5110 causing $domain_data to be set, then another domain was checked against the
5111 same list, then the first domain was re-checked, the value of $domain_data
5112 after the final check could be wrong. In particular, if the second check
5113 failed, it could be set empty. This bug probably also applied to
5116 41. The strip_trailing_dot option was not being applied to the address given
5117 with the -f command-line option.
5119 42. The code for reading a message's header from the spool was incrementing
5120 $received_count, but never initializing it. This meant that the value was
5121 incorrect (doubled) while delivering a message in the same process in which
5122 it was received. In the most common configuration of Exim, this never
5123 happens - a fresh exec is done - but it can happen when
5124 deliver_drop_privilege is set.
5126 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5127 it now includes up to 150 characters of the unexpected data in the log
5130 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5131 and building data strings. The size of both of these buffers was 10 000
5132 bytes - far larger than anybody would *ever* want, thought I. Needless to
5133 say, somebody hit the limit. I have increased the maximum line length to
5134 20 000 and the maximum data length of concatenated lines to 100 000. I have
5135 also fixed two bugs, because there was no checking on these buffers. Tsk,
5136 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5137 buffer is too small.
5139 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5140 lsearch lookups. Now it does.
5142 46. When parsing a route_list item in a manualroute router, a fixed-length
5143 buffer was used for the list of hosts. I made this 1024 bytes long,
5144 thinking that nobody would ever have a list of hosts that long. Wrong.
5145 Somebody had a whole pile of complicated expansion conditions, and the
5146 string was silently truncated, leading to an expansion error. It turns out
5147 that it is easier to change to an unlimited length (owing to other changes
5148 that have happened since this code was originally written) than to build
5149 structure for giving a limitation error. The length of the item that
5150 expands into the list of hosts is now unlimited.
5152 47. The lsearch lookup could not handle data where the length of text line was
5153 more than 4095 characters. Such lines were truncated, leading to shortened
5154 data being returned. It should now handle lines of any length.
5156 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5157 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5160 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5161 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5162 no longer happens in comment lines. A list of replacements is now placed
5163 at the head of all of the source files, except those whose only change is
5164 to replace PERL_COMMAND in the very first #! line.
5166 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5167 messages on the queue, with a bottom-up merge sort, using code contributed
5168 by Michael Haardt. This should make operations like -bp somewhat faster on
5169 large queues. It won't affect queue runners, except when queue_run_in_order
5172 51. Installed eximstats 1.31 in the distribution.
5174 52. Added support for SRV lookups to the dnslookup router.
5176 53. If an ACL referred to $message_body or $message_body_end, the value was not
5177 reset for any messages that followed in the same SMTP session.
5179 54. The store-handling optimization for building very long strings was not
5180 differentiating between the different store pools. I don't think this
5181 actually made any difference in practice, but I've tidied it.
5183 55. While running the routers to verify a sender address, $sender_address
5184 was still set to the sender address. This is wrong, because when routing to
5185 send a bounce to the sender, it would be empty. Therefore, I have changed
5186 it so that, while verifying a sender address, $sender_address is set to <>.
5187 (There is no change to what happens when verifying a recipient address.)
5189 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5190 target A or AAAA records (if not already returned) without resetting the
5191 qualify_single or search_parents options of the DNS resolver. These are
5192 inappropriate in this case because the targets of MX and SRV records must
5193 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5194 target that, when qualified, matched something in the local domain. These
5195 two options are now turned off when doing these lookups.
5197 57. It seems that at least some releases of Reiserfs (which does not have the
5198 concept of a fixed number of inodes) returns zero and not -1 for the
5199 number of available inodes. This interacted badly with check_spool_inodes,
5200 which assumed that -1 was the "no such thing" setting. What I have done is
5201 to check that the total number of inodes is greater than zero before doing
5202 the test of how many are available.
5204 58. When a "warn" ACL statement has a log_message modifier, the message is
5205 remembered, and not repeated. This is to avoid a lot of repetition when a
5206 message has many recipients that cause the same warning to be written.
5207 Howewer, Exim was preserving the list of already written lines for an
5208 entire SMTP session, which doesn't seem right. The memory is now reset if a
5209 new message is started.
5211 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5212 debugging output unless log_rewrite was also set.
5214 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5215 of (int)(handle) when we know that handle contains (void *)(-1).
5217 61. The Exim daemon panic-logs an error return when it closes the incoming
5218 connection. However "connection reset by peer" seems to be common, and
5219 isn't really an error worthy of noting specially, so that particular error
5222 62. When Exim is trying to find all the local interfaces, it used to panic and
5223 die if the ioctl to get the interface flags failed. However, it seems that
5224 on at least one OS (Solaris 9) it is possible to have an interface that is
5225 included in the list of interfaces, but for which you get a failure error
5226 for this call. This happens when the interface is not "plumbed" into a
5227 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5228 failure of the "get flags" call assumes that the interface is down.
5230 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5231 makes life easier for people who are doing arithmetic on fields extracted
5232 from dates, where you often get leading zeros that should not be
5233 interpreted as octal.
5235 64. Added qualify_domain to the redirect router, to override the global
5238 65. If a pathologically long header line contained very many addresses (the
5239 report of this problem mentioned 10 000) and each of them was rewritten,
5240 Exim could use up a very large amount of memory. (It kept on making new
5241 copies of the header line as it rewrote, and never released the old ones.)
5242 At the expense of a bit more processing, the header rewriting function has
5243 been changed so that it no longer eats memory in this way.
5245 66. The generation of the Received: header has been moved from the time that a
5246 message starts to be received, to the time that it finishes. The timestamp
5247 in the Received: header should now be very close to that of the <= log
5248 line. There are two side-effects of this change:
5250 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5251 logged header lines no longer include the local Received: line, because
5252 it has not yet been created. The same applies to a copy of the message
5253 that is returned to a non-SMTP sender when a message is rejected.
5255 (b) When a filter file is tested using -bf, no additional Received: header
5256 is added to the test message. After some thought, I decided that this
5259 This change does not affect the value of $received_for. It is still set
5260 after address rewriting, but before local_scan() is called.
5262 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5264 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5265 gave an unhelpful panic error message, and a defer error. I have managed to
5266 change this behaviour so that it now rejects any supplied certificate,
5267 which seems right, as the list of acceptable certificates is empty.
5269 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5270 gave an unhelpful defer error. I have not managed to make this reject any
5271 supplied certificates, but the error message it gives is "no certificate
5272 supplied", which is not helpful.
5274 70. exigrep's output now also includes lines that are not associated with any
5275 message, but which match the given pattern. Implemented by a patch from
5276 Martin Sluka, which also tidied up the Perl a bit.
5278 71. Recipient callout verification, like sender verification, was using <> in
5279 the MAIL FROM command. This isn't really the right thing, since the actual
5280 sender may affect whether the remote host accepts the recipient or not. I
5281 have changed it to use the actual sender in the callout; this means that
5282 the cache record is now keyed on a recipient/sender pair, not just the
5283 recipient address. There doesn't seem to be a real danger of callout loops,
5284 since a callout by the remote host to check the sender would use <>.
5285 [SEE ABOVE: changed after hitting problems.]
5287 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5288 temporary errors. However, in the case of such a code being given after
5289 the end of a data transmission (i.e. after ".") Exim was failing to write
5290 a retry record for the message. (Yes, there was some broken host that was
5291 actually sending 8xx at this point.)
5293 73. An unknown lookup type in a host list could cause Exim to panic-die when
5294 the list was checked. (An example that provoked this was putting <; in the
5295 middle of a list instead of at the start.) If this happened during a DATA
5296 ACL check, a -D file could be left lying around. This kind of configuration
5297 error no longer causes Exim to die; instead it causes a defer errror. The
5298 incident is still logged to the main and panic logs.
5300 74. Buglet left over from Exim 3 conversion. The message "too many messages
5301 in one connection" was written to the rejectlog but not the mainlog, except
5302 when address rewriting (yes!) was being logged.
5304 75. Added write_rejectlog option.
5306 76. When a system filter was run not as root (that is, when system_filter_user
5307 was set), the values of the $n variables were not being returned to the
5308 main process; thus, they were not subsequently available in the $sn
5311 77. Added +return_path_on_delivery log selector.
5313 78. A connection timeout was being treated differently from recipients deferred
5314 when testing hosts_max_try with a message that was older than the host's
5315 retry timeout. (The host should not be counted, thus allowing all hosts to
5316 be tried at least once before bouncing.) This may have been the cause of an
5317 occasionally reported bug whereby a message would remain on the queue
5318 longer than the retry timeout, but would be bounced if a delivery was
5319 forced. I say "may" because I never totally pinned down the problem;
5320 setting up timeout/retry tests is difficult. See also the next item.
5322 79. The ultimate address timeout was not being applied to errors that involved
5323 a combination of host plus message (for example, a timeout on a MAIL
5324 command). When an address resolved to a number of possible hosts, and they
5325 were not all tried for each delivery (e.g. because of hosts_max_try), a
5326 message could remain on the queue longer than the retry timeout.
5328 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5331 81. Fixed an obscure SMTP outgoing bug which required at least the following
5332 conditions: (a) there was another message waiting for the same server;
5333 (b) the server returned 5xx to all RCPT commands in the first message so
5334 that the message was not completed; (c) the server dropped the connection
5335 or gave a negative response to the RSET that Exim sends to abort the
5336 transaction. The observed case was a dropped connection after DATA that had
5337 been sent in pipelining mode. That is, the server had advertised PIPELINING
5338 but was not implementing it correctly. The effect of the bug was incorrect
5339 behaviour, such as trying another host, and this could lead to a crash.
5345 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5346 and daemon.c were passed as pointers to ints; they should have been
5347 pointers to socklen_t variables (which are typically unsigned ints).
5349 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5352 3. Fixed a really odd bug that affected only the testing scheme; patching a
5353 certain fixed string in the binary changed the value of another string that
5354 happened to be identical to the end of the original first string.
5356 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5357 name", it returns that address as the IP address. On some operating
5358 systems (e.g. Solaris), it also passes back the IP address string as the
5359 "host name". However, on others (e.g. Linux), it passes back an empty
5360 string. Exim wasn't checking for this, and was changing the host name to an
5361 empty string, assuming it had been canonicized.
5363 5. Although rare, it is permitted to have more than one PTR record for a given
5364 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5365 all the names associated with an address, because they do in Solaris.
5366 However, it seems that they do not in Linux for data that comes from the
5367 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5368 I found this out when I moved to a new Linux workstation and tried to run
5369 the Exim test suite.
5371 To get round this problem I have changed the code so that it now does its
5372 own call to the DNS to look up PTR records when searching for a host name.
5373 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5374 addresses that are only in /etc/hosts are still found.
5376 This behaviour is, however, controlled by an option called host_lookup_
5377 order, which defaults to "bydns:byaddr". If people want to use the other
5378 order, or indeed, just use one or the other means of lookup, they can
5379 specify it in this variable.
5381 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5382 some operating systems, this comes back from gethostbyaddr() as an empty
5383 string, and this is what Exim used to test for. However, it seems that in
5384 other systems, "." is yielded. Exim now tests for this case too.
5386 7. The values of check_spool_space and check_log_space are now held internally
5387 as a number of kilobytes instead of an absolute number of bytes. If a
5388 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5389 kilobyte. This means that much larger values can be stored.
5391 8. Exim monitor: an attempt to get the action menu when not actually pointing
5392 at a message produces an empty menu entitled "No message selected". This
5393 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5394 no entries in it ("Shell widget menu has zero width and/or height"). So I
5395 have added a single, blank menu entry in this case.
5397 9. Added ${quote_local_part.
5399 10. MIME decoding is now applied to the contents of Subject: header lines when
5402 11. Now that a reference to $sender_host_address automatically causes a reverse
5403 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5404 host lookup before query-style lookups in lists that might use this
5405 variable. This has therefore been abolished, and the "net-" prefix is no
5406 longer necessary for query-style lookups.
5408 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5409 have been a typo for LFLAGS, so it has been changed.
5411 13. The install script calls Exim with "-C /dev/null" in order to find the
5412 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5413 to be output. Howeve, since Exim outputs its version number before the
5414 error, it didn't break the script. It just looked ugly. I fixed this by
5415 always allowing "-C /dev/null" if the caller is root.
5417 14. Ignore overlarge ACL variable number when reading spool file - insurance
5418 against a later release with more variables having written the file.
5420 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5421 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5422 this, as well as the form without the "IPv6" on the front.
5424 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5425 OS/Makefile-Darwin file.
5427 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5428 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5430 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5431 variable. However, some OS use ints for inodes. I've added cast to long int
5432 to get rid of the compiler warning.
5434 19. I had forgotten to lock out "/../" in configuration file names when
5435 ALT_CONFIG_PREFIX was set.
5437 20. Routers used for verification do not need to specify transports. However,
5438 if such a router generated a host list, and callout was configured, Exim
5439 crashed, because it could not find a port number from the (non-existent)
5440 transport. It now assumes port 25 in this circumstance.
5442 21. Added the -t option to exigrep.
5444 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5445 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5446 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5447 them is set, LOOKUP_LSEARCH is forced.
5449 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5451 24. Added sender and host information to the "rejected by local_scan()" log
5452 line; previously there was no indication of these.
5454 25. Added .include_if_exists.
5456 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5457 Linux. It turns out that not all file systems support this. Apparently some
5458 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5459 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5460 not supported on the file descriptor, is now ignored when Exim is trying to
5461 sync a directory. This applies only to Linux.
5463 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5465 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5466 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5469 29. Added support for maildirsize files from supplied patch (modified a bit).
5471 30. The use of :fail: followed by an empty string could lead Exim to respond to
5472 sender verification failures with (e.g.):
5474 550 Verification failed for <xxx>
5475 550 Sender verify failed
5477 where the first response line was missing the '-' that indicates it is not
5478 the final line of the response.
5480 31. The loop for finding the name of the user that called Exim had a hardwired
5481 limit of 10; it now uses the value of finduser_retries, which is used for
5482 all other user lookups.
5484 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5487 33. Exim was neglecting to zero errno before one call of strtol() when
5488 expanding a string and expecting an integer value. On some systems this
5489 resulted in spurious "integer overflow" errors. Also, it was casting the
5490 result into an int without checking.
5492 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5493 did not work. The code looks as if it has *never* worked, though it appears
5494 to have been documented since at least releast 1.62. I have made it work.
5496 35. The "timeout_DNS" error in retry rules, also documented since at least
5497 1.62, also never worked. As it isn't clear exactly what this means, and
5498 clearly it isn't a major issue, I have abolished the feature by treating it
5499 as "timeout", and writing a warning to the main and panic logs.
5501 36. The display of retry rules for -brt wasn't always showing the error code
5504 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5505 timeout_connect_A, timeout_connect_MX.
5507 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5508 to the empty sender.
5510 39. The daemon was not analysing the content of -oX till after it had closed
5511 stderr and disconnected from the controlling terminal. This meant that any
5512 syntax errors were only noted on the panic log, and the return code from
5513 the command was 0. By re-arranging the code a little, I've made the
5514 decoding happen first, so such errors now appear on stderr, and the return
5515 code is 1. However, the actual setting up of the sockets still happens in
5516 the disconnected process, so errors there are still only recorded on the
5519 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5520 connections (as happens on some IP stacks) was logged at start up time as
5521 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5522 it from "IPv6 and IPv4", which means that two separate sockets are being
5525 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5526 says whether AF_INET or AF_INET6 was passed as an argument.
5528 42. Exiwhat output was messed up when time zones were included in log
5531 43. Exiwhat now gives more information about the daemon's listening ports,
5532 and whether -tls-on-connect was used.
5534 44. The "port" option of the smtp transport is now expanded.
5536 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5537 silently ignored. Now an error message is written to the main and panic
5540 46. There's a new ACL modifier called "logwrite" which writes to a log file
5541 as soon as it is encountered.
5543 47. Added $local_user_uid and $local_user_gid at routing time.
5545 48. Exim crashed when trying to verify a sender address that was being
5548 49. Exim was recognizing only a space character after ".include". It now also
5549 recognizes a tab character.
5551 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5552 extracting the relevant information from the specification. The man page no
5553 longer contains scrambled data for the -d option, and I've added a section
5554 at the front about calling Exim under different names.
5556 51. Added "extra_headers" argument to the "mail" command in filter files.
5558 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5561 53. Installed eximstats 1.29.
5563 54. Added transport_filter_timeout as a generic transport option.
5565 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5566 Cc: header lines. This was required by RFC 822, but it not required by RFC
5569 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5570 incoming messages that did not have them. Now it does so only if the
5571 message originates locally, that is, if there is no associated remote host
5572 address. When Resent- header lines are present, this applies to the Resent-
5573 lines rather than the non-Resent- lines.
5575 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5576 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5578 58. Messages for configuration errors now include the name of the main
5579 configuration file - useful now that there may be more than one file in a
5580 list (.included file names were always shown).
5582 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5583 for those rare installations that do not start the daemon as root or run it
5584 setuid root. I've cut out the call to initgroups() if the daemon is not
5587 60. The Exim user and group can now be bound into the binary as text strings
5588 that are looked up at the start of Exim's processing.
5590 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5592 62. Added $mailstore_basename variable.
5594 63. Installed patch to sieve.c from Michael Haardt.
5596 64. When Exim failed to open the panic log after failing to open the main log,
5597 the original message it was trying to log was written to stderr and debug
5598 output, but if they were not available (the usual case in production), it
5599 was lost. Now it is written to syslog before the two lines that record the
5600 failures to open the logs.
5602 65. Users' Exim filters run in subprocesses under the user's uid. It is
5603 possible for a "deliver" command or an alias in a "personal" command to
5604 provoke an address rewrite. If logging of address rewriting is configured,
5605 this fails because the process is not running as root or exim. There may be
5606 a better way of dealing with this, but for the moment (because 4.30 needs
5607 to be released), I have disabled address rewrite logging when running a
5608 filter in a non-root, non-exim process.
5614 1. The buildconfig auxiliary program wasn't quoting the value set for
5615 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5616 not defined. This bug was masked in 4.22 by the effect that was fixed in
5619 2. Some messages that were rejected after a message id was allocated were
5620 shown as "incomplete" by exigrep. It no longer does this for messages that
5621 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5623 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5624 have allow_domain_literals set, the ID did not get logged in the <= line.
5625 Domain literals are now always recognized in Message-ID: header lines.
5627 4. The first argument for a ${extract expansion item is the key name or field
5628 number. Leading and trailing spaces in this item were not being ignored,
5629 causing some misleading effects.
5631 5. When deliver_drop_privilege was set, single queue runner processes started
5632 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5633 same command in the process it spins off) were not dropping privilege.
5635 6. When the daemon running as "exim" started a queue runner, it always
5636 re-executed Exim in the spun-off process. This is a waste of effort when
5637 deliver_drop_privilege is set. The new process now just calls the
5638 queue-runner function directly.
5644 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5647 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5648 ignored. Though the use of -oP was forcing the writing of a pid file, it
5649 was always written to the default place.
5651 3. If the message "no IP address found for host xxxx" is generated during
5652 incoming verification, it is now followed by identification of the incoming
5653 connection (so you can more easily find what provoked it).
5655 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5657 5. Added some features to "harden" Exim a bit more against certain attacks:
5659 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5660 be put in Local/Makefile. This is like the never_users runtime option,
5661 but it cannot be overridden. The default setting is "root".
5663 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5664 prefix string with which any file named in a -C command line option
5667 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5668 is retained for -C and -D only if the caller of Exim is root. Without
5669 it, the exim user may also use -C and -D and retain privilege.
5671 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5672 command line option is disabled.
5674 6. Macro names set by the -D option must start with an upper case letter, just
5675 like macro names defined in the configuration file.
5677 7. Added "dereference=" facility to LDAP.
5679 8. Two instances of the typo "uknown" in the source files are fixed.
5681 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5682 the Configure-Makefile script screwed up while processing it.
5684 10. Incorporated PCRE 4.4.
5686 11. The SMTP synchronization check was not operating right at the start of an
5687 SMTP session. For example, it could not catch a HELO sent before the client
5688 waited for the greeting. There is now a check for outstanding input at the
5689 point when the greeting is written. Because of the duplex, asynchronous
5690 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5691 way, but not yet received, when the check is performed.
5693 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5694 on TCP/IP sockets, because this apparently causes some broken clients to
5697 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5698 unchanged) from the Cygwin maintainer.
5700 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5701 was supported instead of testing for mbx. Effectively a typo.
5703 15. The spa authenticator server code was not checking that the input it
5704 received was valid base64.
5706 16. The debug output line for the "set" modifier in ACLs was not showing the
5707 name of the variable that was being set.
5709 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5711 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5712 lookup on its own. Something else was needed to trigger the lookup. For
5713 example, a match in host_lookup or the need for a host name in a host list.
5714 Now, if $sender_host_name is referenced and the host name has not yet been
5715 looked up, a lookup is performed. If the lookup fails, the variable remains
5716 empty, and $host_lookup_failed is set to "1".
5718 19. Added "eqi" as a case-independent comparison operator.
5720 20. The saslauthd authentication condition could segfault if neither service
5721 nor realm was specified.
5723 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5724 error message that was logged was misleading, and incoming SMTP
5725 connections were dropped. The message is now more accurate, and temporary
5726 errors are given to SMTP connections.
5728 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5729 (except RSET) with a 421 error, until QUIT is received. However, it was
5730 failing to send a response to QUIT.
5732 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5733 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5734 ACL is now run after the helo_try_verify_hosts code.
5736 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5737 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5738 case-independent, so other case variants are also recognized). Apparently
5739 some systems use these upper case variants.
5741 25. If more than two messages were waiting for the same host, and a transport
5742 filter was specified for the transport, Exim sent two messages over the
5743 same TCP/IP connection, and then failed with "socket operation on non-
5744 socket" when it tried to send the third.
5746 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5748 27. The extern definition of crypt16() in expand.c was not being excluded when
5749 the OS had its own crypt16() function.
5751 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5752 as a preferred synonym for return_size_limit, both as an option and as an
5755 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5757 30. Changed the default configuration ACL to relax the local part checking rule
5758 for addresses that are not in any local domains. For these addresses,
5759 slashes and pipe symbols are allowed within local parts, but the sequence
5760 /../ is explicitly forbidden.
5762 31. SPA server authentication was not clearing the challenge buffer before
5765 32. log_message in a "warn" ACL statement was writing to the reject log as
5766 well as to the main log, which contradicts the documentation and doesn't
5767 seem right (because no rejection is happening). So I have stopped it.
5769 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5770 However, I am unable to do any testing of this.
5772 34. Fixed an infelicity in the appendfile transport. When checking directories
5773 for a mailbox, to see if any needed to be created, it was accidentally
5774 using path names with one or more superfluous leading slashes; tracing
5775 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5777 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5778 contents are added to the log line that is written for every discarded
5779 recipient. (Previously a log_message setting was ignored.)
5781 36. The ${quote: operator now quotes the string if it is empty.
5783 37. The install script runs exim in order to find its version number. If for
5784 some reason other than non-existence or emptiness, which it checks, it
5785 could not run './exim', it was installing it with an empty version number,
5786 i.e. as "exim-". This error state is now caught, and the installation is
5789 38. An argument was missing from the function that creates an error message
5790 when Exim fails to connect to the socket for saslauthd authentication.
5791 This could cause Exim to crash, or give a corrupted message.
5793 39. Added isip, isip4, and isip6 to ${if conditions.
5795 40. The ACL variables $acl_xx are now saved with the message, and can be
5796 accessed later in routers, transports, and filters.
5798 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5799 strings in the file are not string-expanded.
5801 42. If a MAIL command specified a SIZE value that was too large to fit into an
5802 int variable, the check against message_size_limit failed. Such values are
5803 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5804 day this will have to be increased, but I don't think I want to be around
5805 when emails are that large.
5812 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5813 iconv() is not standard in FreeBSD.
5815 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5816 IPv6 enabled. The observed symptom was a segmentation fault on return from
5817 the function os_common_find_running_interfaces() in src/os.c.
5819 3. In the check_special_case() function in daemon.c I had used "errno" as an
5820 argument name, which causes warnings on some systems. This was basically a
5821 typo, since it was named "eno" in the comments!
5823 4. The code that waits for the clock to tick (at a resolution of some fraction
5824 of a second) so as to ensure message-id uniqueness was always waiting for
5825 at least one whole tick, when it could have waited for less. [This is
5826 almost certainly not relevant at current processor speeds, where it is
5827 unlikely to ever wait at all. But we try to future-proof.]
5829 5. The function that sleeps for a time interval that includes fractions of a
5830 second contained a race. It did not block SIGALRM between setting the
5831 timer, and suspending (a couple of lines later). If the interval was short
5832 and the sigsuspend() was delayed until after it had expired, the suspension
5833 never ended. On busy systems this could lead to processes getting stuck for
5836 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5837 process, before it forks any delivery processes. The open lookup caching
5838 mechanism meant that the open file or database connection was passed into
5839 the delivery process. The problem was that delivery processes always tidy
5840 up cached lookup data. This could cause a problem for the next delivery
5841 process started by the queue runner, because the external queue runner
5842 process does not know about the closure. So the next delivery process
5843 still has data in the lookup cache. In the case of a file lookup, there was
5844 no problem because closing a file descriptor in a subprocess doesn't affect
5845 the parent. However, if the lookup was caching a connection to a database,
5846 the connection was closed, and the second delivery process was likely to
5847 see errors such as "PGSQL: query failed: server closed the connection
5848 unexpectedly". The problem has been fixed by closing all cached lookups
5849 in a queue runner before running a delivery process.
5851 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5852 seem to have the "const" qualifier which it has on other OS. I've
5855 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5856 *of the same type* (client or server) with the same public name that an
5857 error should be diagnosed.
5859 9. When Exim looked up a host name for an IP address, but failed to find the
5860 original IP address when looking up the host name (a safety check), it
5861 output the message "<ip address> does not match any IP for NULL", which was
5862 confusing, to say the least. The bug was that the host name should have
5863 appeared instead of "NULL".
5865 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5866 user that is built into the binary, and the -C or -D options is used, root
5867 privilege is dropped before the configuration file is read. In addition,
5868 logging is switched to stderr instead of the normal log files. If the
5869 configuration then re-defines the Exim user, the unprivileged environment
5870 is probably not what is expected, so Exim logs a panic warning message (but
5873 However, if deliver_drop_privilege is set, the unprivileged state may well
5874 be exactly what is intended, so the warning has been cut out in that case,
5875 and Exim is allowed to try to write to its normal log files.
5881 1. smtp_return_error_details was not giving details for temporary sender
5882 or receiver verification errors.
5884 2. Diagnose a configuration error if two authenticators have the same public
5887 3. Exim used not to create the message log file for a message until the first
5888 delivery attempt. This could be confusing when incoming messages were held
5889 for policy or load reasons. The message log file is now created at the time
5890 the message is received, and an initial "Received" line is written to it.
5892 4. The automatically generated man page for command line options had a minor
5893 bug that caused no ill effects; however, a more serious problem was that
5894 the procedure for building the man page automatically didn't always
5895 operate. Consequently, release 4.20 contains an out-of-date version. This
5896 shouldn't happen again.
5898 5. When building Exim with embedded Perl support, the script that builds the
5899 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5900 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5902 6. The freeze_tell option was not being used for messages that were frozen on
5903 arrival, either by an ACL or by local_scan().
5905 7. Added the smtp_incomplete_transaction log selector.
5907 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5908 was accepting AUTH without a new EHLO.
5910 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5911 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5914 10. Logging of TCP/IP connections (when configured) now happens in the main
5915 daemon process instead of the child process, so that the TCP/IP connection
5916 count is more accurate (but it can never be perfect).
5918 11. The use of "drop" in a nested ACL was not being handled correctly in the
5919 outer ACL. Now, if condition failure induced by the nested "drop" causes
5920 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5921 or "require"), the connection is dropped.
5923 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5924 that yield "discard" can now be used with an "accept" or a "discard" verb,
5925 but an error is generated for any others (because I can't see a useful way
5926 to define what should happen).
5928 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5929 are now processed in the same way as lines in the Exim configuration file.
5930 In particular, continuation lines are supported.
5932 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5934 15. Added -ti meaning -t -i.
5936 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5937 domains, and warn by logging if others are found.
5939 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5940 returned from GIFCONF when Exim is trying to find the list of interfaces on
5941 a host. The code in os.c has been modified to copy each ifreq to an aligned
5942 structure in all cases.
5944 Also, in some cases, the returned ifreq's were being copied to a 'struct
5945 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5946 means the last couple of bytes of an IPv6 address could be chopped if the
5947 ifreq contained only a normal sockaddr (14 bytes storage).
5949 18. Named domain lists were not supported in the hosts_treat_as_local option.
5950 An entry such as +xxxx was not recognized, and was treated as a literal
5953 19. Ensure that header lines added by a DATA ACL are included in the reject log
5954 if the ACL subsequently rejects the message.
5956 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5957 MD5 (which is deprecated).
5959 21. When testing a filter file using -bf, Exim was writing a message when it
5960 took the sender from a "From " line in the message, but it was not doing so
5961 when it took $return_path from a Return-Path: header line. It now does.
5963 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5964 with a valid header line field name (a series of printing characters
5965 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5967 23. Changed "disc" in the source to "disk" to conform to the documentation and
5968 the book and for uniformity.
5970 24. Ignore Sendmail's -Ooption=value command line item.
5972 25. When execve() failed while trying to run a command in a pipe transport,
5973 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5974 could be confused with a return value of 69 from the command itself. This
5975 has been changed to 127, the value the shell returns if it is asked to run
5976 a non-existent command. The wording for the related log line suggests a
5977 non-existent command as the problem.
5979 26. If received_header_text expands to an empty string, do not add a Received:
5980 header line to the message. (Well, it adds a token one on the spool, but
5981 marks it "old" so that it doesn't get used or transmitted.)
5983 27. Installed eximstats 1.28 (addition of -nt option).
5985 28. There was no check for failure on the call to getsockname() in the daemon
5986 code. This can fail if there is a shortage of resources on the system, with
5987 ENOMEM, for example. A temporary error is now given on failure.
5989 29. Contrary to the C standard, it seems that in some environments, the
5990 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5991 program. Exim now does this explicitly; it affects the formatting of
5992 timestamps using strftime().
5994 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5995 complaints. I've now initialized all the variables, to avoid this.
5997 32. Header lines added by a system filter were not being "seen" during
5998 transport-time rewrites.
6000 33. The info_callback() function passed to OpenSSL is set up with type void
6001 (*)(SSL *, int, int), as described somewhere. However, when calling the
6002 function (actually a macro) that sets it up, the type void(*)() is
6003 expected. I've put in a cast to prevent warnings from picky compilers.
6005 34. If a DNS black list lookup found a CNAME record, but there were no A
6006 records associated with the domain it pointed at, Exim crashed.
6008 35. If a DNS black list lookup returned more than one A record, Exim ignored
6009 all but the first. It now scans all returned addresses if a particular IP
6010 value is being sought. In this situation, the contents of the
6011 $dnslist_value variable are a list of all the addresses, separated by a
6014 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6015 to accept a host name and all its aliases if the forward lookup for any of
6016 them yielded the IP address of the incoming connection. Now it accepts only
6017 those names whose forward lookup yields the correct IP address. Any other
6018 names are discarded. This closes a loophole whereby a rogue DNS
6019 administrator could create reverse DNS records to break through a
6020 wildcarded host restriction in an ACL.
6022 37. If a user filter or a system filter that ran in a subprocess used any of
6023 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6024 the wrong values were passed to the pipe command ($thisaddress had the
6025 value of $0, $0 had the value of $1, etc). This bug was introduced by
6026 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6028 38. Improved the line breaking for long SMTP error messages from ACLs.
6029 Previously, if there was no break point between 40 and 75 characters, Exim
6030 left the rest of the message alone. Two changes have been made: (a) I've
6031 reduced the minimum length to 35 characters; (b) if it can't find a break
6032 point between 35 and 75 characters, it looks ahead and uses the first one
6033 that it finds. This may give the occasional overlong line, but at least the
6034 remaining text gets split now.
6036 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6037 file descriptors might be low, and that setting 1000 would always raise it.
6038 It turns out that in some environments, the limit is already over 1000 and
6039 that lowering it causes trouble. So now Exim takes care not to decrease it.
6041 40. When delivering a message, the value of $return_path is set to $sender_
6042 address at the start of routing (routers may change the value). By an
6043 oversight, this default was not being set up when an address was tested by
6044 -bt or -bv, which affected the outcome if any router or filter referred to
6047 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6048 log only when "message" or "log_message" are set. However, if one of the
6049 conditions was an address verification, or a call to a nested ACL, the
6050 messages generated by the underlying test were being passed through. This
6051 no longer happens. The underlying message is available in $acl_verify_
6052 message for both "message" and "log_message" expansions, so it can be
6053 passed through if needed.
6055 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6056 new expansion $bh_ to give the encoded byte string without charset
6057 translation. Translation happens only if iconv() is available; HAVE_ICONV
6058 indicates this at build time. HEADERS_CHARSET gives the charset to
6059 translate to; headers_charset can change it in the configuration, and
6060 "headers charset" can change it in an individual filter file.
6062 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6063 that creates RFC 2047 encoded "words" labels them as that charset instead
6064 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6065 expansion operator; (ii) when Exim creates a From: line for a local
6066 message; (iii) when a header line is rewritten to include a "phrase" part.
6068 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6069 buggy, causing it to skip the first lines of messages whose message ID
6070 ended in 'D'. This would not have bitten before Exim release 4.14, because
6071 message IDs were unlikely to end in 'D' before then. The effect was to have
6072 incorrect size information for certain domains.
6074 45. #include "config.h" was missing at the start of the crypt16.c module. This
6075 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6078 46. If there was a timeout during a "random" callout check, Exim treated it as
6079 a failure of the random address, and carried on sending RSET and the real
6080 address. If the delay was just some slowness somewhere, the response to the
6081 original RCPT would be taken as a response to RSET and so on, causing
6082 mayhem of various kinds.
6084 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6085 when I implemented it. It didn't allow for the fact that some option values
6086 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6087 the right test for positive values.
6089 48. Domain names in DNS records are case-independent. Exim always looks them up
6090 in lower case. Some resolvers return domain names in exactly the case they
6091 appear in the zone file, that is, they may contain uppercase letters. Not
6092 all resolvers do this - some return always lower case. Exim was treating a
6093 change of case by a resolver as a change of domain, similar to a widening
6094 of a domain abbreviation. This triggered its re-routing code and so it was
6095 trying to route what was effectively the same domain again. This normally
6096 caused routing to fail (because the router wouldn't handle the domain
6097 twice). Now Exim checks for this case specially, and just changes the
6098 casing of the domain that it ultimately uses when it transmits the message
6101 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6104 50. If a filter generated a file delivery with a non-absolute name (possible if
6105 no home directory exists for the router), the forbid_file option was not
6108 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6109 the existing equality matching.
6111 52. Exim was using ints instead of ino_t variables in some places where it was
6112 dealing with inode numbers.
6114 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6115 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6116 called TMPDIR, and if it finds it is different, it changes its value.
6118 54. The smtp_printf() function is now made available to local_scan() so
6119 additional output lines can be written before returning. There is also an
6120 smtp_fflush() function to enable the detection of a dropped connection.
6121 The variables smtp_input and smtp_batched_input are exported to
6124 55. Changed the default runtime configuration: the message "Unknown user"
6125 has been removed from the ACL, and instead placed on the localuser router,
6126 using the cannot_route_message feature. This means that any verification
6127 failures that generate their own messages won't get overridden. Similarly,
6128 the "Unrouteable address" message that was in the ACL for unverifiable
6129 relay addresses has also been removed.
6131 56. Added hosts_avoid_esmtp to the smtp transport.
6133 57. The exicyclog script was not checking for the esoteric option
6134 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6135 will work only if exicyclog is run under the appropriate euid.
6137 58. Following a discussion on the list, the rules by which Exim recognises line
6138 endings on incoming messages have been changed. The -dropcr and drop_cr
6139 options are now no-ops, retained only for backwards compatibility. The
6140 following line terminators are recognized: LF CRLF CR. However, special
6141 processing applies to CR:
6143 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6144 nor a local message in the state where . is a terminator.
6146 (ii) If a bare CR is encountered in a header line, an extra space is added
6147 after the line terminator so as not to end the header. The reasoning
6148 behind this is that bare CRs in header lines are most likely either
6149 to be mistakes, or people trying to play silly games.
6151 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6152 was being incorrectly given as 18 bytes larger than it should have been.
6153 This is a VOB (very old bug).
6155 60. This may never have affected anything current, but just in case it has:
6156 When the local host is found other than at the start of a list of hosts,
6157 the local host, those with the same MX, and any that follow, are discarded.
6158 When the list in question was part of a longer list of hosts, the following
6159 hosts (not currently being processed) were also being discarded. This no
6160 longer happens. I'm not sure if this situation could ever has previously
6163 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6166 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6167 header. This is something that is recommended in a new Internet Draft, and
6168 is something that is documented as being done by Sendmail. There are two
6169 possible values. For messages generated by the autoreply transport, Exim
6172 Auto-Submitted: auto-replied
6174 whereas for all other generated messages (e.g. bounces) it adds
6176 Auto-Submitted: auto-generated
6178 63. The "personal" condition in filters now includes a test for the
6179 Auto-Submitted: header. If it contains the string "auto-" the message it
6180 not considered personal.
6182 64. Added rcpt_include_affixes as a generic transport option.
6184 65. Added queue_only_override (default true).
6186 66. Added the syslog_duplication option.
6188 67. If what should have been the first header line of a message consisted of
6189 a space followed by a colon, Exim was mis-interpreting it as a header line.
6190 It isn't of course - it is syntactically invalid and should therefore be
6191 treated as the start of the message body. The misbehaviour could have
6192 caused a number of strange effects, including loss of data in subsequent
6193 header lines, and spool format errors.
6195 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6196 client host had authenticated. This control can now be exercised by an ACL
6197 for more flexibility.
6199 69. By default, callouts do not happen when testing with -bh. There is now a
6200 variant, -bhc, which does actually run the callout code, including
6201 consulting and updating the callout cache.
6203 70. Added support for saslauthd authentication, courtesy of Alexander
6206 71. If statvfs() failed on the spool or log directories while checking their
6207 size for availability, Exim confusingly gave the error "space shortage".
6208 Furthermore, in debugging mode it crashed with a floating point exception.
6209 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6210 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6211 serious problem, Exim now writes to the main and panic logs when this
6212 happens, with details of the failure. It then refuses to accept the
6213 incoming message, giving the message "spool directory problem" or "log
6214 directory problem" with a 421 code for SMTP messages.
6216 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6217 0, 1, and 2 exist, because some OS complain for execs without them (see
6218 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6219 descriptors. However, the code omitted to check that the open succeeded,
6220 causing mysterious errors if for some reason the permissions on /dev/null
6221 got screwed. Now Exim writes a message to the main and panic logs, and
6222 bombs out if it can't open /dev/null.
6224 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6225 interact so that it is all more flexible. It is supposed to remain
6226 backwards compatible. Also added extra_local_interfaces.
6228 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6229 to bomb out with an assertion failure - to the client this appears as a
6230 connection drop. This problem occurs in the part of the code that was taken
6231 from the Samba project. Fortunately, the assertion is in a very simple
6232 function, so I have fixed this by reproducing the function inline in the
6233 one place where it is called, and arranging for authentication to fail
6234 instead of killing the process with assert().
6236 75. The SPA client code was not working when the server requested OEM rather
6237 than Unicode encoding.
6239 76. Added code to make require_files with a specific uid setting more usable in
6240 the case where statting the file as root fails - usually a non-root-mounted
6241 NFS file system. When this happens and the failure is EACCES, Exim now
6242 forks a subprocess and does the per-uid checking as the relevant uid.
6244 77. Added process_log_path.
6246 78. If log_file_path was not explicitly set, a setting of check_log_space or
6247 check_log_inodes was ignored.
6249 79. If a space check for the spool or log partitions fails, the incident is now
6250 logged. Of course, in the latter case the data may get lost...
6252 80. Added the %p formatting code to string_format() so that it can be used to
6253 print addresses in debug_print(). Adjusted all the address printing in the
6254 debugging in store.c to use %p rather than %d.
6256 81. There was a concern that a line of code in smtp_in.c could overflow a
6257 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6258 initially expressed, the concern was not well-founded, because trailing
6259 spaces are removed early. However, if the trailing spaces were followed by
6260 a NULL, they did not get removed, so the overflow was possible. Two fixes
6263 (a) I re-wrote the offending code in a cleaner fashion.
6264 (b) If an incoming SMTP command contains a NULL character, it is rejected
6267 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6268 runs initgroups(), so that if the Exim user is in any additional groups,
6269 they will be used during message reception.
6275 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob