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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
92 TL/01 When running the test suite, the README says that variables such as
93 no_msglog_check are global and can be placed anywhere in a specific
94 test's script, however it was observed that placement needed to be near
95 the beginning for it to behave that way. Changed the runtest perl
96 script to read through the entire script once to detect and set these
97 variables, reset to the beginning of the script, and then run through
98 the script parsing/test process like normal.
100 TL/02 The BSD's have an arc4random API. One of the functions to induce
101 adding randomness was arc4random_stir(), but it has been removed in
102 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
103 function when detected.
105 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
106 cause callback expansion.
108 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
109 syntax errors in an expansion can be treated as a string instead of
110 logging or causing an error, due to the internal use of bool_lax
111 instead of bool when processing it.
113 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
114 server certificates when making smtp deliveries.
116 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
118 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
120 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
122 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
123 Merged patch from Sebastian Wiedenroth.
125 JH/05 Fix results-pipe from transport process. Several recipients, combined
126 with certificate use, exposed issues where response data items split
127 over buffer boundaries were not parsed properly. This eventually
128 resulted in duplicates being sent. This issue only became common enough
129 to notice due to the introduction of conection certificate information,
130 the item size being so much larger. Found and fixed by Wolfgang Breyha.
132 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
133 size buffer was used, resulting in syntax errors when an expansion
136 JH/07 Add support for directories of certificates when compiled with a GnuTLS
137 version 3.3.6 or later.
139 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
140 is EXPERIMENTAL_EVENT, the main-configuration and transport options
141 both become "event_action", the variables become $event_name, $event_data
142 and $event_defer_errno. There is a new variable $verify_mode, usable in
143 routers, transports and related events. The tls:cert event is now also
144 raised for inbound connections, if the main configuration event_action
147 TL/06 In test suite, disable OCSP for old versions of openssl which contained
148 early OCSP support, but no stapling (appears to be less than 1.0.0).
150 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
151 server certificate names available under the smtp transport option
152 "tls_verify_cert_hostname" now do not permit multi-component wildcard
155 JH/10 Time-related extraction expansions from certificates now use the main
156 option "timezone" setting for output formatting, and are consistent
157 between OpenSSL and GnuTLS compilations. Bug 1541.
159 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
160 encoded parameter in the incoming message. Bug 1558.
162 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
163 include certificate info, eximon was claiming there were spoolfile
166 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
168 JH/14 Log delivery-related information more consistently, using the sequence
169 "H=<name> [<ip>]" wherever possible.
171 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
172 are problematic for Debian distribution, omit them from the release
175 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
177 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
180 JH/17 Fix a null-indirection in certextract expansions when a nondefault
181 output list separator was used.
186 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
187 checkers that were complaining about end of non-void function with no
190 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
191 This was a regression intruduced in 4.83 by another bugfix.
193 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
195 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
196 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
202 TF/01 Correctly close the server side of TLS when forking for delivery.
204 When a message was received over SMTP with TLS, Exim failed to clear up
205 the incoming connection properly after forking off the child process to
206 deliver the message. In some situations the subsequent outgoing
207 delivery connection happened to have the same fd number as the incoming
208 connection previously had. Exim would try to use TLS and fail, logging
209 a "Bad file descriptor" error.
211 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
212 utilities have not been installed.
214 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
215 temporary space as the ACL may create new global variables.
217 TL/01 LDAP support uses per connection or global context settings, depending
218 upon the detected version of the libraries at build time.
220 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
221 to extract and use the src ip:port in logging and expansions as if it
222 were a direct connection from the outside internet. PPv2 support was
223 updated based on HAProxy spec change in May 2014.
225 JH/02 Add ${listextract {number}{list}{success}{fail}}.
227 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
228 Properly escape header and check for NULL return.
230 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
233 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
235 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
236 characters in header names, implemented as a verify condition.
237 Contributed by Michael Fischer v. Mollard.
239 TL/05 Rename SPF condition results err_perm and err_temp to standardized
240 results permerror and temperror. Previous values are deprecated but
241 still accepted. In a future release, err_perm and err_temp will be
242 completely removed, which will be a backward incompatibility if the
243 ACL tests for either of these two old results. Patch contributed by
244 user bes-internal on the mailing list.
246 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
248 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
249 selectors, in both main and reject logs.
251 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
254 JH/07 Add malware type "sock" for talking to simple daemon.
256 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
258 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
259 routers/transports under cutthrough routing.
261 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
262 numbers. Touch up "bool" conditional to keep the same definition.
264 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
266 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
268 JH/12 Expand items in router/transport headers_add or headers_remove lists
269 individually rather than the list as a whole. Bug 1452.
271 Required for reasonable handling of multiple headers_ options when
272 they may be empty; requires that headers_remove items with embedded
273 colons must have them doubled (or the list-separator changed).
275 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
276 view the policy declared in the DMARC record. Currently, $dmarc_status
277 is a combined value of both the record presence and the result of the
280 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
282 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
283 dnslookup router and the smtp transport (applying to the forward
286 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
287 of ldap servers used for a specific lookup. Patch provided by Heiko
290 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
291 New variable $lookup_dnssec_authenticated for observability.
293 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
294 Patch submitted by Lars Timman.
296 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
298 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
299 Requires trusted mode and valid format message id, aborts otherwise.
300 Patch contributed by Heiko Schlichting.
302 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
303 certextract with support for various fields. Bug 1358.
305 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
306 is requested by default, modifiable by smtp transport option
309 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
310 operate on certificate variables to give certificate fingerprints
311 Also new ${sha256:cert_variable}.
313 JH/23 The PRDR feature is moved from being Experimental into the mainline.
315 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
316 Christian Aistleitner.
318 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
320 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
321 file. Patch from Wolfgang Breyha.
323 JH/25 Expand the coverage of the delivery $host and $host_address to
324 client authenticators run in verify callout. Bug 1476.
326 JH/26 Port service names are now accepted for tls_on_connect_ports, to
327 align with daemon_smtp_ports. Bug 72.
329 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
330 support and error reporting did not work properly.
332 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
333 and is readable. Patch from Andrew Colin Kissa.
335 TL/14 Enhance documentation of ${run expansion and how it parses the
336 commandline after expansion, particularly in the case when an
337 unquoted variable expansion results in an empty value.
339 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
341 PP/02 Fix internal collision of T_APL on systems which support RFC3123
342 by renaming away from it. Addresses GH issue 15, reported by
345 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
347 TL/15 SECURITY: prevent double expansion in math comparison functions
348 (can expand unsanitized data). Not remotely exploitable.
355 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
357 PP/02 Make -n do something, by making it not do something.
358 When combined with -bP, the name of an option is not output.
360 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
363 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
364 $sender_host_name and config options to manage this, and basic check
367 PP/05 DSCP support for outbound connections and control modifier for inbound.
369 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
370 (Only plugin which currently uses this is kerberos4, which nobody should
371 be using, but we should make it available and other future plugins might
372 conceivably use it, even though it would break NAT; stuff *should* be
373 using channel bindings instead).
375 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
376 name; added for Sendmail compatibility; requires admin caller.
377 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
378 just ignore it); requires trusted caller.
379 Also parse but ignore: -Ac -Am -X<logfile>
382 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
384 TL/02 Add +smtp_confirmation as a default logging option.
386 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
387 Patch by Magnus Holmgren from 2007-02-20.
389 TL/04 Bugzilla 1281 - Spec typo.
390 Bugzilla 1283 - Spec typo.
391 Bugzilla 1290 - Spec grammar fixes.
393 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
395 TL/06 Add Experimental DMARC support using libopendmarc libraries.
397 TL/07 Fix an out of order global option causing a segfault. Reported to dev
398 mailing list by by Dmitry Isaikin.
400 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
402 JH/02 Support "G" suffix to numbers in ${if comparisons.
404 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
406 NM/01 Bugzilla 1197 - Spec typo
407 Bugzilla 1196 - Spec examples corrections
409 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
411 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
412 gnutls_enable_pkcs11, but renamed to more accurately indicate its
415 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
416 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
418 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
419 "acl {{name}{arg}...}", and optional args on acl condition
422 JH/05 Permit multiple router/transport headers_add/remove lines.
424 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
426 JH/07 Avoid using a waiting database for a single-message-only transport.
427 Performance patch from Paul Fisher. Bugzilla 1262.
429 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
432 JH/09 Add $headers_added variable, with content from use of ACL modifier
433 add_header (but not yet added to the message). Bugzilla 199.
435 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
436 Pulled from Bugzilla 817 by Wolfgang Breyha.
438 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
440 (nb: this is the same fix as in Exim 4.80.1)
442 JH/11 Add A= logging on delivery lines, and a client_set_id option on
445 JH/12 Add optional authenticated_sender logging to A= and a log_selector
448 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
450 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
451 advertise SMTP AUTH mechanism to us, instead of a generic
452 protocol violation error. Also, make Exim more robust to bad
453 data from the Dovecot auth socket.
455 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
457 When a queue runner is handling a message, Exim first routes the
458 recipient addresses, during which it prunes them based on the retry
459 hints database. After that it attempts to deliver the message to
460 any remaining recipients. It then updates the hints database using
463 So if a recipient address works intermittently, it can get repeatedly
464 deferred at routing time. The retry hints record remains fresh so the
465 address never reaches the final cutoff time.
467 This is a fairly common occurrence when a user is bumping up against
468 their storage quota. Exim had some logic in its local delivery code
469 to deal with this. However it did not apply to per-recipient defers
470 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
472 This change adds a proper retry rule check during routing so that the
473 final cutoff time is checked against the message's age. We only do
474 this check if there is an address retry record and there is not a
475 domain retry record; this implies that previous attempts to handle
476 the address had the retry_use_local_parts option turned on. We use
477 this as an approximation for the destination being like a local
478 delivery, as in LMTP.
480 I suspect this new check makes the old local delivery cutoff check
481 redundant, but I have not verified this so I left the code in place.
483 TF/02 Correct gecos expansion when From: is a prefix of the username.
485 Test 0254 submits a message to Exim with the header
489 When I ran the test suite under the user fanf2, Exim expanded
490 the header to contain my full name, whereas it should have added
491 a Resent-Sender: header. It erroneously treats any prefix of the
492 username as equal to the username.
494 This change corrects that bug.
496 GF/01 DCC debug and logging tidyup
497 Error conditions log to paniclog rather than rejectlog.
498 Debug lines prefixed by "DCC: " to remove any ambiguity.
500 TF/03 Avoid unnecessary rebuilds of lookup-related code.
502 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
503 Bug spotted by Jeremy Harris; was flawed since initial commit.
504 Would have resulted in OCSP responses post-SNI triggering an Exim
505 NULL dereference and crash.
507 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
509 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
510 Bug detection, analysis and fix by Samuel Thibault.
511 Bugzilla 1331, Debian bug #698092.
513 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
515 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
516 Server implementation by Todd Lyons, client by JH.
517 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
518 config variable "prdr_enable" controls whether the server
519 advertises the facility. If the client requests PRDR a new
520 acl_data_smtp_prdr ACL is called once for each recipient, after
521 the body content is received and before the acl_smtp_data ACL.
522 The client is controlled by bolth of: a hosts_try_prdr option
523 on the smtp transport, and the server advertisement.
524 Default client logging of deliveries and rejections involving
525 PRDR are flagged with the string "PRDR".
527 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
528 fclose(). Diagnosis by Todd Lyons.
530 PP/17 Update configure.default to handle IPv6 localhost better.
531 Patch by Alain Williams (plus minor tweaks).
534 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
535 This is now consistent with GnuTLS, and is now documented: the
536 previous undocumented portable approach to treating the option as
537 unset was to force an expansion failure. That still works, and
538 an empty string is now equivalent.
540 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
541 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
542 not performing validation itself.
544 PP/20 Added force_command boolean option to pipe transport.
545 Patch from Nick Koston, of cPanel Inc.
547 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
550 TF/04 Added udpsend ACL modifer and hexquote expansion operator
552 PP/21 Fix eximon continuous updating with timestamped log-files.
553 Broken in a format-string cleanup in 4.80, missed when I repaired the
554 other false fix of the same issue.
555 Report and fix from Heiko Schlichting.
558 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
559 Report from Prashanth Katuri.
561 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
562 It's SecureTransport, so affects any MacOS clients which use the
563 system-integrated TLS libraries, including email clients.
565 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
566 using a MIME ACL for non-SMTP local injection.
567 Report and assistance in diagnosis by Warren Baker.
569 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
571 JH/16 Fix comparisons for 64b. Bugzilla 1385.
573 TL/09 Add expansion variable $authenticated_fail_id to keep track of
574 last id that failed so it may be referenced in subsequent ACL's.
576 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
579 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
580 ldap library initialization, allowing self-signed CA's to be
581 used. Also properly sets require_cert option later in code by
582 using NULL (global ldap config) instead of ldap handle (per
583 session). Bug diagnosis and testing by alxgomz.
585 TL/12 Enhanced documentation in the ratelimit.pl script provided in
586 the src/util/ subdirectory.
588 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
589 renamed to Transport Post Delivery Action by Jeremy Harris, as
592 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
593 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
594 redis_servers = needs to be configured which will be used by the redis
595 lookup. Patch from Warren Baker, of The Packet Hub.
597 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
599 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
600 hostname or reverse DNS when processing a host list. Used suggestions
601 from multiple comments on this bug.
603 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
605 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
606 Missed a few lines, added it to make the runtest require no keyboard
609 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
610 contains upper case chars. Make router use caseful_local_part.
612 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
613 support when GnuTLS has been built with p11-kit.
619 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
621 This, or similar/improved, will also be change PP/11 of 4.82.
627 PP/01 Handle short writes when writing local log-files.
628 In practice, only affects FreeBSD (8 onwards).
629 Bugzilla 1053, with thanks to Dmitry Isaikin.
631 NM/01 Bugzilla 949 - Documentation tweak
633 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
636 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
638 PP/02 Implemented gsasl authenticator.
640 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
642 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
643 `pkg-config foo` for cflags/libs.
645 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
646 with rest of GSASL and with heimdal_gssapi.
648 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
649 `pkg-config foo` for cflags/libs for the TLS implementation.
651 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
652 properties get this fed in as external SSF. A number of robustness
653 and debugging improvements to the cyrus_sasl authenticator.
655 PP/08 cyrus_sasl server now expands the server_realm option.
657 PP/09 Bugzilla 1214 - Log authentication information in reject log.
658 Patch by Jeremy Harris.
660 PP/10 Added dbmjz lookup type.
662 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
664 PP/12 MAIL args handles TAB as well as SP, for better interop with
665 non-compliant senders.
666 Analysis and variant patch by Todd Lyons.
668 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
669 Bug report from Lars Müller <lars@samba.org> (via SUSE),
670 Patch from Dirk Mueller <dmueller@suse.com>
672 PP/13 tls_peerdn now print-escaped for spool files.
673 Observed some $tls_peerdn in wild which contained \n, which resulted
674 in spool file corruption.
676 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
677 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
678 or write after TLS renegotiation, which otherwise led to messages
681 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
682 as a tracking header (ie: a signed header comes before the signature).
683 Patch from Wolfgang Breyha.
685 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
686 comma-sep list; embedded commas doubled.
688 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
690 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
692 Report and patch from Dmitry Banschikov.
694 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
695 Removed SSL_clear() after SSL_new() which led to protocol negotiation
696 failures. We appear to now support TLS1.1+ with Exim.
698 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
699 lets Exim select keys and certificates based upon TLS SNI from client.
700 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
701 before an outbound SMTP session. New log_selector, +tls_sni.
703 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
704 NULL dereference. Report and patch from Alun Jones.
706 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
707 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
708 Not seeing resolver debug output on NetBSD, but suspect this is a
709 resolver implementation change.
711 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
712 Left warnings. Added "eximon gdb" invocation mode.
714 PP/21 Defaulting "accept_8bitmime" to true, not false.
716 PP/22 Added -bw for inetd wait mode support.
718 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
719 locate the relevant includes and libraries. Made this the default.
721 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
722 Bugzilla 1246, report and most of solution from Tomasz Kusy.
724 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
725 This may cause build issues on older platforms.
727 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
728 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
729 gnutls_require_mac & gnutls_require_protocols (no longer supported).
730 Added SNI support via GnuTLS too.
731 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
733 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
735 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
736 Applied second patch from Janne, implementing suggestion to default
737 multiple-strings-in-record handling to match SPF spec.
739 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
741 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
742 read-only, out of scope).
743 Patch from Wolfgang Breyha, report from Stuart Northfield.
745 PP/29 Fix three issues highlighted by clang analyser static analysis.
746 Only crash-plausible issue would require the Cambridge-specific
747 iplookup router and a misconfiguration.
748 Report from Marcin Mirosław.
750 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
752 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
753 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
754 As part of this, removing so much warning spew let me fix some minor
755 real issues in debug logging.
757 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
758 assignment on my part. Fixed.
760 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
761 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
762 Janne Snabb (who went above and beyond: thank you).
764 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
765 string otherwise requires a connection and a bunch more work and it's
766 relatively easy to get wrong. Should also expose TLS library linkage
769 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
770 64-bit ${eval} (JH/03).
772 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
773 GNU libc to support some of the 64-bit stuff, should not lead to
774 conflicts. Defined before os.h is pulled in, so if a given platform
775 needs to override this, it can.
777 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
778 protection layer was required, which is not implemented.
779 Bugzilla 1254, patch from Wolfgang Breyha.
781 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
782 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
783 tls_dhparam take prime identifiers. Also unbreak combination of
784 OpenSSL+DH_params+TLSSNI.
786 PP/39 Disable SSLv2 by default in OpenSSL support.
792 PP/01 Solaris build fix for Oracle's LDAP libraries.
793 Bugzilla 1109, patch from Stephen Usher.
795 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
797 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
800 TF/02 Fix a couple more cases where we did not log the error message
801 when unlink() failed. See also change 4.74-TF/03.
803 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
804 lock up or crash if it happened to be inside a call to libc when it
805 got a SIGUSR1 from exiwhat.
807 The SIGUSR1 handler appends the current process status to the process
808 log which is later printed by exiwhat. It used to use the general
809 purpose logging code to do this, but several functions it calls are
810 not safe for signals.
812 The new output code in the SIGUSR1 handler is specific to the process
813 log, and simple enough that it's easy to inspect for signal safety.
814 Removing some special cases also simplifies the general logging code.
815 Removing the spurious timestamps from the process log simplifies
818 TF/04 Improved ratelimit ACL condition.
820 The /noupdate option has been deprecated in favour of /readonly which
821 has clearer semantics. The /leaky, /strict, and /readonly update modes
822 are mutually exclusive. The update mode is no longer included in the
823 database key; it just determines when the database is updated. (This
824 means that when you upgrde Exim will forget old rate measurements.)
826 Exim now checks that the per_* options are used with an update mode that
827 makes sense for the current ACL. For example, when Exim is processing a
828 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
829 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
830 must specify per_mail/readonly. If you omit the update mode it defaults to
831 /leaky where that makes sense (as before) or /readonly where required.
833 The /noupdate option is now undocumented but still supported for
834 backwards compatibility. It is equivalent to /readonly except that in
835 ACLs where /readonly is required you may specify /leaky/noupdate or
836 /strict/noupdate which are treated the same as /readonly.
838 A useful new feature is the /count= option. This is a generalization
839 of the per_byte option, so that you can measure the throughput of other
840 aggregate values. For example, the per_byte option is now equivalent
841 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
843 The per_rcpt option has been generalized using the /count= mechanism
844 (though it's more complicated than the per_byte equivalence). When it is
845 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
846 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
847 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
848 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
849 Note that using per_rcpt with a non-readonly update mode in more than
850 one ACL will cause the recipients to be double-counted. (The per_mail
851 and per_byte options don't have this problem.)
853 The handling of very low rates has changed slightly. If the computed rate
854 is less than the event's count (usually one) then this event is the first
855 after a long gap. In this case the rate is set to the same as this event's
856 count, so that the first message of a spam run is counted properly.
858 The major new feature is a mechanism for counting the rate of unique
859 events. The new per_addr option counts the number of different
860 recipients that someone has sent messages to in the last time period. It
861 behaves like per_rcpt if all the recipient addresses are different, but
862 duplicate recipient addresses do not increase the measured rate. Like
863 the /count= option this is a general mechanism, so the per_addr option
864 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
865 example, measure the rate that a client uses different sender addresses
866 with the options per_mail/unique=$sender_address. There are further
867 details in the main documentation.
869 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
871 TF/06 Removed a few PCRE remnants.
873 TF/07 Automatically extract Exim's version number from tags in the git
874 repository when doing development or release builds.
876 PP/02 Raise smtp_cmd_buffer_size to 16kB.
877 Bugzilla 879. Patch from Paul Fisher.
879 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
880 Heavily based on revision 40f9a89a from Simon Arlott's tree.
883 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
885 PP/05 Variable $av_failed, true if the AV scanner deferred.
886 Bugzilla 1078. Patch from John Horne.
888 PP/06 Stop make process more reliably on build failure.
889 Bugzilla 1087. Patch from Heiko Schlittermann.
891 PP/07 Make maildir_use_size_file an _expandable_ boolean.
892 Bugzilla 1089. Patch from Heiko Schlittermann.
894 PP/08 Handle ${run} returning more data than OS pipe buffer size.
895 Bugzilla 1131. Patch from Holger Weiß.
897 PP/09 Handle IPv6 addresses with SPF.
898 Bugzilla 860. Patch from Wolfgang Breyha.
900 PP/10 GnuTLS: support TLS 1.2 & 1.1.
902 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
905 PP/11 match_* no longer expand right-hand-side by default.
906 New compile-time build option, EXPAND_LISTMATCH_RHS.
907 New expansion conditions, "inlist", "inlisti".
909 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
911 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
913 PP/14 fix log_write() format string regression from TF/03.
914 Bugzilla 1152. Patch from Dmitry Isaikin.
920 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
922 PP/02 Harmonised TLS library version reporting; only show if debugging.
923 Layout now matches that introduced for other libraries in 4.74 PP/03.
925 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
927 PP/04 New "dns_use_edns0" global option.
929 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
932 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
933 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
935 TK/01 Updated PolarSSL code to 0.14.2.
936 Bugzilla 1097. Patch from Andreas Metzler.
938 PP/07 Catch divide-by-zero in ${eval:...}.
941 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
944 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
945 format-string attack -- SECURITY: remote arbitrary code execution.
947 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
948 time unintentionally subject to list matching rules, letting the header
949 cause arbitrary Exim lookups (of items which can occur in lists, *not*
950 arbitrary string expansion). This allowed for information disclosure.
952 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
953 INT_MIN/-1 -- value coerced to INT_MAX.
959 NM/01 Workround for PCRE version dependency in version reporting
962 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
963 This fixes portability to compilers other than gcc, notably
964 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
966 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
967 makefiles for portability to HP-UX and POSIX correctness.
969 PP/01 Permit LOOKUP_foo enabling on the make command-line.
970 Also via indented variable definition in the Makefile.
971 (Debugging by Oliver Heesakkers).
973 PP/02 Restore caching of spamd results with expanded spamd_address.
974 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
976 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
977 Improves build reliability. Fix from: Frank Elsner
979 NM/02 Fix wide character breakage in the rfc2047 coding
980 Fixes bug 1064. Patch from Andrey N. Oktyabrski
982 NM/03 Allow underscore in dnslist lookups
983 Fixes bug 1026. Patch from Graeme Fowler
985 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
986 Code patches from Adam Ciarcinski of NetBSD.
988 NM/04 Fixed exiqgrep to cope with mailq missing size issue
991 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
992 is logged, to avoid truncation. Patch from John Horne.
994 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
995 Patch from Jakob Hirsch.
997 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
998 SQL string expansion failure details.
999 Patch from Andrey Oktyabrski.
1001 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1002 Patch from Simon Arlott.
1004 PP/09 New lookups functionality failed to compile on old gcc which rejects
1005 extern declarations in function scope.
1006 Patch from Oliver Fleischmann
1008 PP/10 Use sig_atomic_t for flags set from signal handlers.
1009 Check getgroups() return and improve debugging.
1010 Fixed developed for diagnosis in bug 927 (which turned out to be
1013 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1014 Patch from Mark Zealey.
1016 PP/12 Bugzilla 1056: Improved spamd server selection.
1017 Patch from Mark Zealey.
1019 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1020 Based on patch from Heiko Schlittermann.
1022 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1023 Patch from Uwe Doering, sign-off by Michael Haardt.
1025 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1026 variable declaration deep within a block. Bug and patch from
1029 PP/15 lookups-Makefile IRIX compatibilty coercion.
1031 PP/16 Make DISABLE_DKIM build knob functional.
1033 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1034 Patch by Simon Arlott
1036 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1037 variable argument macros. Our copy now differs from upstream.
1043 TF/01 Failure to get a lock on a hints database can have serious
1044 consequences so log it to the panic log.
1046 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1047 controlled using the smtp_confirmation log selector.
1049 TF/03 Include the error message when we fail to unlink a spool file.
1051 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1052 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1053 for maintaining out-of-tree patches for some time.
1055 PP/01 Bugzilla 139: Documentation and portability issues.
1056 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1057 Handle per-OS dynamic-module compilation flags.
1059 PP/02 Let /dev/null have normal permissions.
1060 The 4.73 fixes were a little too stringent and complained about the
1061 permissions on /dev/null. Exempt it from some checks.
1062 Reported by Andreas M. Kirchwitz.
1064 PP/03 Report version information for many libraries, including
1065 Exim version information for dynamically loaded libraries. Created
1066 version.h, now support a version extension string for distributors
1067 who patch heavily. Dynamic module ABI change.
1069 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1070 privilege escalation vulnerability whereby the Exim run-time user
1071 can cause root to append content of the attacker's choosing to
1074 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1077 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1078 If dropping privileges for untrusted macros, we disabled normal logging
1079 on the basis that it would fail; for the Exim run-time user, this is not
1080 the case, and it resulted in successful deliveries going unlogged.
1081 Fixed. Reported by Andreas Metzler.
1087 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1088 only prepend for the Resent-* case. Fixes regression introduced in
1089 Exim 4.70 by NM/22 for Bugzilla 607.
1091 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1092 increasing numbers of administrators be bitten by this.
1094 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1096 PP/03 Bugzilla 994: added openssl_options main configuration option.
1098 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1100 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1102 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1104 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1105 without a peer certificate, leading to a segfault because of an
1106 assumption that peers always have certificates. Be a little more
1107 paranoid. Problem reported by Martin Tscholak.
1109 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1110 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1111 NB: ClamAV planning to remove STREAM in "middle of 2010".
1112 CL also introduces -bmalware, various -d+acl logging additions and
1113 more caution in buffer sizes.
1115 PP/09 Implemented reverse_ip expansion operator.
1117 PP/10 Bugzilla 937: provide a "debug" ACL control.
1119 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1121 PP/12 Bugzilla 973: Implement --version.
1123 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1125 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1127 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1129 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1130 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1131 ignore trailing whitespace.
1133 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1135 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1138 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1139 Notification from Dr Andrew Aitchison.
1141 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1142 ExtendedDetectionInfo response format.
1143 Notification from John Horne.
1145 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1148 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1149 XSL and documented dependency on system catalogs, with examples of how
1152 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1155 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1156 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1157 configuration file which is writeable by the Exim user or group.
1159 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1160 of configuration files to cover files specified with the -C option if
1161 they are going to be used with root privileges, not just the default
1164 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1165 option (effectively making it always true).
1167 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1168 files to be used while preserving root privileges.
1170 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1171 that rogue child processes cannot use them.
1173 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1174 run-time user, instead of root.
1176 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1177 Exim run-time user without dropping privileges.
1179 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1180 result string, instead of calling string_vformat() twice with the same
1183 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1184 for other users. Others should always drop root privileges if they use
1185 -C on the command line, even for a whitelisted configure file.
1187 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1189 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1195 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1196 $data_path, and $header_path variables; fixed documentation bugs and
1199 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1200 exipick to access non-standard spools, including the "frozen" queue
1203 NM/01 Bugzilla 965: Support mysql stored procedures.
1204 Patch from Alain Williams
1206 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1208 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1209 Patch from Andreas Metzler
1211 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1212 Patch from Kirill Miazine
1214 NM/05 Bugzilla 671: Added umask to procmail example.
1216 JJ/03 installed exipick 20100323.0, fixing doc bug
1218 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1219 directory. Notification and patch from Dan Rosenberg.
1221 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1223 TK/02 Improve log output when DKIM signing operation fails.
1225 MH/01 Treat the transport option dkim_domain as a colon separated
1226 list, not as a single string, and sign the message with each element,
1227 omitting multiple occurences of the same signer.
1229 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1230 Bugzilla 985, 986. Patch by Simon Arlott
1232 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1233 Patch by Simon Arlott
1235 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1236 MBX locking. Notification from Dan Rosenberg.
1242 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1244 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1246 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1248 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1250 NM/04 Bugzilla 915: Flag broken perl installation during build.
1256 TK/01 Added patch by Johannes Berg that expands the main option
1257 "spamd_address" if it starts with a dollar sign.
1259 TK/02 Write list of recipients to X-Envelope-Sender header when building
1260 the mbox-format spool file for content scanning (suggested by Jakob
1263 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1264 (http://www.dcc-servers.net/) support via dccifd. Activated by
1265 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1267 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1268 by Mark Daniel Reidel <mr@df.eu>.
1270 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1271 When building exim an external PCRE library is now needed -
1272 PCRE is a system library on the majority of modern systems.
1273 See entry on PCRE_LIBS in EDITME file.
1275 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1276 conversation. Added nologin parameter to request.
1277 Patch contributed by Kirill Miazine.
1279 TF/01 Do not log submission mode rewrites if they do not change the address.
1281 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1283 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1284 log files in place. Contributed by Roberto Lima.
1286 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1288 TF/03 Bugzilla 615: When checking the local_parts router precondition
1289 after a local_part_suffix or local_part_prefix option, Exim now
1290 does not use the address's named list lookup cache, since this
1291 contains cached lookups for the whole local part.
1293 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1294 Robert Millan. Documentation is in experimental-spec.txt.
1296 TF/04 Bugzilla 668: Fix parallel build (make -j).
1298 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1300 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1301 Patch provided by Jan Srzednicki.
1303 TF/05 Leading white space used to be stripped from $spam_report which
1304 wrecked the formatting. Now it is preserved.
1306 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1307 that they are available at delivery time.
1309 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1311 TF/08 TLS error reporting now respects the incoming_interface and
1312 incoming_port log selectors.
1314 TF/09 Produce a more useful error message if an SMTP transport's hosts
1315 setting expands to an empty string.
1317 NM/06 Bugzilla 744: EXPN did not work under TLS.
1318 Patch provided by Phil Pennock.
1320 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1321 Patch provided by Richard Godbee.
1323 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1324 acl_smtp_notquit, added index entry.
1326 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1327 Patch provided by Eugene Bujak.
1329 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1330 accept(). Patch provided by Maxim Dounin.
1332 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1333 Patch provided by Phil Pennock.
1335 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1337 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1338 Patch provided by Brad "anomie" Jorsch.
1340 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1341 Patch provided by Dean Brooks.
1343 TK/05 Add native DKIM support (does not depend on external libraries).
1345 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1346 Patch provided by Graeme Fowler.
1348 NM/16 Bugzilla 851: Documentation example syntax fix.
1350 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1352 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1355 NM/19 Bugzilla 745: TLS version reporting.
1356 Patch provided by Phil Pennock.
1358 NM/20 Bugzilla 167: bool: condition support.
1359 Patch provided by Phil Pennock.
1361 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1362 clients. Patch provided by Phil Pennock.
1364 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1365 Patch provided by Brad "anomie" Jorsch.
1367 NM/23 Bugzilla 687: Fix misparses in eximstats.
1368 Patch provided by Heiko Schlittermann.
1370 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1371 Patch provided by Heiko Schlittermann.
1373 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1374 plus update to original patch.
1376 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1378 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1379 Patch provided by David Brownlee.
1381 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1383 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1385 NM/30 Bugzilla 888: TLS documentation bugfixes.
1387 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1389 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1390 Unlike the original bugzilla I have changed all shell scripts in src tree.
1392 NM/33 Bugzilla 898: Transport filter timeout fix.
1393 Patch by Todd Rinaldo.
1395 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1396 Patch by Serge Demonchaux.
1398 NM/35 Bugzilla 39: Base64 decode bug fixes.
1399 Patch by Jakob Hirsch.
1401 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1403 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1405 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1407 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1413 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1414 ALT-N's libdkim that I have put here:
1415 http://duncanthrax.net/exim-experimental/
1417 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1418 were called 'true' and it seems that C99 defines that as a reserved
1419 keyword to be used with 'bool' variable types. That means you could
1420 not include C99-style headers which use bools without triggering
1421 build errors in sieve.c.
1423 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1424 as mailq or other aliases. Changed the --help handling significantly
1425 to do whats expected. exim_usage() emits usage/help information.
1427 SC/01 Added the -bylocaldomain option to eximstats.
1429 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1431 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1433 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1439 PH/01 Another patch from the Sieve maintainer.
1441 PH/02 When an IPv6 address is converted to a string for single-key lookup
1442 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1443 dots are used instead of colons so that keys in lsearch files need not
1444 contain colons. This was done some time before quoting was made available
1445 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1446 (notated using the quote facility) so as to distinguish them from IPv4
1447 keys. This meant that lookups for IP addresses in host lists did not work
1448 for iplsearch lookups.
1450 This has been fixed by arranging for IPv6 addresses to be expressed with
1451 colons if the lookup type is iplsearch. This is not incompatible, because
1452 previously such lookups could never work.
1454 The situation is now rather anomolous, since one *can* have colons in
1455 ordinary lsearch keys. However, making the change in all cases is
1456 incompatible and would probably break a number of configurations.
1458 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1461 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1462 conversion specification without a maximum field width, thereby enabling
1463 a rogue spamd server to cause a buffer overflow. While nobody in their
1464 right mind would setup Exim to query an untrusted spamd server, an
1465 attacker that gains access to a server running spamd could potentially
1466 exploit this vulnerability to run arbitrary code as the Exim user.
1468 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1469 $primary_hostname instead of what libspf2 thinks the hosts name is.
1471 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1472 a directory entry by the name of the lookup key. Previously, if a
1473 symlink pointed to a non-existing file or a file in a directory that
1474 Exim lacked permissions to read, a lookup for a key matching that
1475 symlink would fail. Now it is enough that a matching directory entry
1476 exists, symlink or not. (Bugzilla 503.)
1478 PH/03 The body_linecount and body_zerocount variables are now exported in the
1481 PH/04 Added the $dnslist_matched variable.
1483 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1484 This means they are set thereafter only if the connection becomes
1487 PH/06 Added the client_condition to authenticators so that some can be skipped
1488 by clients under certain conditions.
1490 PH/07 The error message for a badly-placed control=no_multiline_responses left
1491 "_responses" off the end of the name.
1493 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1495 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1496 (without spaces) instead of just copying the configuration text.
1498 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1500 PH/11 Added $max_received_linelength.
1502 PH/12 Added +ignore_defer and +include_defer to host lists.
1504 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1505 way in which PCRE > 7.0 is built.
1507 PH/14 Implemented queue_only_load_latch.
1509 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1510 MAIL command. The effect was to mangle the value on 64-bit systems.
1512 PH/16 Another patch from the Sieve maintainer.
1514 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1516 PH/18 If a system quota error occurred while trying to create the file for
1517 a maildir delivery, the message "Mailbox is full" was not appended to the
1518 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1519 only to a quota excession during the actual writing of the file.
1521 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1522 characters?) which causes problems in log lines. The DN values are now
1523 passed through string_printing() before being added to log lines.
1525 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1526 and InterBase are left for another time.)
1528 PH/21 Added message_body_newlines option.
1530 PH/22 Guard against possible overflow in moan_check_errorcopy().
1532 PH/23 POSIX allows open() to be a macro; guard against that.
1534 PH/24 If the recipient of an error message contained an @ in the local part
1535 (suitably quoted, of course), incorrect values were put in $domain and
1536 $local_part during the evaluation of errors_copy.
1542 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1543 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1546 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1547 issue a MAIL command.
1549 PH/02 In an ACL statement such as
1551 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1553 if a client was not listed at all, or was listed with a value other than
1554 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1555 the condition was not true (as it should be), so access was not denied.
1556 The bug was that the ! inversion was incorrectly passed on to the second
1557 item. This has been fixed.
1559 PH/03 Added additional dnslists conditions == and =& which are different from
1560 = and & when the dns lookup returns more than one IP address.
1562 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1563 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1565 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1566 FSYNC, which compiles an option called disable_fsync that allows for
1567 bypassing fsync(). The documentation is heavily laced with warnings.
1569 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1571 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1572 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1573 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1574 including adding "make clean"; (3) Added -fPIC when compiling the test
1575 dynamically loaded module, to get rid of a warning.
1577 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1578 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1579 = 0s. The bug is otherwise harmless.
1581 PH/07 There was a bug in the dovecot authenticator such that the value of
1582 $auth1 could be overwritten, and so not correctly preserved, after a
1583 successful authentication. This usually meant that the value preserved by
1584 the server_setid option was incorrect.
1586 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1588 PH/09 Installed PCRE release 7.0.
1590 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1591 run for batched SMTP input. It is now run at the start of every message
1592 in the batch. While fixing this I discovered that the process information
1593 (output by running exiwhat) was not always getting set for -bs and -bS
1594 input. This is fixed, and it now also says "batched" for BSMTP.
1596 PH/11 Added control=no_pipelining.
1598 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1599 patch, slightly modified), and move the expansion of helo_data till after
1600 the connection is made in the smtp transport (so it can use these
1603 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1605 PH/14 Added log_selector = +pid.
1607 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1609 PH/16 Add ${if forany and ${if forall.
1611 PH/17 Added dsn_from option to vary the From: line in DSNs.
1613 PH/18 Flush SMTP output before performing a callout, unless control =
1614 no_callout_flush is set.
1616 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1617 was true (the default) a successful delivery failed to delete the retry
1618 item, thus causing premature timeout of the address. The bug is now
1621 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1623 PH/21 Long custom messages for fakedefer and fakereject are now split up
1624 into multiline reponses in the same way that messages for "deny" and
1625 other ACL rejections are.
1627 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1628 with slight modification.
1630 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1631 draft, changing the syntax and factoring some duplicate code".
1633 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1634 for deliveries of the second and subsequent messages over the same SMTP
1637 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1638 ${reduce, with only minor "tidies".
1640 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1642 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1643 expansion side effects.
1645 PH/27 When a message times out after an over-quota error from an Exim-imposed
1646 quota, the bounce message says "mailbox is full". This message was not
1647 being given when it was a system quota that was exceeded. It now should
1650 MH/03 Made $recipients available in local_scan(). local_scan() already has
1651 better access to the recipient list through recipients_list[], but
1652 $recipients can be useful in postmaster-provided expansion strings.
1654 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1655 in the case of a MAIL command with additional options following the
1656 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1657 were accidentally chopped off.
1659 PH/29 SMTP synchronization checks are implemented when a command is read -
1660 there is a check that no more input is waiting when there shouldn't be
1661 any. However, for some commands, a delay in an ACL can mean that it is
1662 some time before the response is written. In this time, more input might
1663 arrive, invalidly. So now there are extra checks after an ACL has run for
1664 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1665 pipelining has not been advertised.
1667 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1669 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1670 returned in the SMTP response when smtp_return_error_details was false.
1671 This has been fixed.
1673 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1674 instead of the C I/O that was originally supplied, because problems were
1675 reported on Solaris.
1677 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1678 Exim which did not show up earlier: it was assuming that a call to
1679 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1680 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1681 was a macro that became an assignment, so it seemed to work. This has
1682 changed to a proper function call with a void return, hence the compile
1683 error. Exim's code has been fixed.
1685 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1688 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1690 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1692 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1693 because it turns out to be too big or there is a timeout) there is no
1694 "Completed" line in the log. When some messages of this type were
1695 selected by exigrep, they were listed as "not completed". Others were
1696 picked up by some special patterns. I have improved the selection
1697 criteria to be more general.
1699 PH/38 The host_find_failed option in the manualroute router can now be set
1700 to "ignore", to completely ignore a host whose IP address cannot be
1701 found. If all hosts are ignored, the behaviour is controlled by the new
1702 host_all_ignored option.
1704 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1705 homing or because of multiple MX records with /mx) generated more than
1706 one IP address, and the following item turned out to be the local host,
1707 all the secondary addresses of the first item were incorrectly removed
1708 from the list, along with the local host and any following hosts (which
1709 is what is supposed to happen).
1711 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1712 whoever called Exim into the -H file. In the case of the daemon it was
1713 behaving confusingly. When first started, it used values for whoever
1714 started the daemon, but after a SIGHUP it used the Exim user (because it
1715 calls itself on a restart). I have changed the code so that it now always
1718 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1719 message are rejected with the same error (e.g. no authentication or bad
1720 sender address), and a DATA command is nevertheless sent (as can happen
1721 with PIPELINING or a stupid MUA), the error message that was given to the
1722 RCPT commands is included in the rejection of the DATA command. This is
1723 intended to be helpful for MUAs that show only the final error to their
1726 PH/42 Another patch from the Sieve maintainer.
1728 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1729 Eximstats - Fixed some broken HTML links and added missing column headers
1731 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1732 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1734 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1736 PH/43 Yet another patch from the Sieve maintainer.
1738 PH/44 I found a way to check for a TCP/IP connection going away before sending
1739 the response to the final '.' that terminates a message, but only in the
1740 case where the client has not sent further data following the '.'
1741 (unfortunately, this is allowed). However, in many cases there won't be
1742 any further data because there won't be any more messages to send. A call
1743 to select() can be used: if it shows that the input is "ready", there is
1744 either input waiting, or the socket has been closed. An attempt to read
1745 the next input character can distinguish the two cases. Previously, Exim
1746 would have sent an OK response which the client would never have see.
1747 This could lead to message repetition. This fix should cure that, at
1748 least in a lot of common cases.
1750 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1751 advertised in response to EHLO.
1757 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1758 fixed by 4.65/MH/01 (is this a record?) are fixed:
1760 (i) An empty string was always treated as zero by the numeric comparison
1761 operators. This behaviour has been restored.
1763 (ii) It is documented that the numeric comparison operators always treat
1764 their arguments as decimal numbers. This was broken in that numbers
1765 starting with 0 were being interpreted as octal.
1767 While fixing these problems I realized that there was another issue that
1768 hadn't been noticed. Values of message_size_limit (both the global option
1769 and the transport option) were treated as octal if they started with 0.
1770 The documentation was vague. These values are now always treated as
1771 decimal, and I will make that clear in the documentation.
1777 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1778 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1781 MH/01 Don't check that the operands of numeric comparison operators are
1782 integers when their expansion is in "skipping" mode (fixes bug
1783 introduced by 4.64-PH/07).
1785 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1786 child addresses, Exim now panics and dies. Previously, because the count
1787 is held in a short int, deliveries were likely to be lost. As such a
1788 large number of recipients for a single message is ridiculous
1789 (performance will be very, very poor), I have chosen to impose a limit
1790 rather than extend the field.
1796 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1797 leftover -K file (the existence of which was triggered by #402).
1798 While we were at it, introduced process PID as part of the -K
1799 filename. This should rule out race conditions when creating
1802 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1803 processing considerably. Previous code took too long for large mails,
1804 triggering a timeout which in turn triggers #401.
1806 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1807 in the DK code in transports.c. sendfile() is not really portable,
1808 hence the _LINUX specificness.
1810 TF/01 In the add_headers option to the mail command in an Exim filter,
1811 there was a bug that Exim would claim a syntax error in any
1812 header after the first one which had an odd number of characters
1815 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1816 callout verification, Exim cached a "reject" for the entire domain. This
1817 is correct for most verifications, but it is not correct for a recipient
1818 verification with use_sender or use_postmaster set, because in that case
1819 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1820 case of MAIL FROM:<> rejection from other early rejections (e.g.
1821 rejection of HELO). When verifying a recipient using a non-null MAIL
1822 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1823 Whatever the result of the callout, the value of the domain cache is
1824 left unchanged (for any other kind of callout, getting as far as trying
1825 RCPT means that the domain itself is ok).
1827 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1830 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1831 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1832 session, the daemon ignores ECONNECT errors and logs others; it now
1833 ignores EPIPE as well.
1835 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1836 (quoted-printable decoding).
1838 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1839 later the small subsequent patch to fix an introduced bug.
1841 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1843 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1845 PH/08 An error is now given if message_size_limit is specified negative.
1847 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1848 to be given (somewhat) arbitrary names.
1850 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1853 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1854 miscellaneous code fixes
1856 PH/10 Added the log_reject_target ACL modifier to specify where to log
1859 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1860 hostname. This is wrong, because it relates to the incoming message (and
1861 probably the interface on which it is arriving) and not to the outgoing
1862 callout (which could be using a different interface). This has been
1863 changed to use the value of the helo_data option from the smtp transport
1864 instead - this is what is used when a message is actually being sent. If
1865 there is no remote transport (possible with a router that sets up host
1866 addresses), $smtp_active_hostname is used.
1868 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1869 tweaks were necessary in order to get it to work (see also 21 below):
1870 (a) The code assumed that strncpy() returns a negative number on buffer
1871 overflow, which isn't the case. Replaced with Exim's string_format()
1873 (b) There were several signed/unsigned issues. I just did the minimum
1874 hacking in of casts. There is scope for a larger refactoring.
1875 (c) The code used strcasecmp() which is not a standard C function.
1876 Replaced with Exim's strcmpic() function.
1877 (d) The code set only $1; it now sets $auth1 as well.
1878 (e) A simple test gave the error "authentication client didn't specify
1879 service in request". It would seem that Dovecot has changed its
1880 interface. Fortunately there's a specification; I followed it and
1881 changed what the client sends and it appears to be working now.
1883 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1886 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1887 address A is aliased to B and C, where B exists and C does not. Without
1888 -v the output is "A verified" because verification stops after a
1889 successful redirection if more than one address is generated. However,
1890 with -v the child addresses are also verified. Exim was outputting "A
1891 failed to verify" and then showing the successful verification for C,
1892 with its parentage. It now outputs "B failed to verify", showing B's
1893 parentage before showing the successful verification of C.
1895 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1896 look up a TXT record in a specific list after matching in a combined
1899 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1900 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1901 they consult the DNS. I had assumed they would set it the way they
1902 wanted; and indeed my experiments on Linux seem to show that in some
1903 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1904 To be on the safe side, however, I have now made the interface to
1905 host_find_byname() similar to host_find_bydns(), with an argument
1906 containing the DNS resolver options. The host_find_byname() function now
1907 sets these options at its start, just as host_find_bydns() does. The smtp
1908 transport options dns_qualify_single and dns_search_parents are passed to
1909 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1910 of host_find_byname() use the default settings of RES_DEFNAMES
1911 (qualify_single) but not RES_DNSRCH (search_parents).
1913 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1914 spool_read_header() do less string testing, by means of a preliminary
1915 switch on the second character of optional "-foo" lines. (This is
1916 overdue, caused by the large number of possibilities that now exist.
1917 Originally there were few.) While I was there, I also converted the
1918 str(n)cmp tests so they don't re-test the leading "-" and the first
1919 character, in the hope this might squeeze out yet more improvement.
1921 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1922 flag allowing group syntax was set by the header_syntax check but not
1923 turned off, possible causing trouble later; (2) The flag was not being
1924 set at all for the header_verify test, causing "group"-style headers to
1925 be rejected. I have now set it in this case, and also caused header_
1926 verify to ignore an empty address taken from a group. While doing this, I
1927 came across some other cases where the code for allowing group syntax
1928 while scanning a header line wasn't quite right (mostly, not resetting
1929 the flag correctly in the right place). These bugs could have caused
1930 trouble for malformed header lines. I hope it is now all correct.
1932 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1933 with the "reply" argument non-NULL. The code, however (which originally
1934 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1935 but it didn't always do it. This confused somebody who was copying the
1936 code for some other use. I have removed all the tests.
1938 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1939 feature that was used to support insecure browsers during the U.S. crypto
1940 embargo. It requires special client support, and Exim is probably the
1941 only MTA that supported it -- and would never use it because real RSA is
1942 always available. This code has been removed, because it had the bad
1943 effect of slowing Exim down by computing (never used) parameters for the
1944 RSA_EXPORT functionality.
1946 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1947 authenticator to fail if there's a tab character in the incoming data
1948 (there should never be unless someone is messing about, as it's supposed
1949 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1950 if the connection is using TLS or if the remote IP is the same as the
1951 local IP, and the "valid-client-cert option" if a client certificate has
1954 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1955 authenticators. This can be used for authorization after authentication
1956 succeeds. (In the case of plaintext, it servers for both authentication
1959 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1960 if any retry times were supplied.
1962 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1963 connection, where there is no client IP address to check. In this
1964 situation, the verify now always succeeds.
1966 PH/25 Applied John Jetmore's -Mset patch.
1968 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1970 PH/27 In a string expansion for a processed (not raw) header when multiple
1971 headers of the same name were present, leading whitespace was being
1972 removed from all of them, but trailing whitespace was being removed only
1973 from the last one. Now trailing whitespace is removed from each header
1974 before concatenation. Completely empty headers in a concatenation (as
1975 before) are ignored.
1977 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1978 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1980 PH/29 [Removed. This was a change that I later backed out, and forgot to
1981 correct the ChangeLog entry (that I had efficiently created) before
1982 committing the later change.]
1984 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1985 address errors (4xx response to RCPT) over the same connection as other
1986 messages routed to the same hosts. Such deliveries are always "forced",
1987 so retry times are not inspected. This resulted in far too many retries
1988 for the affected addresses. The effect occurred only when there were more
1989 hosts than the hosts_max_try setting in the smtp transport when it had
1990 the 4xx errors. Those hosts that it had tried were not added to the list
1991 of hosts for which the message was waiting, so if all were tried, there
1992 was no problem. Two fixes have been applied:
1994 (i) If there are any address or message errors in an SMTP delivery, none
1995 of the hosts (tried or untried) are now added to the list of hosts
1996 for which the message is waiting, so the message should not be a
1997 candidate for sending over the same connection that was used for a
1998 successful delivery of some other message. This seems entirely
1999 reasonable: after all the message is NOT "waiting for some host".
2000 This is so "obvious" that I'm not sure why it wasn't done
2001 previously. Hope I haven't missed anything, but it can't do any
2002 harm, as the worst effect is to miss an optimization.
2004 (ii) If, despite (i), such a delivery is accidentally attempted, the
2005 routing retry time is respected, so at least it doesn't keep
2006 hammering the server.
2008 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2009 in ${readsocket because some servers need this prod.
2011 PH/32 Added some extra debug output when updating a wait-xxx database.
2013 PH/33 The hint "could be header name not terminated by colon", which has been
2014 given for certain expansion errors for a long time, was not being given
2015 for the ${if def:h_colon_omitted{... case.
2017 PH/34 The spec says: "With one important exception, whenever a domain list is
2018 being scanned, $domain contains the subject domain." There was at least
2019 one case where this was not true.
2021 PH/35 The error "getsockname() failed: connection reset by peer" was being
2022 written to the panic log as well as the main log, but it isn't really
2023 panic-worthy as it just means the connection died rather early on. I have
2024 removed the panic log writing for the ECONNRESET error when getsockname()
2027 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2028 runs only) independently of the message's sender address. This meant
2029 that, if the 4xx error was in fact related to the sender, a different
2030 message to the same recipient with a different sender could confuse
2031 things. In particualar, this can happen when sending to a greylisting
2032 server, but other circumstances could also provoke similar problems.
2033 I have changed the default so that the retry time for these errors is now
2034 based a combination of the sender and recipient addresses. This change
2035 can be overridden by setting address_retry_include_sender=false in the
2038 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2039 remote server are returned as part of bounce messages. This was not
2040 happening for LMTP over a pipe (the lmtp transport), but now it is the
2041 same for both kinds of LMTP.
2043 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2044 in header lines that were in fact CNAMEs. This is no longer the case.
2046 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2047 and queue runs started by the daemon processed all messages. This has
2048 been fixed so that -R and -S can now usefully be given with -q<time>.
2050 PH/40 Import PCRE release 6.7 (fixes some bugs).
2052 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2054 PH/42 Give an error if -q is specified more than once.
2056 PH/43 Renamed the variables $interface_address and $interface_port as
2057 $received_ip_address and $received_port, to make it clear that these
2058 values apply to message reception, and not to the outgoing interface when
2059 a message is delivered. (The old names remain recognized, of course.)
2061 PH/44 There was no timeout on the connect() call when using a Unix domain
2062 socket in the ${readsocket expansion. There now is.
2064 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2065 be meaningful with "accept".
2067 SC/01 Eximstats V1.43
2068 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2070 SC/02 Eximstats V1.44
2071 Use a glob alias rather than an array ref in the generated
2072 parser. This improves both readability and performance.
2074 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2075 Collect SpamAssassin and rejection statistics.
2076 Don't display local sender or destination tables unless
2077 there is data to show.
2078 Added average volumes into the top table text output.
2080 SC/04 Eximstats V1.46
2081 Collect data on the number of addresses (recipients)
2082 as well as the number of messages.
2084 SC/05 Eximstats V1.47
2085 Added 'Message too big' to the list of mail rejection
2086 reasons (thanks to Marco Gaiarin).
2088 SC/06 Eximstats V1.48
2089 Mainlog lines which have GMT offsets and are too short to
2090 have a flag are now skipped.
2092 SC/07 Eximstats V1.49 (Alain Williams)
2093 Added the -emptyok flag.
2095 SC/08 Eximstats V1.50
2096 Fixes for obtaining the IP address from reject messages.
2098 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2099 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2100 whitesspace changes from 4.64-PH/27
2102 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2105 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2106 are found, allow negative numbers in numeric criteria)
2108 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2110 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2111 to match changes made in 4.64-PH/43
2113 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2115 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2116 "domain" is actually an IP address, and force a failure. This locks out
2117 those revolvers/nameservers that support "A-for-A" lookups, in
2118 contravention of the specifications.
2120 PH/48 When a host name was looked up from an IP address, and the subsequent
2121 forward lookup of the name timed out, the host name was left in
2122 $sender_host_name, contrary to the specification.
2124 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2125 restricted to single-key lookups, Exim was not diagnosing an error if
2126 * or *@ was used with a query-style lookup.
2128 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2130 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2131 long ago, but noone interested enough thought of it. Let's just say that
2132 the "1.1" means that there are some new functions that weren't there at
2133 some point in the past.
2135 PH/51 Error processing for expansion failure of helo_data from an smtp
2136 transport during callout processing was broken.
2138 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2139 tested/used via the -bh/-bhc/-bs options.
2141 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2142 bug, fixed in subsequent PCRE releases).
2144 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2145 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2147 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2153 SC/01 Use a glob alias rather than an array ref in eximstats generated
2154 parser. This improves both readability and performance.
2156 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2157 Don't display local sender or destination tables in eximstats unless
2158 there is data to show.
2159 Added average volumes into the eximstats top table text output.
2161 SC/03 Collect data on the number of addresses (recipients) as well
2162 as the number of messages in eximstats.
2164 TF/01 Correct an error in the documentation for the redirect router. Exim
2165 does (usually) call initgroups() when daemonizing.
2167 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2168 with consistent privilege compared to when running as a daemon.
2170 TF/03 Note in the spec that $authenticated_id is not set for local
2171 submissions from trusted users.
2173 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2174 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2176 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2177 by adding some example configuration directives to the default
2178 configuration file. A little bit of work is required to uncomment the
2179 directives and define how usernames and passwords are checked, but
2180 there is now a framework to start from.
2182 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2183 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2184 without this. I don't know how relevant this is to other LDAP libraries.
2186 PH/02 Add the verb name to the "unknown ACL verb" error.
2188 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2190 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2192 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2193 directory not expanded when it should be if an expanded home directory
2194 was set for the address (which is overridden by the transport).
2196 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2199 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2200 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2201 because it is too late at that time, and has no effect.
2203 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2204 security issue with \' (bugzilla #107). I could not use the
2205 PQescapeStringConn() function, because it needs a PGconn value as one of
2208 PH/08 When testing addresses using -bt, indicate those final addresses that
2209 are duplicates that would not cause an additional delivery. At least one
2210 person was confused, thinking that -bt output corresponded to deliveries.
2211 (Suppressing duplicates isn't a good idea as you lose the information
2212 about possibly different redirections that led to the duplicates.)
2214 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2215 systems where poll() doesn't work, in particular OS X.
2217 PH/10 Added more information to debugging output for retry time not reached.
2219 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2220 operations in malware.c.
2222 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2225 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2226 syslog_duplication set false, log lines that would normally be written
2227 both the the main log and to the reject log were not written to syslog at
2230 PH/14 In the default configuration, change the use of "message" in ACL warn
2231 statements to "add_header".
2233 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2234 not followed by a command (e.g. "seen endif").
2236 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2237 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2240 PH/17 Added extra conditions to the default value of delay_warning_condition
2244 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2245 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2246 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2249 The Auto-Submitted: and various List- headers are standardised, whereas I
2250 don't think Precedence: ever was.
2252 PH/18 Refactored debugging code in route_finduser() to show more information,
2253 in particular, the error code if getpwnam() issues one.
2255 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2256 This is apparently needed in addition to the PH/07 change above to avoid
2257 any possible encoding problems.
2259 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2260 but not after initializing Perl.
2262 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2263 output them only if debugging. By default they are written stderr,
2264 apparently, which is not desirable.
2266 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2269 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2272 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2274 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2275 authenticated or an ident call has been made. Suppress the default
2276 values for $authenticated_id and $authenticated_sender (but permit -oMai
2277 and -oMas) when testing with -bh.
2279 PH/24 Re-jigged the order of the tests in the default configuration so that the
2280 tests for valid domains and recipients precede the DNS black list and CSA
2281 tests, on the grounds that those ones are more expensive.
2283 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2284 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2285 This bug exists in every version of Exim that I still have, right back to
2288 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2289 However, an attempt to turn on case-sensitivity in a regex key by
2290 including (?-i) didn't work because the subject string was already
2291 lowercased, and the effects were non-intuitive. It turns out that a
2292 one-line patch can be used to allow (?-i) to work as expected.
2298 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2299 other effects) broke the use of negated acl sub-conditions.
2301 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2304 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2305 "Deny" causes Exim to reject the incoming connection with a 554 error.
2306 Unfortunately, if there is a major crisis, such as a disk failure,
2307 tcp-wrappers gives "deny", whereas what one would like would be some
2308 kind of temporary error. A kludge has been added to help with this.
2309 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2310 554 error is used if errno is still zero or contains ENOENT (which occurs
2311 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2314 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2316 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2317 errors. Otherwise a message that provokes a temporary error (when other
2318 messages do not) can cause a whole host to time out.
2320 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2321 addresses were routed directly to files or pipes from a redirect router.
2322 File deliveries just didn't batch; pipe deliveries might have suffered
2325 PH/06 A failure to get a lock for a hints database would erroneously always say
2326 "Failed to get write lock", even when it was really a read lock.
2328 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2329 of 0600. This has been changed to use the value of the lockfile_mode
2330 option (which defaults to 0600).
2332 PH/08 Applied small patch from the Sieve maintainer.
2334 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2335 folder from quota calculations, a direct delivery into this folder messed
2336 up the contents of the maildirsize file. This was because the regex was
2337 used only to exclude .Trash (or whatever) when the size of the mailbox
2338 was calculated. There was no check that a delivery was happening into an
2339 excluded directory. This bug has been fixed by ignoring all quota
2340 processing for deliveries into excluded directories.
2342 PH/10 Added the maildirfolder_create_regex option to appendfile.
2348 PH/01 The code for finding all the local interface addresses on a FreeBSD
2349 system running IPv6 was broken. This may well have applied to all BSD
2350 systems, as well as to others that have similar system calls. The broken
2351 code found IPv4 interfaces correctly, but gave incorrect values for the
2352 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2353 that it would not match correctly against @[] and not recognize the IPv6
2356 PH/02 The ipliteral router was not recognizing addresses of the form user@
2357 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2359 PH/03 Added disable_ipv6.
2361 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2362 decoded form, because it is most often used to construct To: headers
2363 lines in autoreplies, and the decoded form may well be syntactically
2364 invalid. However, $reply_address has leading white space removed, and all
2365 newlines turned into spaces so that the autoreply transport does not
2368 PH/05 If group was specified without a user on a router, and no group or user
2369 was specified on a transport, the group from the router was ignored.
2371 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2372 for visible compile-time settings that can be used to change these
2373 numbers, for those that want even more. Backwards compatibility with old
2374 spool files has been maintained. However, going back to a previous Exim
2375 release will lost any variables that are in spool files.
2377 PH/07 Two small changes when running in the test harness: increase delay when
2378 passing a TCP/IP connection to a new process, in case the original
2379 process has to generate a bounce, and remove special handling of
2380 127.0.0.2 (sic), which is no longer necessary.
2382 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2383 be the same on different OS.
2385 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2388 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2389 whether --show-vars was specified or not
2391 JJ/02 exipick: Added support for new ACL variable spool format introduced
2394 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2395 syntactically invalid From: or Reply-to: line, and a filter used this to
2396 generate an autoreply, and therefore failed to obtain an address for the
2397 autoreply, Exim could try to deliver to a non-existent relative file
2398 name, causing unrelated and misleading errors. What now happens is that
2399 it logs this as a hard delivery error, but does not attempt to create a
2402 PH/11 The exinext utility has a -C option for testing purposes, but although
2403 the given file was scanned by exinext itself; it wasn't being passed on
2404 when Exim was called.
2406 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2407 an end-of-file indication when reading a command response.
2409 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2410 compiled. In many other places in Exim, IPv6 addresses are always
2411 recognized, so I have changed this. It also means that IPv4 domain
2412 literals of the form [IPV4:n.n.n.n] are now always recognized.
2414 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2415 used if the router is not running as root, for example, when verifying at
2416 ACL time, or when using -bh. The debugging output from this situation was
2417 non-existent - all you got was a failure to exec. I have made two
2420 (a) Failures to set uid/gid, the current directory, or a process leader
2421 in a subprocess such as that created by queryprogram now generate
2422 suitable debugging ouput when -d is set.
2424 (b) The queryprogram router detects when it is not running as root,
2425 outputs suitable debugging information if -d is set, and then runs
2426 the subprocess without attempting to change uid/gid.
2428 PH/15 Minor change to Makefile for building test_host (undocumented testing
2431 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2432 additional section of a DNS packet that returns MX or SRV records.
2433 Instead, it always explicitly searches for A/AAAA records. This avoids
2434 major problems that occur when a DNS server includes only records of one
2435 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2436 fixed another bug: if SRV records were looked up and the corresponding
2437 address records were *not* found in the additional section, the port
2438 values from the SRV records were lost.
2440 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2441 using the correct key (the original address) when searching the retry
2442 rules in order to find which one to use for generating the retry hint.
2444 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2445 adding the default one. Similarly, if it contains a Reply-To: header, the
2446 errors_reply_to option, if set, is not used.
2448 PH/19 When calculating a retry time, Exim used to measure the "time since
2449 failure" by looking at the "first failed" field in the retry record. Now
2450 it does not use this if it is later than than the arrival time of the
2451 message. Instead it uses the arrival time. This makes for better
2452 behaviour in cases where some deliveries succeed, thus re-setting the
2453 "first failed" field. An example is a quota failure for a huge message
2454 when small messages continue to be delivered. Without this change, the
2455 "time since failure" will always be short, possible causing more frequent
2456 delivery attempts for the huge message than are intended.
2457 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2459 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2460 $1, $2, $3) because the numerical variables can be reset during some
2461 expansion items (e.g. "match"), thereby losing the authentication data.
2463 PH/21 Make -bV show the size of off_t variables so that the test suite can
2464 decide whether to run tests for quotas > 2G.
2466 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2467 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2468 filecount value is greater than 2G or if a quota value is greater than 2G
2469 on a system where the size of off_t is not greater than 4, a panic error
2472 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2473 never match. The debug and -bh output now contains an explicit error
2474 message indicating a malformed IPv4 address or mask.
2476 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2477 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2478 PH/23 above applies.
2480 PH/25 Do not write to syslog when running in the test harness. The only
2481 occasion when this arises is a failure to open the main or panic logs
2482 (for which there is an explicit test).
2484 PH/26 Added the /no_tell option to "control=freeze".
2486 PH/27 If a host name lookup failed very early in a connection, for example, if
2487 the IP address matched host_lookup and the reverse lookup yielded a name
2488 that did not have a forward lookup, an error message of the form "no IP
2489 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2490 could be logged. Now it outputs the IP address instead of "NULL".
2492 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2493 allows the sender and the authenticated sender to be set when
2494 submitting a message from within Exim. Since child_open_exim() is
2495 documented for local_scan(), the new function should be too.
2497 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2498 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2499 results in an empty string is now treated as unset.
2501 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2503 PH/31 Added sender_verify_fail logging option.
2505 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2506 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2507 tidied the source and removed it altogether.
2509 PH/33 When a queue run was abandoned because the load average was too high, a
2510 log line was always written; now it is written only if the queue_run log
2511 selector is set. In addition, the log line for abandonment now contains
2512 information about the queue run such as the pid. This is always present
2513 in "start" and "stop" lines but was omitted from the "abandon" line.
2515 PH/34 Omit spaces between a header name and the colon in the error message that
2516 is given when verify = headers_syntax fails (if there are lots of them,
2517 the message gets confusing).
2519 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2520 names, as there are now some PTR records that contain slashes. This check
2521 is only to protect against broken name servers that fall over on strange
2522 characters, so the fact that it applies to all lookups doesn't matter.
2524 PH/36 Now that the new test suite is complete, we can remove some of the
2525 special code in Exim that was needed for the old test suite. For example,
2526 sorting DNS records because real resolvers return them in an arbitrary
2527 order. The new test suite's fake resolver always returns records in the
2530 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2531 bounces) except when queue_only is set, to avoid logging races between
2532 the different processes.
2534 PH/38 Panic-die if .include specifies a non-absolute path.
2536 PH/39 A tweak to the "H" retry rule from its user.
2538 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2539 a label. They prevented compilation on older perls.
2541 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2542 a warning to be raised on newish perls.
2544 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2545 on queue. Changes to match documented behaviour of showing count of
2546 messages matching specified criteria.
2548 PH/40 Changed the default ident timeout from 30s to 5s.
2550 PH/41 Added support for the use of login_cap features, on those BSD systems
2551 that have them, for controlling the resources used by pipe deliveries.
2553 PH/42 The content-scanning code uses fopen() to create files in which to put
2554 message data. Previously it was not paying any attention to the mode of
2555 the files. Exim runs with umask(0) because the rest of the code creates
2556 files with open(), and sets the required mode explicitly. Thus, these
2557 files were ending up world-writeable. This was not a big issue, because,
2558 being within the spool directory, they were not world-accessible. I have
2559 created a function called modefopen, which takes an additional mode
2560 argument. It sets umask(777), creates the file, chmods it to the required
2561 mode, then resets the umask. All the relevant calls to fopen() in the
2562 content scanning code have been changed to use this function.
2564 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2565 to 24 hours. This avoids potential overflow problems when processing G
2566 and H retry rules. I suspect nobody ever tinkers with this value.
2568 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2570 PH/45 When the plaintext authenticator is running as a client, the server's
2571 challenges are checked to ensure they are valid base64 strings. By
2572 default, the authentication attempt is cancelled if an invalid string is
2573 received. Setting client_ignore_invalid_base64 true ignores these errors.
2574 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2575 they are received. Thus, the responses can be made to depend on the
2576 challenges. If an invalid string is ignored, an empty string is placed in
2579 PH/46 Messages that are created by the autoreply transport now contains a
2580 References: header, in accordance with RFCs 2822 and 3834.
2582 PH/47 Added authenticated_sender_force to the smtp transport.
2584 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2586 PH/49 Installed latest patch from the Sieve maintainer.
2588 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2589 was also set, the appendfile transport was unnecessarily scanning a
2590 directory of message files (e.g. for maildir delivery) to find the count
2591 of files (along with the size), even though it did not need this
2592 information. It now does the scan only if it needs to find either the
2593 size of the count of files.
2595 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2597 PH/52 Two bugs concerned with error handling when the smtp transport is
2600 (i) Exim was not creating retry information for temporary errors given
2601 for individual recipients after the DATA command when the smtp transport
2602 was used in LMTP mode. This meant that they could be retried too
2603 frequently, and not timed out correctly.
2605 (ii) Exim was setting the flag that allows error details to be returned
2606 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2607 recipients that were returned after the DATA command.
2609 PH/53 This is related to PH/52, but is more general: for any failing address,
2610 when detailed error information was permitted to be returned to the
2611 sender, but the error was temporary, then after the final timeout, only
2612 "retry timeout exceeded" was returned. Now it returns the full error as
2613 well as "retry timeout exceeded".
2615 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2616 do this, and (what is worse) MTAs that accept it.
2618 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2619 will now be deprecated.
2621 PH/56 New os.c-cygwin from the Cygwin maintainer.
2623 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2624 formats (previously only available in exim formats via -bpr, -bpru,
2625 and -bpra. Now also available in native and exiqgrep formats)
2627 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2628 with very large, slow to parse queues
2630 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2632 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2634 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2635 responses to authentication challenges, though it was showing the
2636 challenges; (ii) I've removed the CR characters from the debug output for
2639 PH/58 Allow for the insertion of a newline as well as a space when a string
2640 is turned into more than one encoded-word during RFC 2047 encoding. The
2641 Sieve code now uses this.
2643 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2644 data_4xx, lost_connection, tls_required.
2646 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2647 message was being sent as an SMTP response.
2649 PH/61 Add -l and -k options to exicyclog.
2651 PH/62 When verifying, if an address was redirected to one new address, so that
2652 verification continued, and the new address failed or deferred after
2653 having set something in $address_data, the value of $address_data was not
2654 passed back to the ACL. This was different to the case when no
2655 redirection occurred. The value is now passed back in both cases.
2657 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2658 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2659 use login_cap.h, so on its own it isn't the distinguishing feature. The
2660 new name refers directly to the setclassresources() function.
2662 PH/65 Added configuration files for NetBSD3.
2664 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2666 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2667 is preferred over IPv4.
2669 PH/68 The bounce_return_message and bounce_return_body options were not being
2670 honoured for bounces generated during the reception of non-SMTP messages.
2671 In particular, this applied to messages rejected by the ACL. This bug has
2672 been fixed. However, if bounce_return_message is true and bounce_return_
2673 body is false, the headers that are returned for a non-SMTP message
2674 include only those that have been read before the error was detected.
2675 (In the case of an ACL rejection, they have all been read.)
2677 PH/69 The HTML version of the specification is now built in a directory called
2678 spec_html instead of spec.html, because the latter looks like a path with
2679 a MIME-type, and this confuses some software.
2681 PH/70 Catch two compiler warnings in sieve.c.
2683 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2684 function verify_get_ident() calls ip_connect() to connect a socket, but
2685 if the "connect()" function timed out, ip_connect() used to close the
2686 socket. However, verify_get_ident() also closes the socket later, and in
2687 between Exim writes to the log, which may get opened at this point. When
2688 the socket was closed in ip_connect(), the log could get the same file
2689 descriptor number as the socket. This naturally causes chaos. The fix is
2690 not to close the socket in ip_connect(); the socket should be closed by
2691 the function that creates it. There was only one place in the code where
2692 this was missing, in the iplookup router, which I don't think anybody now
2693 uses, but I've fixed it anyway.
2695 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2696 well as to direct DNS lookups. Otherwise the handling of names in host
2697 lists is inconsistent and therefore confusing.
2703 PH/01 Two changes to the default runtime configuration:
2705 (1) Move the checks for relay_from_hosts and authenticated clients from
2706 after to before the (commented out) DNS black list checks.
2708 (2) Add control=submission to the relay_from_hosts and authenticated
2709 clients checks, on the grounds that messages accepted by these
2710 statements are most likely to be submissions.
2712 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2714 (1) Generate an error if the third argument for the ${prvs expansion is
2717 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2720 (3) Reset the variables that are obtained from the first argument of
2721 ${prvscheck and used in the second argument before leaving the code,
2722 because their memory is reclaimed, so using them afterwards may do
2725 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2726 one (it's much easier than Tom thought :-).
2728 (5) Because of (4), we can now allow for the use of $prvscheck_result
2729 inside the third argument.
2731 PH/03 For some reason, the default setting of PATH when running a command from
2732 a pipe transport was just "/usr/bin". I have changed it to
2735 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2736 anything to be listed in the output from -bV.
2738 PH/05 When a filter generated an autoreply, the entire To: header line was
2739 quoted in the delivery log line, like this:
2741 => >A.N.Other <ano@some.domain> <original@ddress> ...
2743 This has been changed so that it extracts the operative address. There
2744 may be more than one such address. If so, they are comma-separated, like
2747 => >ano@some.domain,ona@other.domain <original@ddress> ...
2749 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2750 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2751 not being looked up in the rDNS to get a host name, Exim was showing the
2752 IP address twice in Received: lines, even though the IP addresses were
2753 identical. For example:
2755 Received: from [1.2.3.4] (helo=[1.2.3.4])
2757 However, if the real host name was known, it was omitting the HELO data
2758 if it matched the actual IP address. This has been tidied up so that it
2759 doesn't show the same IP address twice.
2761 PH/07 When both +timestamp and +memory debugging was on, the value given by
2762 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2763 called by the debug printing, thereby overwriting the timestamp buffer.
2764 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2766 PH/08 When the original message was included in an autoreply transport, it
2767 always said "this is a copy of the message, including all the headers",
2768 even if body_only or headers_only was set. It now gives an appropriate
2771 PH/09 Applied a patch from the Sieve maintainer which:
2773 o fixes some comments
2774 o adds the (disabled) notify extension core
2775 o adds some debug output for the result of if/elsif tests
2776 o points to the current vacation draft in the documentation
2777 and documents the missing references header update
2781 o fixes a bug in processing the envelope test (when testing
2782 multiple envelope elements, the last element determinted the
2785 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2786 Electronic Mail") by including:
2788 Auto-submitted: auto-generated
2790 in the messages that it generates (bounce messages and others, such as
2791 warnings). In the case of bounce messages for non-SMTP mesages, there was
2792 also a typo: it was using "Auto_submitted" (underscore instead of
2793 hyphen). Since every message generated by Exim is necessarily in response
2794 to another message, thes have all been changed to:
2796 Auto-Submitted: auto-replied
2798 in accordance with these statements in the RFC:
2800 The auto-replied keyword:
2802 - SHOULD be used on messages sent in direct response to another
2803 message by an automatic process,
2805 - MUST NOT be used on manually-generated messages,
2807 - MAY be used on Delivery Status Notifications (DSNs) and Message
2808 Disposition Notifications (MDNs),
2810 - MUST NOT be used on messages generated by automatic or periodic
2811 processes, except for messages which are automatic responses to
2814 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2815 to the default Received: header definition.
2817 PH/12 Added log selector acl_warn_skipped (default on).
2819 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2820 variables because (a) they are in the wrong storage pool and (b) even if
2821 they were copied, it wouldn't work properly because of the caching.
2823 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2824 checking when decoding. Apparently there are clients that generate
2825 overlong encoded strings. Why am I not surprised?
2827 PH/15 If the first argument of "${if match_address" was not empty, but did not
2828 contain an "@" character, Exim crashed. Now it writes a panic log message
2829 and treats the condition as false.
2831 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2833 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2834 extension "envelope-auth". The code is finished and in agreement with
2835 other implementations, but there is no documentation so far and in fact,
2836 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2837 not changing the active code.
2839 Print executed "if" and "elsif" statements when debugging is used. This
2840 helps a great deal to understand what a filter does.
2842 Document more things not specified clearly in RFC3028. I had all this
2843 sorted out, when out of a sudden new issues came to my mind. Oops."
2845 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2848 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2849 canonical form (as documented). However, after a host name lookup from
2850 the IP address, check_host() was doing a simple string comparison with
2851 addresses acquired from the DNS when checking that the found name did
2852 have the original IP as one of its addresses. Since any found IPv6
2853 addresses are likely to be in abbreviated form, the comparison could
2854 fail. Luckily, there already exists a function for doing the comparison
2855 by converting both addresses to binary, so now that is used instead of
2856 the text comparison.
2858 PH/20 There was another similar case to PH/19, when a complete host name was
2859 given in a host list; looking up its IP address could give an abbreviated
2860 form, whereas the current host's name might or might not be abbreviated.
2861 The same fix has been applied.
2867 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2868 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2871 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2872 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2874 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2876 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2877 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2878 Also, for "Auto-Submitted", treat anything other than "no" as
2879 non-personal, in accordance with RFC 3834. (Previously it treated
2880 anything starting "auto-" as non-personal.)
2882 TF/01 The control=submission/name=... option had a problem with syntax
2883 errors if the name included a slash character. The /name= option
2884 now slurps the rest of the string, so it can include any characters
2885 but it must come last in the list of options (after /sender_retain
2888 PH/05 Some modifications to the interface to the fake nameserver for the new
2896 TK/01 Added the "success_on_redirect" address verification option. See
2897 NewStuff for rationale and an example.
2899 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2901 PH/02 Patch to exigrep to allow it to work on syslog lines.
2903 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2904 fread() to skip over the body file's header line, because in Cygwin the
2905 header line is locked and is inaccessible.
2907 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2908 co-exist for some time) to make it clear that it is the Exim ID that is
2909 referenced, not the Message-ID: header line.
2911 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2912 string_format() function, because snprintf() does not exist on all
2915 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2916 ${stat: expansion item.
2918 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2919 protocol synchronization error", to keep the pedants happy.
2921 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2922 well as for IRIX systems, when gcc is being used. See the host.c source
2925 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2927 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2930 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2931 transport and to the smtp transport in LMTP mode.
2933 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2935 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2936 run a filter in a subprocess. This could lead to confusion in subsequent
2937 lookups in the parent process. There should also be a search_tidyup() at
2938 the end of the subprocess.
2940 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2941 only if the host matched helo_try_verify_hosts, which caused the
2942 verification to occur when the EHLO/HELO command was issued. The ACL just
2943 tested the remembered result. Now, if a previous verification attempt has
2944 not happened, "verify = helo" does it there and then.
2946 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2948 TK/03 Fix log output including CR from clamd.
2950 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2951 exist provoked a memory error which could cause a segfault.
2953 PH/15 Installed PCRE 6.2
2955 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2957 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2958 of the problem. Specifically, suggested +O2 rather than +O1 for the
2961 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2963 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2964 record, the port was not correctly logged when the outgoing_port log
2965 selector was set (it logged the transort's default port).
2967 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2968 fallback_hosts, and "hosts" in the smtp transport.
2970 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2971 host errors such as "Connection refused".
2973 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2974 authentication with radiusclient 0.4.9:
2976 - Error returned from rc_read_config was caught wrongly
2977 - Username/password not passed on to radius server due to wrong length.
2979 The presumption is that some radiusclient API changes for 4.51/PH/17
2980 were not taken care of correctly. The code is still untested by me (my
2981 Linux distribution still has 0.3.2 of radiusclient), but it was
2982 contributed by a Radius user.
2984 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2985 expanding the "port" option of the smtp transport.
2987 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2988 while reading a MIME header. Thanks to Tom Hughes for a patch.
2990 PH/24 Include config.h inside local_scan.h so that configuration settings are
2993 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2994 that in an ACL for RCPT (for example), you can examine exactly what was
2997 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2998 commands, but it was not correctly comparing the address with the actual
2999 client host address. Thus, it would show the EHLO address in Received:
3000 header lines when this was not necessary.
3002 PH/27 Added the % operator to ${eval:}.
3004 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3005 it should be ignoring failures (because with -C, for example, it has lost
3006 privilege). It wasn't ignoring creation failures other than "already
3009 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3012 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3013 because an input line was too long, either on its own, or by virtue of
3014 too many continuations, the temporary file was not being removed, and the
3015 return code was incorrect.
3017 PH/31 Missing "BOOL" in function definition in filtertest.c.
3019 PH/32 Applied Sieve patches from the maintainer.
3021 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3023 PH/33 Added "verify = not_blind".
3025 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3026 Local/Makefile (with some defaults set). These are used in built scripts
3027 such as exicyclog, but they have never been used in the exim_install
3028 script (though there are many overriding facilities there). I have
3029 arranged that the exim_install script now takes note of these two
3032 PH/35 Installed configuration files for Dragonfly.
3034 PH/36 When a locally submitted message by a trusted user did not contain a
3035 From: header, and the sender address was obtained from -f or from an SMTP
3036 MAIL command, and the trusted user did not use -F to supply a sender
3037 name, $originator_name was incorrectly used when constructing a From:
3038 header. Furthermore, $originator_name was used for submission mode
3039 messages from external hosts without From: headers in a similar way,
3040 which is clearly wrong.
3042 PH/37 Added control=suppress_local_fixups.
3044 PH/38 When log_selector = +received_sender was set, and the addition of the
3045 sender made the log line's construction buffer exactly full, or one byte
3046 less than full, an overflow happened when the terminating "\n" was
3049 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3050 when the result of a list match is failure because a DNS lookup failed.
3052 PH/40 RM_COMMAND is now used in the building process.
3054 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3055 the "build-* directories that it finds.
3057 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3058 address in a domain literal was a prefix of an interface address.
3060 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3061 when verifying a sender address, unless rewrite_headers is false.
3063 PH/44 Wrote a long comment about why errors_to addresses are verified as
3064 recipients, not senders.
3066 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3067 the ratelimit ACL was added.
3069 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3071 PH/46 Added extra information about PostgreSQL errors to the error string.
3073 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3074 avoiding the need to install special zones in a real server. This is
3075 backwards compatible; if it can't find the fake resolver, it drops back.
3076 Thus, both old and new test suites can be run.
3078 TF/02 Added util/ratelimit.pl
3080 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3081 clock is set back in time.
3083 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3084 Candler <B.Candler@pobox.com>.
3086 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3087 OK for addresses that are the result of redirections.
3089 PH/48 A number of further additions for the benefit of the new test suite,
3090 including a fake gethostbyname() that interfaces to the fake DNS resolver
3093 TF/06 The fix for widen_domains has also been applied to qualify_single and
3094 search_parents which are the other dnslookup options that can cause
3097 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3100 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3102 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3103 service" if it encountered an MX record with an empty target hostname.
3104 The message is now "an MX or SRV record indicated no SMTP service".
3106 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3107 if the DNS of the sending site is misconfigured. This is quite a
3108 common situation. This change restores the behaviour of treating a
3109 helo verification defer as a failure.
3111 PH/51 If self=fail was set on a router, the bounce message did not include the
3112 actual error message.
3118 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3120 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3121 command itself ended in error, the underlying message about the transport
3122 filter timeout was being overwritten with the pipe command error. Now the
3123 underlying error message should be appended to the second error message.
3125 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3127 PH/02 Reduce the amount of output that "make" produces by default. Full output
3128 can still be requested.
3130 PH/03 The warning log line about a condition test deferring for a "warn" verb
3131 was being output only once per connection, rather than after each
3132 occurrence (because it was using the same function as for successful
3133 "warn" verbs). This seems wrong, so I have changed it.
3135 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3136 it should not have, which might have caused a crash in the right
3137 circumstances, but probably never did.
3139 PH/04 Installed a modified version of Tony Finch's patch to make submission
3140 mode fix the return path as well as the Sender: header line, and to
3141 add a /name= option so that you can make the user's friendly name appear
3144 TF/03 Added the control = fakedefer ACL modifier.
3146 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3147 Mark Lowes for thorough testing.
3149 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3151 TK/03 Merged latest SRS patch from Miles Wilton.
3153 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3154 with the definition in sysexits.h (which is #included earlier).
3155 Fortunately, Exim does not actually use EX_OK. The code used to try to
3156 preserve the sysexits.h value, by assumimg that macro definitions were
3157 scanned for macro replacements. I have been disabused of this notion,
3158 so now the code just undefines EX_OK before #including unistd.h.
3160 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3161 in the smtp transport. When a block could not be written in a single
3162 write() function, the timeout was being re-applied to each part-write.
3163 This seems wrong - if the receiver was accepting one byte at a time it
3164 would take for ever. The timeout is now adjusted when this happens. It
3165 doesn't have to be particularly precise.
3167 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3168 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3170 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3171 without a domain if the check to <postmaster@domain> fails.
3173 SC/01 Eximstats: added -xls and the ability to specify output files
3174 (patch written by Frank Heydlauf).
3176 SC/02 Eximstats: use FileHandles for outputing results.
3178 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3180 SC/04 Eximstats: fixed display of large numbers with -nvr option
3182 SC/05 Eximstats: fixed merging of reports with empty tables.
3184 SC/06 Eximstats: added the -include_original_destination flag
3186 SC/07 Eximstats: removed tabs and trailing whitespace.
3188 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3190 TK/06 MBOX spool code: Add real "From " MBOX separator line
3191 so the .eml file is really in mbox format (even though
3192 most programs do not really care). Patch from Alex Miller.
3194 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3195 The latter is generated from $received_to and is only set if the
3196 message has one envelope recipient. SA can use these headers,
3197 obviously out-of-the-box. Patch from Alex Miller.
3199 PH/08 The ${def test on a variable was returning false if the variable's
3200 value was "0", contrary to what the specification has always said!
3201 The result should be true unless the variable is empty.
3203 PH/09 The syntax error of a character other than { following "${if
3204 def:variable_name" (after optional whitespace) was not being diagnosed.
3205 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3206 accidental colon was present, for example, could give incorrect results.
3208 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3209 result is used (not including appendfile, where other changes are about
3212 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3213 This involved changing a lot of size variables from int to off_t. It
3214 should work with maildirs and everything.
3216 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3217 spamd dying while we are connected to it.
3219 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3222 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3223 function for BDB 4.3.
3225 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3227 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3228 the sources for PCRE have been re-arranged and more files are now
3231 PH/15 The code I had for printing potentially long long variables in PH/11
3232 above was not the best (it lost precision). The length of off_t variables
3233 is now inspected at build time, and an appropriate printing format (%ld
3234 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3235 to be "long long int" or "long int". This is needed for the internal
3236 formatting function string_vformat().
3238 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3239 the configuration file to be ":syslog", then the script "guesses" where
3240 the logs files are, rather than using the compiled in default. In our
3241 case the guess is not the same as the compiled default, so the script
3242 suddenly stopped working when I started to use syslog. The patch checks
3243 to see if log_file_path is "". If so, it attempts to read it from exim
3244 with no configuration file to get the compiled in version, before it
3245 falls back to the previous guessing code."
3247 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3248 implementing BATV in an Exim configuration. See NewStuff for the gory
3251 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3252 Makefile that are specific to HP-UX.
3254 PH/18 If the "use_postmaster" option was set for a recipient callout together
3255 with the "random" option, the postmaster address was used as the MAIL
3256 FROM address for the random test, but not for the subsequent recipient
3257 test. It is now used for both.
3259 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3260 patch removes a few documentation additions to RFC 3028, because the
3261 latest draft now contains them. It adds the new en;ascii-case comparator
3262 and a new error check for 8bit text in MIME parts. Comparator and
3263 require names are now matched exactly. I enabled the subaddress
3264 extension, but it is not well tested yet (read: it works for me)."
3266 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3267 rework some of the code of TK/09 above to avoid the hardwired use of
3268 "%lld" and "long long". Replaced the call to snprintf() with a call to
3271 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3272 records point to non-existent hosts", "retry timeout exceeded", and
3273 "retry time not reached for any host after a long failure period".
3275 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3276 experimental DomainKeys support:
3278 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3279 (2) On an error such as an illegally used "control", the wrong name for
3280 the control was given.
3282 These problems did NOT occur unless DomainKeys support was compiled.
3284 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3286 PH/24 Added ${if match_ip condition.
3288 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3289 databases so that it will be absolutely obvious if a crash occurs in the
3290 DB library. This is a regular occurrence (often caused by mis-matched
3293 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3294 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3295 picked up on a user's system that detects such things. There doesn't seem
3296 to be a gcc warning option for this - only an attribute that has to be
3297 put on the function's prototype. It seems that in Fedora Core 4 they have
3298 set this on a number of new functions. No doubt there will be more in due
3301 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3302 specify a transport. However, if an address that was verified by such a
3303 router was the subject of a callout, Exim crashed because it tried to
3304 read the rcpt_include_affixes from the non-existent transport. Now it
3305 just assumes that the setting of that option is false. This bug was
3306 introduced by 4.51/PH/31.
3308 PH/28 Changed -d+all to exclude +memory, because that information is very
3309 rarely of interest, but it makes the output a lot bigger. People tend to
3310 do -d+all out of habit.
3312 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3313 code in os-type was giving problems when libc.so lives in lib64, like on
3316 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3317 aren't the modern standard, and it seems that some systems' include files
3318 don't always have them. Exim was already checking for some of the newer
3319 ones like T_AAAA, and defining it itself. I've added checks for all the
3320 record types that Exim uses.
3322 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3323 not automatically generating a new one, as it is supposed to. This
3324 prevented TLS from working. If the file did exist, but contained invalid
3325 data, a new version was generated, as expected. It was only the case of a
3326 non-existent file that was broken.
3328 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3329 with a change in libdomainkeys > 0.64.
3331 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3332 from DNS. If the selector record carries the flag, it now has
3333 precedence over the domain-wide flag.
3335 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3337 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3338 the use of a port name that isn't defined in /etc/services) occurred, the
3339 message was deferred as in a normal delivery, and thus remained on the
3340 spool, instead of being failed because of the mua_wrapper setting. This
3341 is now fixed, and I tidied up some of the mua_wrapper messages at the
3344 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3345 the messages in a hash of arrays rather than using individual hashes.
3346 This is a bit cleaner and results in dramatic memory savings, albeit
3347 at a slight CPU cost.
3349 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3350 as requested by Marc Sherman.
3352 SC/10 Eximstats: added histograms for user specified patterns as requested
3355 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3357 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3358 fopen() in the content-scanning modules that did not already have it.
3364 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3365 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3367 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3369 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3371 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3372 to test Sieve filters that use "vacation".
3374 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3375 that changes the way the GnuTLS parameters are stored in the cache file.
3376 The new format can be generated externally. For backward compatibility,
3377 if the data in the cache doesn't make sense, Exim assumes it has read an
3378 old-format file, and it generates new data and writes a new file. This
3379 means that you can't go back to an older release without removing the
3382 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3383 work if there are any delivery delays because "one_time" forces the
3384 parent to be marked "delivered", so its unseen clone is never tried
3385 again. For this reason, Exim now forbids the simultaneous setting of
3388 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3389 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3390 entry if you want to know the details. The fix, however, neglected to
3391 consider the case when local delivery batching is involved. The test for
3392 "previously delivered" was not happening when checking to see if an
3393 address could be batched with a previous (undelivered) one; under
3394 certain circumstances this could lead to multiple deliveries to the same
3397 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3398 in its include files, and this causes problems building Exim.
3400 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3401 header_syntax) but Exim was just ignoring anything given after a slash.
3402 In particular, this caused confusion with an attempt to use "verify =
3403 reverse_host_lookup/defer_ok". An error is now given when options are
3404 supplied for verify items that do not have them. (Maybe reverse_host_
3405 lookup should have a defer_ok option, but that's a different point.)
3407 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3408 defined by RFC 821) to 2048, because there were problems with some AUTH
3409 commands, and RFC 1869 says the size should be increased for extended
3410 SMTP commands that take arguments.
3412 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3415 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3416 "unknown" error; now it says that the functionality isn't in the binary.
3418 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3419 an address' error message when a string expansion fails (syntax or
3420 whatever). Otherwise the password may appear in the log. Following change
3421 PH/42 below, there is no longer a chance of it appearing in a bounce
3424 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3426 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3427 of its addresses was ever tried. (Bugzilla bug #2.)
3429 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3430 the result incorrectly in the debug output. (It correctly added a newline
3431 to what was transported.)
3433 TF/01 Added $received_time.
3435 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3436 commented out examples of how to interface to a virus scanner and to
3437 SpamAssassin. Also added commented examples of av_scanner and
3438 spamd_address settings.
3440 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3441 and controls are allowed in which ACLs. There were a couple of minor
3442 errors. Some of the entries in the conditions table (which is a table of
3443 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3444 negation of where the condition IS allowed.
3446 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3448 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3449 header file does not have a version number, so I've had to invent a new
3450 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3451 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3452 radiusclient), but it was contributed by a Radius user.
3454 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3455 files or directories, for OpenSSL.
3457 PH/19 When an Exim process that is running as root has to create an Exim log
3458 file, it does so in a subprocess that runs as exim:exim so as to get the
3459 ownership right at creation (otherwise, other Exim processes might see
3460 the file with the wrong ownership). There was no test for failure of this
3461 fork() call, which would lead to the process getting stuck as it waited
3462 for a non-existent subprocess. Forks do occasionally fail when resources
3463 run out. I reviewed all the other calls to fork(); they all seem to check
3466 PH/20 When checking for unexpected SMTP input at connect time (before writing
3467 the banner), Exim was not dealing correctly with a non-positive return
3468 from the read() function. If the client had disconnected by this time,
3469 the result was a log entry for a synchronization error with an empty
3470 string after "input=" when read() returned zero. If read() returned -1
3471 (an event I could not check), uninitialized data bytes were printed.
3472 There were reports of junk text (parts of files, etc) appearing after
3475 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3477 PH/22 Added support for macro redefinition, and (re)definition in between
3478 driver and ACL definitions.
3480 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3481 forgetting to use the resulting value; it was using the unexpanded value.
3483 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3484 hadn't been configured. The fix is from Juergen Kreileder, who
3485 understands it better than I do:
3487 "Here's what I see happening with three configured cyrus_sasl
3488 authenticators configured (plain, login, cram-md5):
3490 On startup auth_cyrus_sasl_init() gets called for each of these.
3491 This means three calls to sasl_listmech() without a specified mech_list.
3492 => SASL tests which mechs of all available mechs actually work
3493 => three warnings about OTP not working
3494 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3496 With the patch, sasl_listmech() also gets called three times. But now
3497 SASL's mech_list option is set to the server_mech specified in the the
3498 authenticator. Or in other words, the answer from sasl_listmech()
3499 gets limited to just the mech you're testing for (which is different
3501 => the return list contains just 'plain' or 'login', 'cram-md5' or
3502 nothing depending on the value of ob->server_mech.
3504 I've just tested the patch: Authentication still works fine,
3505 unavailable mechs specified in the exim configuration are still
3506 caught, and the auth.log warnings about OTP are gone."
3508 PH/25 When debugging is enabled, the contents of the command line are added
3509 to the debugging output, even when log_selector=+arguments is not
3512 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3513 answer is "GNU", and only if the return is "GNU/something" is the answer
3516 PH/27 $acl_verify_message is now set immediately after the failure of a
3517 verification in an ACL, and so is available in subsequent modifiers. In
3518 particular, the message can be preserved by coding like this:
3520 warn !verify = sender
3521 set acl_m0 = $acl_verify_message
3523 Previously, $acl_verify_message was set only while expanding "message"
3524 and "log_message" when a very denied access.
3526 PH/28 Modified OS/os.c-Linux with
3528 -#ifndef OS_LOAD_AVERAGE
3529 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3531 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3532 nomenclature these days.)
3534 PH/29 Installed patch from the Sieve maintainer that adds the options
3535 sieve_useraddress and sieve_subaddress to the redirect router.
3537 PH/30 In these circumstances:
3538 . Two addresses routed to the same list of hosts;
3539 . First host does not offer TLS;
3540 . First host accepts first address;
3541 . First host gives temporary error to second address;
3542 . Second host offers TLS and a TLS session is established;
3543 . Second host accepts second address.
3544 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3545 and peerdn, if requested) that were in fact used only for the second
3548 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3549 attention to any local part prefix or suffix that was matched by the
3550 router that accepted the address. It now behaves in the same way as it
3551 does for delivery: the affixes are removed from the local part unless
3552 rcpt_include_affixes is set on the transport.
3554 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3555 timeout during the DATA phase of an incoming message.
3557 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3558 applied a patch sanctioned by the Sieve maintainer.
3560 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3561 the uid or gid is negative. A case of a negative gid caused this to be
3562 noticed. The fix allows for either to be negative.
3564 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3565 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3568 PH/36 The change PH/12 above was broken. Fixed it.
3570 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3571 the grounds that routing the same address twice would always produce the
3572 same answer. This might have been true once, but it is certainly no
3573 longer true now. Routing a child address may depend on the previous
3574 routing that produced that child. Some complicated redirection strategies
3575 went wrong when messages had multiple recipients, and made Exim's
3576 behaviour dependent on the order in which the addresses were given.
3578 I have moved the duplicate checking until after the routing is complete.
3579 Exim scans the addresses that are assigned to local and remote
3580 transports, and removes any duplicates. This means that more work will be
3581 done, as duplicates will always all be routed, but duplicates are
3582 presumably rare, so I don't expect this is of any significance.
3584 For deliveries to pipes, files, and autoreplies, the duplicate checking
3585 still happens during the routing process, since they are not going to be
3588 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3589 It corrects a timeout issue with spamd. This is Ian's comment: "The
3590 background is that sometimes spamd either never reads data from a
3591 connection it has accepted, or it never writes response data. The exiscan
3592 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3593 blindly assumes that writes won't block so it may never time out."
3595 PH/39 Allow G after quota size as well as K and M.
3597 PH/40 The value set for $authenticated_id in an authenticator may not contain
3598 binary zeroes or newlines because the value is written to log lines and
3599 to spool files. There was no check on this. Now the value is run through
3600 the string_printing() function so that such characters are converted to
3601 printable escape sequences.
3603 PH/41 $message_linecount is a new variable that contains the total number of
3604 lines in the message. Compare $body_linecount, which is the count for the
3607 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3608 bounce and delay warning messages, except in certain special cases, which
3611 (a) An SMTP error message from a remote host;
3612 (b) A message specified in a :fail: redirection;
3613 (c) A message specified in a "fail" command in a system filter;
3614 (d) A message specified in a FAIL return from the queryprogram router;
3615 (e) A message specified by the cannot_route_message router option.
3617 In these cases only, Exim does include the error details in bounce and
3618 warning messages. There are also a few cases where bland messages such
3619 as "unrouteable address" or "local delivery error" are given.
3621 PH/43 $value is now also set for the "else" part of a ${run expansion.
3623 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3624 being worked on, but at least Exim now implements the latest version to
3627 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3628 process to complete was treated as a delivery failure, a timeout while
3629 writing the message to the pipe was logged, but erroneously treated as a
3630 successful delivery. Such timeouts include transport filter timeouts. For
3631 consistency with the overall process timeout, these timeouts are now
3632 treated as errors, giving rise to delivery failures by default. However,
3633 there is now a new Boolean option for the pipe transport called
3634 timeout_defer, which, if set TRUE, converts the failures into defers for
3635 both kinds of timeout. A transport filter timeout is now identified in
3638 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3639 systems where "make" and "gmake" are different, calling "gmake" at top
3640 level broke things. I've arranged for the value of $(MAKE) to be passed
3641 from the Makefile to this script so that it can call the same version of
3645 A note about Exim versions 4.44 and 4.50
3646 ----------------------------------------
3648 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3649 changes of various kinds. As a consequence, a big documentation update was
3650 needed. This delayed the release for rather longer than seemed good, especially
3651 in the light of a couple of (minor) security issues. Therefore, the changes
3652 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3653 release. So 4.44 and 4.50 are in effect two different branches that both start
3656 I have left the 4.50 change log unchanged; it contains all the changes since
3657 4.43. The change log for 4.44 is below; many of its items are identical to
3658 those for 4.50. This seems to be the most sensible way to preserve the
3659 historical information.
3665 1. Minor wording change to the doc/README.SIEVE file.
3667 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3668 computation of the current number of files was incorrect.
3670 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3671 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3674 4. Give more explanation in the error message when the command for a transport
3675 filter fails to execute.
3677 5. There are several places where Exim runs a non-Exim command in a
3678 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3679 was being done only for the command run by the queryprogram router. It is
3680 now done for all such subprocesses. The other cases are: ${run, transport
3681 filters, and the commands run by the lmtp and pipe transports.
3683 6. Added CONFIGURE_GROUP build-time option.
3685 7. Some older OS have a limit of 256 on the maximum number of file
3686 descriptors. Exim was using setrlimit() to set 1000 as a large value
3687 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3688 systems. I've change it so that if it can't get 1000, it tries for 256.
3690 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3691 was an oversight, and furthermore, ever since the addition of extra
3692 controls (e.g. 4.43/32), the checks on when to allow different forms of
3693 "control" were broken. There should now be diagnostics for all cases when a
3694 control that does not make sense is encountered.
3696 9. Added the /retain_sender option to "control=submission".
3698 10. $recipients is now available in the predata ACL (oversight).
3700 11. Tidy the search cache before the fork to do a delivery from a message
3701 received from the command line. Otherwise the child will trigger a lookup
3702 failure and thereby defer the delivery if it tries to use (for example) a
3703 cached ldap connection that the parent has called unbind on.
3705 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3706 of $address_data from the recipient verification was clobbered by the
3707 sender verification.
3709 13. The value of address_data from a sender verification is now available in
3710 $sender_address_data in subsequent conditions in the ACL statement.
3712 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3714 15. Added a new option "connect=<time>" to callout options, to set a different
3717 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3718 was its contents. (It was OK if the option was not defined at all.)
3720 17. A "Completed" log line is now written for messages that are removed from
3721 the spool by the -Mrm option.
3723 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3724 information about exactly what failed.
3726 19. Added -dd to debug only the daemon process.
3728 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3729 handles timeouts, both on the server side and network timeouts. Renamed the
3730 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3732 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3733 It is now set to "smtps".
3735 22. $host_address is now set to the target address during the checking of
3736 ignore_target_hosts.
3738 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3739 being passed; this would have caused $sender_host_name to have been used if
3740 matching the list had actually called for a host name (not very likely,
3741 since this list is usually IP addresses). A host name is now passed as
3744 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3745 code that specifies a non-restarting handler (typically sigaction() in
3746 modern systems) in an attempt to fix a rare and obscure crash bug.
3748 25. Narrowed the window for a race in the daemon that could cause it to ignore
3749 SIGCHLD signals. This is not a major problem, because they are used only to
3750 wake it up if nothing else does.
3752 26. A malformed maildirsize file could cause Exim to calculate negative values
3753 for the mailbox size or file count. Odd effects could occur as a result.
3754 The maildirsize information is now recalculated if the size or filecount
3757 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3758 support for a long time. Removed HAVE_SYS_VFS_H.
3760 28. Installed the latest version of exipick from John Jetmore.
3762 29. In an address list, if the pattern was not a regular expression, an empty
3763 subject address (from a bounce message) matched only if the pattern was an
3764 empty string. Non-empty patterns were not even tested. This was the wrong
3765 because it is perfectly reasonable to use an empty address as part of a
3766 database query. An empty address is now tested by patterns that are
3767 lookups. However, all the other forms of pattern expect the subject to
3768 contain a local part and a domain, and therefore, for them, an empty
3769 address still always fails if the pattern is not itself empty.
3771 30. Exim went into a mad DNS loop when attempting to do a callout where the
3772 host was specified on an smtp transport, and looking it up yielded more
3773 than one IP address.
3775 31. Re-factored the code for checking spool and log partition space into a
3776 function that finds that data and another that does the check. The former
3777 is then used to implement four new variables: $spool_space, $log_space,
3778 $spool_inodes, and $log_inodes.
3780 32. The RFC2047 encoding function was originally intended for short strings
3781 such as real names; it was not keeping to the 75-character limit for
3782 encoded words that the RFC imposes. It now respects the limit, and
3783 generates multiple encoded words if necessary. To be on the safe side, I
3784 have increased the buffer size for the ${rfc2047: expansion operator from
3787 33. It is now permitted to omit both strings after an "if" condition; if the
3788 condition is true, the result is "true". As before, when the second string
3789 is omitted, a false condition yields an empty string. This makes it less
3790 cumbersome to write custom ACL and router conditions.
3792 34. Failure to deliver a bounce message always caused it to be frozen, even if
3793 there was an errors_to setting on the router. The errors_to setting is now
3796 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3797 canonical form (fully expanded) before being placed in
3798 $sender_host_address.
3800 36. The table in the code that translates DNS record types into text (T_A to
3801 "A" for instance) was missing entries for NS and CNAME. It is just possible
3802 that this could have caused confusion if both these types were looked up
3803 for the same domain, because the text type is used as part of Exim's
3804 per-process caching. But the chance of anyone hitting this buglet seems
3807 37. The dnsdb lookup has been extended in a number of ways.
3809 (1) There is a new type, "zns", which walks up the domain tree until it
3810 finds some nameserver records. It should be used with care.
3812 (2) There is a new type, "mxh", which is like "mx" except that it returns
3813 just the host names, not the priorities.
3815 (3) It is now possible to give a list of domains (or IP addresses) to be
3816 looked up. The behaviour when one of the lookups defers can be
3817 controlled by a keyword.
3819 (4) It is now possible to specify the separator character for use when
3820 multiple records are returned.
3822 38. The dnslists ACL condition has been extended: it is now possible to supply
3823 a list of IP addresses and/or domains to be looked up in a particular DNS
3826 39. Added log_selector=+queue_time_overall.
3828 40. When running the queue in the test harness, wait just a tad after forking a
3829 delivery process, to get repeatability of debugging output.
3831 41. Include certificate and key file names in error message when GnuTLS fails
3832 to set them up, because the GnuTLS error message doesn't include the name
3833 of the failing file when there is a problem reading it.
3835 42. Allow both -bf and -bF in the same test run.
3837 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3839 44. The "Exiscan patch" is now merged into the mainline Exim source.
3841 45. Sometimes the final signoff response after QUIT could fail to get
3842 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3843 < 0 before doing a fflush(). This bug looks as though it goes back to the
3844 introduction of TLS in release 3.20, but "sometimes" must have been rare
3845 because the tests only now provoked it.
3847 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3848 (this can affect the format of dates).
3850 47. exim_tidydb, when checking for the continued existence of a message for
3851 which it has found a message-specific retry record, was not finding
3852 messages that were in split spool directories. Consequently, it was
3853 deleting retry records that should have stayed in existence.
3855 48. Steve fixed some bugs in eximstats.
3857 49. The SPA authentication driver was not abandoning authentication and moving
3858 on to the next authenticator when an expansion was forced to fail,
3859 contradicting the general specification for all authenticators. Instead it
3860 was generating a temporary error. It now behaves as specified.
3862 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3863 (the order specifies the preference for clients). The order is now AES256,
3864 AES128, 3DES, ARCFOUR128.
3866 51. Small patch to Sieve code - explicitly set From: when generating an
3869 52. Exim crashed if a remote delivery caused a very long error message to be
3870 recorded - for instance if somebody sent an entire SpamAssassin report back
3871 as a large number of 550 error lines. This bug was coincidentally fixed by
3872 increasing the size of one of Exim's internal buffers (big_buffer) that
3873 happened as part of the Exiscan merge. However, to be on the safe side, I
3874 have made the code more robust (and fixed the comments that describe what
3877 53. Now that there can be additional text after "Completed" in log lines (if
3878 the queue_time_overall log selector is set), a one-byte patch to exigrep
3879 was needed to allow it to recognize "Completed" as not the last thing in
3882 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3883 patch that reportedly fixes this has been added. I am not expert enough to
3884 create a test for it. This is what the patch creator wrote:
3886 "I found a little strange behaviour of ldap code when working with
3887 Windows 2003 AD Domain, where users was placed in more than one
3888 Organization Units. When I tried to give exim partial DN, the exit code
3889 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3890 But simultaneously result of request was absolutely normal ldap result,
3891 so I produce this patch..."
3893 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3894 REFERENCE, so I have modified the code to exclude the patch when that macro
3897 55. Some experimental protocols are using DNS PTR records for new purposes. The
3898 keys for these records are domain names, not reversed IP addresses. The
3899 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3900 leaves it alone. Component reversal etc. now happens only for IP addresses.
3903 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3905 57. Double the size of the debug message buffer (to 2048) so that more of very
3906 long debug lines gets shown.
3908 58. The exicyclog utility now does better if the number of log files to keep
3909 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3911 59. Two changes related to the smtp_active_hostname option:
3913 (1) $smtp_active_hostname is now available as a variable.
3914 (2) The default for smtp_banner uses $smtp_active_hostname instead
3915 of $primary_hostname.
3917 60. The host_aton() function is supposed to be passed a string that is known
3918 to be a valid IP address. However, in the case of IPv6 addresses, it was
3919 not checking this. This is a hostage to fortune. Exim now panics and dies
3920 if the condition is not met. A case was found where this could be provoked
3921 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3922 components; fortuitously, this particular loophole had already been fixed
3923 by change 4.50/55 above.
3925 If there are any other similar loopholes, the new check in host_aton()
3926 itself should stop them being exploited. The report I received stated that
3927 data on the command line could provoke the exploit when Exim was running as
3928 exim, but did not say which command line option was involved. All I could
3929 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3930 running as the user.
3933 61. There was a buffer overflow vulnerability in the SPA authentication code
3934 (which came originally from the Samba project). I have added a test to the
3935 spa_base64_to_bits() function which I hope fixes it.
3938 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3939 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3941 63. The daemon start-up calls getloadavg() while still root for those OS that
3942 need the first call to be done as root, but it missed one case: when
3943 deliver_queue_load_max is set with deliver_drop_privilege. This is
3944 necessary for the benefit of the queue runner, because there is no re-exec
3945 when deliver_drop_privilege is set.
3947 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3948 This has been fixed.
3950 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3951 was in use, was not putting the data itself into the right store pool;
3952 consequently, it could be overwritten for a subsequent message in the same
3953 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3956 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3958 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3959 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3960 the function were treating the return as a boolean value, which happened to
3961 work because 0=false and not-0=true, but is not correct code.
3963 68. The host_aton() function was not handling scoped IPv6 addresses (those
3964 with, for example, "%eth0" on the end) correctly.
3966 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3967 negated items (that is, ~something) in unsigned ints. Some compilers
3968 apparently mutter when there is no cast.
3970 70. If an address verification called from an ACL failed, and did not produce a
3971 user-specific message (i.e. there was only a "system" message), nothing was
3972 put in $acl_verify_message. In this situation, it now puts the system
3975 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3976 session; change 4.31/43 added the unwanted input to the log line - except
3977 that it did not do this in the start of session case. It now does.
3979 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3980 This is wrong and can cause the other end to generate a synchronization
3981 error if it is another Exim or anything else that does the synchronization
3982 check. A QUIT command is no longer sent after a timeout.
3984 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3985 during host lookups.
3987 74. The defer_ok option of callout verification was not working if it was used
3988 when verifying addresses in header lines, that is, for this case:
3990 verify = header_sender/callout=defer_ok
3992 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3993 those file descriptors could be used for SMTP connections. If anything
3994 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3995 could be sent to the SMTP client, causing chaos. The daemon now opens
3996 stdin, stdout, and stderr to /dev/null when it puts itself into the
3999 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4000 log by default. The user can override this with suitable Perl magic.
4002 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4003 the log message when discard triggers, was not working for the DATA ACL or
4004 for the non-SMTP ACL.
4006 78. Error message wording change in sieve.c.
4008 79. If smtp_accept_max_per_host was set, the number of connections could be
4009 restricted to fewer than expected, because the daemon was trying to set up
4010 a new connection before checking whether the processes handling previous
4011 connections had finished. The check for completed processes is now done
4012 earlier. On busy systems, this bug wouldn't be noticed because something
4013 else would have woken the daemon, and it would have reaped the completed
4016 80. If a message was submitted locally by a user whose login name contained one
4017 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4018 It caused a spool format error. I have fixed the spool reading code. A
4019 related problem was that the "from" clause in the Received: line became
4020 illegal because of the space(s). It is now covered by ${quote_local_part.
4022 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4025 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4027 83. Patch from Sieve maintainer to fix unterminated string problem in
4028 "vacation" handling.
4030 84. Some minor changes to the Linux configuration files to help with other
4031 OS variants using glibc.
4033 85. One more patch for Sieve to update vacation handling to latest spec.
4036 ----------------------------------------------------
4037 See the note above about the 4.44 and 4.50 releases.
4038 ----------------------------------------------------
4044 1. Change 4.43/35 introduced a bug that caused file counts to be
4045 incorrectly computed when quota_filecount was set in an appendfile
4048 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4049 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4052 3. Give more explanation in the error message when the command for a transport
4053 filter fails to execute.
4055 4. There are several places where Exim runs a non-Exim command in a
4056 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4057 was being done only for the command run by the queryprogram router. It is
4058 now done for all such subprocesses. The other cases are: ${run, transport
4059 filters, and the commands run by the lmtp and pipe transports.
4061 5. Some older OS have a limit of 256 on the maximum number of file
4062 descriptors. Exim was using setrlimit() to set 1000 as a large value
4063 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4064 systems. I've change it so that if it can't get 1000, it tries for 256.
4066 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4067 was an oversight, and furthermore, ever since the addition of extra
4068 controls (e.g. 4.43/32), the checks on when to allow different forms of
4069 "control" were broken. There should now be diagnostics for all cases when a
4070 control that does not make sense is encountered.
4072 7. $recipients is now available in the predata ACL (oversight).
4074 8. Tidy the search cache before the fork to do a delivery from a message
4075 received from the command line. Otherwise the child will trigger a lookup
4076 failure and thereby defer the delivery if it tries to use (for example) a
4077 cached ldap connection that the parent has called unbind on.
4079 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4080 of $address_data from the recipient verification was clobbered by the
4081 sender verification.
4083 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4084 was its contents. (It was OK if the option was not defined at all.)
4086 11. A "Completed" log line is now written for messages that are removed from
4087 the spool by the -Mrm option.
4089 12. $host_address is now set to the target address during the checking of
4090 ignore_target_hosts.
4092 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4093 being passed; this would have caused $sender_host_name to have been used if
4094 matching the list had actually called for a host name (not very likely,
4095 since this list is usually IP addresses). A host name is now passed as
4098 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4099 code that specifies a non-restarting handler (typically sigaction() in
4100 modern systems) in an attempt to fix a rare and obscure crash bug.
4102 15. Narrowed the window for a race in the daemon that could cause it to ignore
4103 SIGCHLD signals. This is not a major problem, because they are used only to
4104 wake it up if nothing else does.
4106 16. A malformed maildirsize file could cause Exim to calculate negative values
4107 for the mailbox size or file count. Odd effects could occur as a result.
4108 The maildirsize information is now recalculated if the size or filecount
4111 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4112 support for a long time. Removed HAVE_SYS_VFS_H.
4114 18. Updated exipick to current release from John Jetmore.
4116 19. Allow an empty sender to be matched against a lookup in an address list.
4117 Previously the only cases considered were a regular expression, or an
4120 20. Exim went into a mad DNS lookup loop when doing a callout where the
4121 host was specified on the transport, if the DNS lookup yielded more than
4124 21. The RFC2047 encoding function was originally intended for short strings
4125 such as real names; it was not keeping to the 75-character limit for
4126 encoded words that the RFC imposes. It now respects the limit, and
4127 generates multiple encoded words if necessary. To be on the safe side, I
4128 have increased the buffer size for the ${rfc2047: expansion operator from
4131 22. Failure to deliver a bounce message always caused it to be frozen, even if
4132 there was an errors_to setting on the router. The errors_to setting is now
4135 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4136 canonical form (fully expanded) before being placed in
4137 $sender_host_address.
4139 24. Updated eximstats to version 1.33
4141 25. Include certificate and key file names in error message when GnuTLS fails
4142 to set them up, because the GnuTLS error message doesn't include the name
4143 of the failing file when there is a problem reading it.
4145 26. Expand error message when OpenSSL has problems setting up cert/key files.
4148 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4149 (this can affect the format of dates).
4151 28. exim_tidydb, when checking for the continued existence of a message for
4152 which it has found a message-specific retry record, was not finding
4153 messages that were in split spool directories. Consequently, it was
4154 deleting retry records that should have stayed in existence.
4156 29. eximstats updated to version 1.35
4157 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4158 1.35 - bugfix such that pie charts by volume are generated correctly
4160 30. The SPA authentication driver was not abandoning authentication and moving
4161 on to the next authenticator when an expansion was forced to fail,
4162 contradicting the general specification for all authenticators. Instead it
4163 was generating a temporary error. It now behaves as specified.
4165 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4166 (the order specifies the preference for clients). The order is now AES256,
4167 AES128, 3DES, ARCFOUR128.
4169 31. Small patch to Sieve code - explicitly set From: when generating an
4172 32. Exim crashed if a remote delivery caused a very long error message to be
4173 recorded - for instance if somebody sent an entire SpamAssassin report back
4174 as a large number of 550 error lines. This bug was coincidentally fixed by
4175 increasing the size of one of Exim's internal buffers (big_buffer) that
4176 happened as part of the Exiscan merge. However, to be on the safe side, I
4177 have made the code more robust (and fixed the comments that describe what
4180 33. Some experimental protocols are using DNS PTR records for new purposes. The
4181 keys for these records are domain names, not reversed IP addresses. The
4182 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4183 leaves it alone. Component reversal etc. now happens only for IP addresses.
4186 34. The host_aton() function is supposed to be passed a string that is known
4187 to be a valid IP address. However, in the case of IPv6 addresses, it was
4188 not checking this. This is a hostage to fortune. Exim now panics and dies
4189 if the condition is not met. A case was found where this could be provoked
4190 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4191 components; fortuitously, this particular loophole had already been fixed
4192 by change 4.50/55 or 4.44/33 above.
4194 If there are any other similar loopholes, the new check in host_aton()
4195 itself should stop them being exploited. The report I received stated that
4196 data on the command line could provoke the exploit when Exim was running as
4197 exim, but did not say which command line option was involved. All I could
4198 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4199 running as the user.
4202 35. There was a buffer overflow vulnerability in the SPA authentication code
4203 (which came originally from the Samba project). I have added a test to the
4204 spa_base64_to_bits() function which I hope fixes it.
4207 36. The daemon start-up calls getloadavg() while still root for those OS that
4208 need the first call to be done as root, but it missed one case: when
4209 deliver_queue_load_max is set with deliver_drop_privilege. This is
4210 necessary for the benefit of the queue runner, because there is no re-exec
4211 when deliver_drop_privilege is set.
4213 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4214 was in use, was not putting the data itself into the right store pool;
4215 consequently, it could be overwritten for a subsequent message in the same
4216 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4219 38. Sometimes the final signoff response after QUIT could fail to get
4220 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4221 < 0 before doing a fflush(). This bug looks as though it goes back to the
4222 introduction of TLS in release 3.20, but "sometimes" must have been rare
4223 because the tests only now provoked it.
4229 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4230 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4231 Now it returns an integer. A number of calls were still expecting a T/F
4232 return. Fortuitously, in all cases, the tests worked in OK situations,
4233 which is the norm. However, things would have gone wrong on any write
4234 failures on the smtp file descriptor. This function is used when sending
4235 messages over SMTP and also when doing verify callouts.
4237 2. When Exim is called to do synchronous delivery of a locally submitted
4238 message (the -odf or -odi options), it no longer closes stderr before doing
4241 3. Implemented the mua_wrapper option.
4243 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4245 5. Implemented the functions header_remove(), header_testname(),
4246 header_add_at_position(), and receive_remove_recipient(), and exported them
4249 6. If an ACL "warn" statement specified the addition of headers, Exim already
4250 inserted X-ACL-Warn: at the start if there was no header name. However, it
4251 was not making this test for the second and subsequent header lines if
4252 there were newlines in the string. This meant that an invalid header could
4253 be inserted if Exim was badly configured.
4255 7. Allow an ACL "warn" statement to add header lines at the start or after all
4256 the Received: headers, as well as at the end.
4258 8. Added the rcpt_4xx retry error code.
4260 9. Added postmaster_mailfrom=xxx to callout verification option.
4262 10. Added mailfrom=xxxx to the callout verification option, for verify=
4265 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4266 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4268 12. Inserted some casts to stop certain compilers warning when using pointer
4269 differences as field lengths or precisions in printf-type calls (mostly
4270 affecting debugging statements).
4272 13. Added optional readline() support for -be (dynamically loaded).
4274 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4275 same clock tick as a message's arrival, so that its received time was the
4276 same as the "first fail" time on the retry record, and that message
4277 remained on the queue past the ultimate address timeout, every queue runner
4278 would try a delivery (because it was past the ultimate address timeout) but
4279 after another failure, the ultimate address timeout, which should have then
4280 bounced the address, did not kick in. This was a "< instead of <=" error;
4281 in most cases the first failure would have been in the next clock tick
4282 after the received time, and all would be well.
4284 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4285 being recognized when the domain list was tested by the match_domain
4286 condition in an expansion string.
4288 16. Added the ${str2b64: operator.
4290 17. Exim was always calling setrlimit() to set a large limit for the number of
4291 processes, without checking whether the existing limit was already
4292 adequate. (It did check for the limit on file descriptors.) Furthermore,
4293 errors from getrlimit() and setrlimit() were being ignored. Now they are
4294 logged to the main and panic logs, but Exim does carry on, to try to do its
4295 job under whatever limits there are.
4297 18. Imported PCRE 5.0.
4299 19. Trivial typo in log message " temporarily refused connection" (the leading
4302 20. If the log selector return_path_on_delivery was set and an address was
4303 redirected to /dev/null, the delivery process crashed because it assumed
4304 that a return path would always be set for a "successful" delivery. In this
4305 case, the whole delivery is bypassed as an optimization, and therefore no
4308 21. Internal re-arrangement: the function for sending a challenge and reading
4309 a response while authentication was assuming a zero-terminated challenge
4310 string. It's now changed to take a pointer and a length, to allow for
4311 binary data in such strings.
4313 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4315 23. Exim was not respecting finduser_retries when seeking the login of the
4316 uid under which it was called; it was always trying 10 times. (The default
4317 setting of finduser_retries is zero.) Also, it was sleeping after the final
4318 failure, which is pointless.
4320 24. Implemented tls_on_connect_ports.
4322 25. Implemented acl_smtp_predata.
4324 26. If the domain in control=submission is set empty, Exim assumes that the
4325 authenticated id is a complete email address when it generates From: or
4326 Sender: header lines.
4328 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4329 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4330 chown and chgrp in /bin and hostname in /usr/bin.
4332 28. Exim was keeping the "process log" file open after each use, just as it
4333 does for the main log. This opens the possibility of it remaining open for
4334 long periods when the USR1 signal hits a daemon. Occasional processlog
4335 errors were reported, that could have been caused by this. Anyway, it seems
4336 much more sensible not to leave this file open at all, so that is what now
4339 29. The long-running daemon process does not normally write to the log once it
4340 has entered its main loop, and it closes the log before doing so. This is
4341 so that log files can straightforwardly be renamed and moved. However,
4342 there are a couple of unusual error situations where the daemon does write
4343 log entries, and I had neglected to close the log afterwards.
4345 30. The text of an SMTP error response that was received during a remote
4346 delivery was being truncated at 512 bytes. This is too short for some of
4347 the long messages that one sometimes sees. I've increased the limit to
4350 31. It is now possible to make retry rules that apply only when a message has a
4351 specific sender, in particular, an empty sender.
4353 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4354 it possible to be selective about when SMTP synchronization is enforced.
4356 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4358 32. Implemented hosts_connection_nolog.
4360 33. Added an ACL for QUIT.
4362 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4365 35. Added mailbox_size and mailbox_filecount to appendfile.
4367 36. Added control = no_multiline_responses to ACLs.
4369 37. There was a bug in the logic of the code that waits for the clock to tick
4370 in the case where the clock went backwards by a substantial amount such
4371 that the microsecond fraction of "now" was more than the microsecond
4372 fraction of "then" (but the whole seconds number was less).
4374 38. Added support for the libradius Radius client library this is found on
4375 FreeBSD (previously only the radiusclient library was supported).
4381 1. When certain lookups returned multiple values in the form name=value, the
4382 quoting of the values was not always being done properly. Specifically:
4383 (a) If the value started with a double quote, but contained no whitespace,
4385 (b) If the value contained whitespace other than a space character (i.e.
4386 tabs or newlines or carriage returns) it was not quoted.
4387 This fix has been applied to the mysql and pgsql lookups by writing a
4388 separate quoting function and calling it from the lookup code. The fix
4389 should probably also be applied to nisplus, ibase and oracle lookups, but
4390 since I cannot test any of those, I have not disturbed their existing code.
4392 2. A hit in the callout cache for a specific address caused a log line with no
4393 reason for rejecting RCPT. Now it says "Previous (cached) callout
4394 verification failure".
4396 3. There was an off-by-one bug in the queryprogram router. An over-long
4397 return line was truncated at 256 instead of 255 characters, thereby
4398 overflowing its buffer with the terminating zero. As well as fixing this, I
4399 have increased the buffer size to 1024 (and made a note to document this).
4401 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4402 when Exim is waiting for an SMTP response from a remote server, Exim
4403 restarts its select() call on the socket, thereby resetting its timeout.
4404 This is not a problem when such interrupts are rare. Somebody set up a cron
4405 job to run exiwhat every 2 minutes, which is less than the normal select()
4406 timeout (5 or 10 minutes). This meant that the select() timeout never
4407 kicked in because it was always reset. I have fixed this by comparing the
4408 time when an interrupt arrives with the time at the start of the first call
4409 to select(). If more time than the timeout has elapsed, the interrupt is
4410 treated as a timeout.
4412 5. Some internal re-factoring in preparation for the addition of Sieve
4413 extensions (by MH). In particular, the "personal" test is moved to a
4414 separate function, and given an option for scanning Cc: and Bcc: (which is
4415 not set for Exim filters).
4417 6. When Exim created an email address using the login of the caller as the
4418 local part (e.g. when creating a From: or Sender: header line), it was not
4419 quoting the local part when it contained special characters such as @.
4421 7. Installed new OpenBSD configuration files.
4423 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4424 try to make them clearer.
4426 9. Callout options, other than the timeout value, were being ignored when
4427 verifying sender addresses in header lines. For example, when using
4429 verify = header_sender/callout=no_cache
4431 the cache was (incorrectly) being used.
4433 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4434 only the Cygwin environment.
4436 11. When return_path_on_delivery was set as a log selector, if different remote
4437 addresses in the same message used different return paths and parallel
4438 remote delivery occurred, the wrong values would sometimes be logged.
4439 (Whenever a remote delivery process finished, the return path value from
4440 the most recently started remote delivery process was logged.)
4442 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4443 lines when AUTH and/or TLS are in use. This is the "received protocol"
4444 field. Exim used to use "asmtp" for authenticated SMTP, without any
4445 indication (in the protocol name) for TLS use. Now it follows the RFC and
4446 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4447 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4448 names appear in log lines as well as in Received: header lines.
4450 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4451 extensions, and comparison tests, and to fix some bugs.
4453 14. Changes to the "personal" filter test:
4455 (1) The test was buggy in that it was just doing the equivalent of
4456 "contains" tests on header lines. For example, if a user's address was
4457 anne@some.where, the "personal" test would incorrectly be true for
4459 To: susanne@some.where
4461 This test is now done by extracting each address from the header in turn,
4462 and checking the entire address. Other tests that are part of "personal"
4463 are now done using regular expressions (for example, to check local parts
4464 of addresses in From: header lines).
4466 (2) The list of non-personal local parts in From: addresses has been
4467 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4468 taken from the Sieve specification recommendations.
4470 (3) If the message contains any header line starting with "List-" it is
4471 treated as non-personal.
4473 (4) The test for "circular" in the Subject: header line has been removed
4474 because it now seems ill-conceived.
4476 15. Minor typos in src/EDITME comments corrected.
4478 16. Installed latest exipick from John Jetmore.
4480 17. If headers_add on a router specified a text string that was too long for
4481 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4482 of string_sprintf() is now avoided.
4484 18. $message_body_size was not set (it was always zero) when running the DATA
4485 ACL and the local_scan() function.
4487 19. For the "mail" command in an Exim filter, no default was being set for
4488 the once_repeat time, causing a random time value to be used if "once" was
4489 specified. (If the value happened to be <= 0, no repeat happened.) The
4490 default is now 0s, meaning "never repeat". The "vacation" command was OK
4491 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4492 (I found it when inspecting the code).
4494 20. There is now an overall timeout for performing a callout verification. It
4495 defaults to 4 times the callout timeout, which applies to individual SMTP
4496 commands during the callout. The overall timeout applies when there is more
4497 than one host that can be tried. The timeout is checked before trying the
4498 next host. This prevents very long delays if there are a large number of
4499 hosts and all are timing out (e.g. when the network connections are timing
4500 out). The value of the overall timeout can be changed by specifying an
4501 additional sub-option for "callout", called "maxwait". For example:
4503 verify = sender/callout=5s,maxwait=20s
4505 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4506 to the end before writing, but this should make it even safer).
4508 22. Exim was forgetting that it had advertised PIPELINING for the second and
4509 subsequent messages on an SMTP connection. It was also not resetting its
4510 memory on STARTTLS and an internal HELO.
4512 23. When Exim logs an SMTP synchronization error within a session, it now
4513 records whether PIPELINING has been advertised or not.
4515 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4516 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4517 rather than long int.
4519 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4521 26. Added the never_mail option to autoreply.
4527 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4528 crash if the getsockname() call failed; this can happen if a connection is
4529 closed very soon after it is established. The problem was simply in the
4530 order in which certain operations were done, causing Exim to try to write
4531 to the SMTP stream before it had set up the file descriptor. The bug has
4532 been fixed by making things happen in the correct order.
4538 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4539 before the connection was closed, thus losing the rejection response.
4541 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4542 some early Solaris releases, but causes trouble in current releases where
4543 socklen_t is defined.
4545 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4548 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4551 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4552 that releasing the top of it at the end releases what was used for sub-
4553 expansions (unless the block got too big). However, discard this block if
4554 the first thing is a variable or header, so that we can use its block when
4555 it is dynamic (useful for very large $message_headers, for example).
4557 6. Lookups now cache *every* query, not just the most recent. A new, separate
4558 store pool is used for this. It can be recovered when all lookup caches are
4559 flushed. Lookups now release memory at the end of their result strings.
4560 This has involved some general refactoring of the lookup sources.
4562 7. Some code has been added to the store_xxx() functions to reduce the amount
4563 of flapping under certain conditions.
4565 8. log_incoming_interface used to affect only the <= reception log lines. Now
4566 it causes the local interface and port to be added to several more SMTP log
4567 lines, for example "SMTP connection from", and rejection lines.
4569 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4571 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4573 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4574 could be overwritten at the end of the current message (or the start of a
4575 new message if it was set in a HELO ACL). The value is now preserved for
4576 the duration of the SMTP connection.
4578 12. If a transport had a headers_rewrite setting, and a matching header line
4579 contained an unqualified address, that address was qualified, even if it
4580 did not match any rewriting rules. The underlying bug was that the values
4581 of the flags that permit the existence of unqualified sender and recipient
4582 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4583 non-local messages, and by -bnq for local messages) were not being
4584 preserved with the message after it was received.
4586 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4587 "next input=" as part of the text comprising the host identity instead of
4588 the correct text. The code was using the same buffer for two different
4589 strings. However, depending on which order the printing function evaluated
4590 its arguments, the bug did not always show up. Under Linux, for example, my
4591 test suite worked just fine.
4593 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4594 doesn't work with some older versions of Perl. It has been changed to "my",
4595 which in any case is probably the better facility to use.
4597 15. A really picky compiler found some instances of statements for creating
4598 error messages that either had too many or two few arguments for the format
4601 16. The size of the buffer for calls to the DNS resolver has been increased
4602 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4603 for addresses that have a lot of PTR records. This alleviates a problem; it
4604 does not fully solve it.
4606 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4607 buffer now truncates the list and logs the incident, which is the same
4608 action as happens when Exim is looking up a host name and its aliases.
4609 Previously in this situation something unpredictable would happen;
4610 sometimes it was "internal error: store_reset failed".
4612 18. If a server dropped the connection unexpectedly when an Exim client was
4613 using GnuTLS and trying to read a response, the client delivery process
4614 crashed while trying to generate an error log message.
4616 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4617 string, for example:
4619 warn message = H1: something\nH2: something
4621 the text was added as a single header line from Exim's point of view
4622 though it ended up OK in the delivered message. However, searching for the
4623 second and subsequent header lines using $h_h2: did not work. This has been
4624 fixed. Similarly, if a system filter added multiple headers in this way,
4625 the routers could not see them.
4627 20. Expanded the error message when iplsearch is called with an invalid key to
4628 suggest using net-iplsearch in a host list.
4630 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4631 ACLs are no longer actually imposed (and a message to that effect is
4634 22. If a "gecos" field in a passwd entry contained escaped characters, in
4635 particular, if it contained a \" sequence, Exim got it wrong when building
4636 a From: or a Sender: header from that name. A second bug also caused
4637 incorrect handling when an unquoted " was present following a character
4638 that needed quoting.
4640 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4641 was not being matched caselessly.
4643 24. Arranged for all hyphens in the exim.8 source to be escaped with
4646 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4647 properly. Recipient callout cache records were still being keyed to include
4648 the sender, even when use_sender was set false. This led to far more
4649 callouts that were necessary. The sender is no longer included in the key
4650 when use_sender is false.
4652 26. Added "control = submission" modifier to ACLs.
4654 27. Added the ${base62d: operator to decode base 62 numbers.
4656 28. dnsdb lookups can now access SRV records.
4658 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4659 the configuration file.
4661 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4662 (-v) mode. This makes the output for a verbose queue run more intelligible.
4664 31. Added a use_postmaster feature to recipient callouts.
4666 32. Added the $body_zerocount variable, containing the number of binary zero
4667 bytes in the message body.
4669 33. The time of last modification of the "new" subdirectory is now used as the
4670 "mailbox time last read" when there is a quota error for a maildir
4673 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4675 35. Added +ignore_unknown as a special item in host lists.
4677 36. Code for decoding IPv6 addresses in host lists is now included, even if
4678 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4679 address was recognized as an IP address, but was then not correctly decoded
4680 into binary, causing unexpected and incorrect effects when compared with
4687 1. Very minor rewording of debugging text in manualroute to say "list of
4688 hosts" instead of "hostlist".
4690 2. If verify=header_syntax was set, and a header line with an unqualified
4691 address (no domain) and a large number of spaces between the end of the
4692 name and the colon was received, the reception process suffered a buffer
4693 overflow, and (when I tested it) crashed. This was caused by some obsolete
4694 code that should have been removed. The fix is to remove it!
4696 3. When running in the test harness, delay a bit after writing a bounce
4697 message to get a bit more predictability in the log output.
4699 4. Added a call to search_tidyup() just before forking a reception process. In
4700 theory, someone could use a lookup in the expansion of smtp_accept_max_
4701 per_host which, without the tidyup, could leave open a database connection.
4703 5. Added the variables $recipient_data and $sender_data which get set from a
4704 lookup success in an ACL "recipients" or "senders" condition, or a router
4705 "senders" option, similar to $domain_data and $local_part_data.
4707 6. Moved the writing of debug_print from before to after the "senders" test
4710 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4711 problems for message scanning, either using a data ACL, or using
4712 local_scan() because the Received: header was not generated till after they
4713 were called (in order to set the time as the time of reception completion).
4714 I have revised the way this works. The header is now generated after the
4715 body is received, but before the ACL or local_scan() are called. After they
4716 are run, the timestamp in the header is updated.
4722 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4723 before starting a queue runner without re-exec. This happened only when
4724 deliver_drop_privilege was set or when the Exim user was set to root. The
4725 effect of the bug was that timeouts during subsequent deliveries caused
4726 crashes instead of being properly handled. The handler is now left at its
4727 default (and expected) setting.
4729 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4730 message, again when deliver_drop_privilege is set or Exim is run as root.
4731 The bug described in (1) was not present in this case, but the tidying up
4732 of the other signals was missing. I have made the two cases consistent.
4734 3. The ignore_target_hosts setting on a manualroute router was being ignored
4735 for hosts that were looked up using the /MX notation.
4737 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4740 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4741 operated on the sender address. After changing the $sender_address to <>
4742 for the sender address verify, Exim was re-instated it as the original
4743 (before rewriting) address, but remembering that it had rewritten it, so it
4744 wasn't rewriting it again. This bug also had the effect of breaking the
4745 sender address verification caching when the sender address was rewritten.
4747 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4748 This has been changed so that if the ip literal address matches
4749 ignore_target_hosts, the router declines.
4751 7. Added expansion conditions match_domain, match_address, and match_local_
4752 part (NOT match_host).
4754 8. The placeholder for the Received: header didn't have a length field set.
4756 9. Added code to Exim itself and to exim_lock to test for a specific race
4757 condition that could lead to file corruption when using MBX delivery. The
4758 issue is with the lockfile that is created in /tmp. If this file is removed
4759 after a process has opened it but before that process has acquired a lock,
4760 there is the potential for a second process to recreate the file and also
4761 acquire a lock. This could lead to two Exim processes writing to the file
4762 at the same time. The added code performs the same test as UW imapd; it
4763 checks after acquiring the lock that its file descriptor still refers to
4764 the same named file.
4766 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4767 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4770 11. Added the smtp_active_hostname option. If used, this will typically be made
4771 to depend on the incoming interface address. Because $interface_address is
4772 not set up until the daemon has forked a reception process, error responses
4773 that can happen earlier (such as "too many connections") no longer contain
4776 12. If an expansion in a condition on a "warn" statement fails because a lookup
4777 defers, the "warn" statement is abandoned, and the next ACL statement is
4778 processed. Previously this caused the whole ACL to be aborted.
4780 13. Added the iplsearch lookup type.
4782 14. Added ident_timeout as a log selector.
4784 15. Added tls_certificate_verified as a log selector.
4786 16. Added a global option tls_require_ciphers (compare the smtp transport
4787 option of the same name). This controls incoming TLS connections.
4789 17. I finally figured out how to make tls_require_ciphers do a similar thing
4790 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4791 before starting the TLS session.
4793 18. Tabs are now shown as \t in -bP output.
4795 19. If the log selector return_path_on_delivery was set, Exim crashed when
4796 bouncing a message because it had too many Received: header lines.
4798 20. If two routers both had headers_remove settings, and the first one included
4799 a superfluous trailing colon, the final name in the first list and the
4800 first name in the second list were incorrectly joined into one item (with a
4801 colon in the middle).
4807 1. Added -C and -D options to the exinext utility, mainly to make it easier
4808 to include in the automated testing, but these could be helpful when
4809 multiple configurations are in use.
4811 2. The exinext utility was not formatting the output nicely when there was
4812 an alternate port involved in the retry record key, nor when there was a
4813 message id as well (for retries that were specific to a specific message
4814 and a specific host). It was also confused by IPv6 addresses, because of
4815 the additional colons they contain. I have fixed the IPv4 problem, and
4816 patched it up to do a reasonable job for IPv6.
4818 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4819 delivery, the log line now contains "pipelined" if PIPELINING was used.
4821 4. An SMTP transport process used to panic and die if the bind() call to set
4822 an explicit outgoing interface failed. This has been changed; it is now
4823 treated in the same way as a connect() failure.
4825 5. A reference to $sender_host_name in the part of a conditional expansion
4826 that was being skipped was still causing a DNS lookup. This no longer
4829 6. The def: expansion condition was not recognizing references to header lines
4830 that used bh_ and bheader_.
4832 7. Added the _cache feature to named lists.
4834 8. The code for checking quota_filecount in the appendfile transport was
4835 allowing one more file than it should have been.
4837 9. For compatibility with Sendmail, the command line option
4845 and sets the incoming protocol and host name (for trusted callers). The
4846 host name and its colon can be omitted when only the protocol is to be set.
4847 Note the Exim already has two private options, -pd and -ps, that refer to
4848 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4849 "s", but I don't think that's a major issue.
4851 10. A number of refactoring changes to the code, none of which should affect
4854 (a) The number of logging options was getting close to filling up the
4855 32-bit word that was used as a bit map. I have split them into two classes:
4856 those that are passed in the argument to log_write(), and those that are
4857 only ever tested independently outside of that function. These are now in
4858 separate 32-bit words, so there is plenty of room for expansion again.
4859 There is no change in the user interface or the logging behaviour.
4861 (b) When building, for example, log lines, the code previously used a
4862 macro that called string_cat() twice, in order to add two strings. This is
4863 not really sufficiently general. Furthermore, there was one instance where
4864 it was actually wrong because one of the argument was used twice, and in
4865 one call a function was used. (As it happened, calling the function twice
4866 did not affect the overall behaviour.) The macro has been replaced by a
4867 function that can join an arbitrary number of extra strings onto a growing
4870 (c) The code for expansion conditions now uses a table and a binary chop
4871 instead of a serial search (which was left over from when there were very
4872 few conditions). Also, it now recognizes conditions like "pam" even when
4873 the relevant support is not compiled in: a suitably worded error message is
4874 given if an attempt is made to use such a condition.
4876 11. Added ${time_interval:xxxxx}.
4878 12. A bug was causing one of the ddress fields not to be passed back correctly
4879 from remote delivery subprocesses. The field in question was not being
4880 subsequently used, so this caused to problems in practice.
4882 13. Added new log selectors queue_time and deliver_time.
4884 14. Might have fixed a bug in maildirsizefile handling that threw up
4885 "unexpected character" debug warnings, and recalculated the data
4886 unnecessarily. In any case, I expanded the warning message to give more
4889 15. Added the message "Restricted characters in address" to the statements in
4890 the default ACL that block characters like @ and % in local parts.
4892 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4893 Three changes have been made:
4895 (a) There was a serious bug; a negative response to MAIL caused the whole
4896 recipient domain to be cached as invalid, thereby blocking all messages
4897 to all local parts at the same domain, from all senders. This bug has
4898 been fixed. The domain is no longer cached after a negative response to
4899 MAIL if the sender used is not empty.
4901 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4904 (c) A new callout option, "use_sender" has been added for people who want
4905 the modified behaviour.
4911 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4914 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4915 indeed breaks things for older releases.
4917 3. Added additional logging to the case where there is a problem reading data
4918 from a filter that is running in a subprocess using a pipe, in order to
4919 try to track down a specific problem.
4921 4. Testing facility fudge: when running in the test harness and attempting
4922 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4923 getting "No route to host". Convert this to a timeout.
4925 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4928 6. Some OS don't have socklen_t but use size_t instead. This affects the
4929 fifth argument of getsockopt() amongst other things. This is now
4930 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4931 can be set for individual OS. I have set it for SunOS5, OSF1, and
4932 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4933 some earlier ones do not.
4935 7. Change 4.30/15 was not doing the test caselessly.
4937 8. The standard form for an IPv6 address literal was being rejected by address
4938 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4939 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4940 this, as well as the form without the "IPv6" on the front (but only when
4941 address literals are enabled, of course).
4943 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4945 10. Exim crashed if a message with an empty sender address specified by -f
4946 encountered a router with an errors_to setting. This could be provoked only
4947 by a command such as
4951 where an empty string was supplied; "<>" did not hit this bug.
4953 11. Installed PCRE release 4.5.
4955 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4956 remained set. It is now erased.
4958 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4959 times from message ids (which are base 36 rather than the normal 62).
4961 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4962 were being counted as actual protocol errors, and logged if the log
4963 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4964 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4965 and DATA following a set of rejected RCPTs do not count as protocol errors.
4966 In other words, Exim assumes they were pipelined, though this may not
4967 actually be the case. Of course, in all cases the client gets an
4968 appropriate error code.
4970 15. If a lookup fails in an ACL condition, a message about the failure may
4971 be available; it is used if testing the ACL cannot continue, because most
4972 such messages specify what the cause of the deferral is. However, some
4973 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4974 that caused an old message to be retained and used if a later statement
4975 caused a defer, replacing the real cause of the deferral.
4977 16. If an IP address had so many PTR records that the DNS lookup buffer
4978 was not large enough to hold them, Exim could crash while trying to process
4979 the truncated data. It now detects and logs this case.
4981 17. Further to 4.21/58, another change has been made: if (and only if) the
4982 first line of a message (the first header line) ends with CRLF, a bare LF
4983 in a subsequent header line has a space inserted after it, so as not to
4984 terminate the header.
4986 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4987 unnecessarily, used atoi() instead of strtol(), and didn't check the
4988 termination when getting file sizes from file names by regex.
4990 19. Completely re-implemented the support for maildirsize files, in the light
4991 of a number of problems with the previous contributed implementation
4992 (4.30/29). In particular:
4994 . If the quota is zero, the maildirsize file is maintained, but no quota is
4997 . If the maildir directory does not exist, it is created before any attempt
4998 to write a maildirsize file.
5000 . The quota value in the file is just a cache; if the quota is changed in
5001 the transport, the new value overrides.
5003 . A regular expression is available for excluding directories from the
5006 20. The autoreply transport checks the characters in options that define the
5007 message's headers; it allows continued headers, but it was checking with
5008 isspace() after an embedded newline instead of explicitly looking for a
5011 21. If all the "regular" hosts to which an address was routed had passed their
5012 expiry times, and had not reached their retry times, the address was
5013 bounced, even if fallback hosts were defined. Now Exim should go on to try
5016 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5017 equivalent code in the SMTP transport. Some hosts send humungous responses
5018 to HELO/EHLO, more than 1024 it seems.
5020 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5021 gives compiler warnings in some environments. I've now done it "properly",
5024 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5025 (because of problems with the built-in one) was declared to return uschar *
5026 instead of char *, causing compiler failure.
5028 25. Fixed a file descriptor leak when processing alias/forward files.
5030 26. Fixed a minor format string issue in dbfn.c.
5032 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5034 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5035 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5038 29. When a sender address is verified, it is cached, to save repeating the test
5039 when there is more than one recipient in a message. However, when the
5040 verification involves a callout, it is possible for different callout
5041 options to be set for different recipients. It is too complicated to keep
5042 track of this in the cache, so now Exim always runs a verification when a
5043 callout is required, relying on the callout cache for the optimization.
5044 The overhead is duplication of the address routing, but this should not be
5047 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5048 to be verified with callout=postmaster, and the main callout worked but the
5049 postmaster check failed, the verification correctly failed. However, if a
5050 subsequent RCPT command asked for sender verification *without* the
5051 postmaster check, incorrect caching caused this verification also to fail,
5054 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5055 it was not caching the DNS options (qualify_single, search_parents) that
5056 were used when the lookup failed. A subsequent lookup with different
5057 options therefore always gave the same answer, though there were cases
5058 where it should not have. (Example: a "domains = !$mx_any" option on a
5059 dnslookup router: the "domains" option is always processed without any
5060 widening, but the router might have qualify_single set.) Now Exim uses the
5061 cached value only when the same options are set.
5063 32. Added John Jetmore's "exipick" utility to the distribution.
5065 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5066 than a timeout (e.g. a certificate is required, and is not provided), an
5067 Exim server now closes the connection immediately. Previously it waited for
5068 the client to close - but if the client is SSL, it seems that they each
5069 wait for each other, leading to a delay before one of them times out.
5071 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5072 maintained 0.8.x compatibility because I don't think many are using it, and
5073 it is clearly obsolete.
5075 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5078 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5079 client certificate was expired. A simple patch fixes this, though I don't
5080 understand the full logic of why the verify callback is called multiple
5083 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5084 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5085 which causes problems with some clients (such as the Certicom SSL Plus
5086 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5087 disables the coutermeasure allowing Eudora to connect."
5089 38. Exim was not checking that a write() to a log file succeeded. This could
5090 lead to Bad Things if a log got too big, in particular if it hit a file
5091 size limit. Exim now panics and dies if it cannot write to a log file, just
5092 as it does if it cannot open a log file.
5094 39. Modified OS/Makefile-Linux so that it now contains
5096 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5098 The two -D definitions ensure that Exim is compiled with large file
5099 support, which makes it possible to handle log files that are bigger than
5102 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5103 instance) a domain was checked against a named list that involved a lookup,
5104 causing $domain_data to be set, then another domain was checked against the
5105 same list, then the first domain was re-checked, the value of $domain_data
5106 after the final check could be wrong. In particular, if the second check
5107 failed, it could be set empty. This bug probably also applied to
5110 41. The strip_trailing_dot option was not being applied to the address given
5111 with the -f command-line option.
5113 42. The code for reading a message's header from the spool was incrementing
5114 $received_count, but never initializing it. This meant that the value was
5115 incorrect (doubled) while delivering a message in the same process in which
5116 it was received. In the most common configuration of Exim, this never
5117 happens - a fresh exec is done - but it can happen when
5118 deliver_drop_privilege is set.
5120 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5121 it now includes up to 150 characters of the unexpected data in the log
5124 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5125 and building data strings. The size of both of these buffers was 10 000
5126 bytes - far larger than anybody would *ever* want, thought I. Needless to
5127 say, somebody hit the limit. I have increased the maximum line length to
5128 20 000 and the maximum data length of concatenated lines to 100 000. I have
5129 also fixed two bugs, because there was no checking on these buffers. Tsk,
5130 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5131 buffer is too small.
5133 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5134 lsearch lookups. Now it does.
5136 46. When parsing a route_list item in a manualroute router, a fixed-length
5137 buffer was used for the list of hosts. I made this 1024 bytes long,
5138 thinking that nobody would ever have a list of hosts that long. Wrong.
5139 Somebody had a whole pile of complicated expansion conditions, and the
5140 string was silently truncated, leading to an expansion error. It turns out
5141 that it is easier to change to an unlimited length (owing to other changes
5142 that have happened since this code was originally written) than to build
5143 structure for giving a limitation error. The length of the item that
5144 expands into the list of hosts is now unlimited.
5146 47. The lsearch lookup could not handle data where the length of text line was
5147 more than 4095 characters. Such lines were truncated, leading to shortened
5148 data being returned. It should now handle lines of any length.
5150 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5151 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5154 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5155 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5156 no longer happens in comment lines. A list of replacements is now placed
5157 at the head of all of the source files, except those whose only change is
5158 to replace PERL_COMMAND in the very first #! line.
5160 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5161 messages on the queue, with a bottom-up merge sort, using code contributed
5162 by Michael Haardt. This should make operations like -bp somewhat faster on
5163 large queues. It won't affect queue runners, except when queue_run_in_order
5166 51. Installed eximstats 1.31 in the distribution.
5168 52. Added support for SRV lookups to the dnslookup router.
5170 53. If an ACL referred to $message_body or $message_body_end, the value was not
5171 reset for any messages that followed in the same SMTP session.
5173 54. The store-handling optimization for building very long strings was not
5174 differentiating between the different store pools. I don't think this
5175 actually made any difference in practice, but I've tidied it.
5177 55. While running the routers to verify a sender address, $sender_address
5178 was still set to the sender address. This is wrong, because when routing to
5179 send a bounce to the sender, it would be empty. Therefore, I have changed
5180 it so that, while verifying a sender address, $sender_address is set to <>.
5181 (There is no change to what happens when verifying a recipient address.)
5183 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5184 target A or AAAA records (if not already returned) without resetting the
5185 qualify_single or search_parents options of the DNS resolver. These are
5186 inappropriate in this case because the targets of MX and SRV records must
5187 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5188 target that, when qualified, matched something in the local domain. These
5189 two options are now turned off when doing these lookups.
5191 57. It seems that at least some releases of Reiserfs (which does not have the
5192 concept of a fixed number of inodes) returns zero and not -1 for the
5193 number of available inodes. This interacted badly with check_spool_inodes,
5194 which assumed that -1 was the "no such thing" setting. What I have done is
5195 to check that the total number of inodes is greater than zero before doing
5196 the test of how many are available.
5198 58. When a "warn" ACL statement has a log_message modifier, the message is
5199 remembered, and not repeated. This is to avoid a lot of repetition when a
5200 message has many recipients that cause the same warning to be written.
5201 Howewer, Exim was preserving the list of already written lines for an
5202 entire SMTP session, which doesn't seem right. The memory is now reset if a
5203 new message is started.
5205 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5206 debugging output unless log_rewrite was also set.
5208 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5209 of (int)(handle) when we know that handle contains (void *)(-1).
5211 61. The Exim daemon panic-logs an error return when it closes the incoming
5212 connection. However "connection reset by peer" seems to be common, and
5213 isn't really an error worthy of noting specially, so that particular error
5216 62. When Exim is trying to find all the local interfaces, it used to panic and
5217 die if the ioctl to get the interface flags failed. However, it seems that
5218 on at least one OS (Solaris 9) it is possible to have an interface that is
5219 included in the list of interfaces, but for which you get a failure error
5220 for this call. This happens when the interface is not "plumbed" into a
5221 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5222 failure of the "get flags" call assumes that the interface is down.
5224 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5225 makes life easier for people who are doing arithmetic on fields extracted
5226 from dates, where you often get leading zeros that should not be
5227 interpreted as octal.
5229 64. Added qualify_domain to the redirect router, to override the global
5232 65. If a pathologically long header line contained very many addresses (the
5233 report of this problem mentioned 10 000) and each of them was rewritten,
5234 Exim could use up a very large amount of memory. (It kept on making new
5235 copies of the header line as it rewrote, and never released the old ones.)
5236 At the expense of a bit more processing, the header rewriting function has
5237 been changed so that it no longer eats memory in this way.
5239 66. The generation of the Received: header has been moved from the time that a
5240 message starts to be received, to the time that it finishes. The timestamp
5241 in the Received: header should now be very close to that of the <= log
5242 line. There are two side-effects of this change:
5244 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5245 logged header lines no longer include the local Received: line, because
5246 it has not yet been created. The same applies to a copy of the message
5247 that is returned to a non-SMTP sender when a message is rejected.
5249 (b) When a filter file is tested using -bf, no additional Received: header
5250 is added to the test message. After some thought, I decided that this
5253 This change does not affect the value of $received_for. It is still set
5254 after address rewriting, but before local_scan() is called.
5256 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5258 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5259 gave an unhelpful panic error message, and a defer error. I have managed to
5260 change this behaviour so that it now rejects any supplied certificate,
5261 which seems right, as the list of acceptable certificates is empty.
5263 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5264 gave an unhelpful defer error. I have not managed to make this reject any
5265 supplied certificates, but the error message it gives is "no certificate
5266 supplied", which is not helpful.
5268 70. exigrep's output now also includes lines that are not associated with any
5269 message, but which match the given pattern. Implemented by a patch from
5270 Martin Sluka, which also tidied up the Perl a bit.
5272 71. Recipient callout verification, like sender verification, was using <> in
5273 the MAIL FROM command. This isn't really the right thing, since the actual
5274 sender may affect whether the remote host accepts the recipient or not. I
5275 have changed it to use the actual sender in the callout; this means that
5276 the cache record is now keyed on a recipient/sender pair, not just the
5277 recipient address. There doesn't seem to be a real danger of callout loops,
5278 since a callout by the remote host to check the sender would use <>.
5279 [SEE ABOVE: changed after hitting problems.]
5281 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5282 temporary errors. However, in the case of such a code being given after
5283 the end of a data transmission (i.e. after ".") Exim was failing to write
5284 a retry record for the message. (Yes, there was some broken host that was
5285 actually sending 8xx at this point.)
5287 73. An unknown lookup type in a host list could cause Exim to panic-die when
5288 the list was checked. (An example that provoked this was putting <; in the
5289 middle of a list instead of at the start.) If this happened during a DATA
5290 ACL check, a -D file could be left lying around. This kind of configuration
5291 error no longer causes Exim to die; instead it causes a defer errror. The
5292 incident is still logged to the main and panic logs.
5294 74. Buglet left over from Exim 3 conversion. The message "too many messages
5295 in one connection" was written to the rejectlog but not the mainlog, except
5296 when address rewriting (yes!) was being logged.
5298 75. Added write_rejectlog option.
5300 76. When a system filter was run not as root (that is, when system_filter_user
5301 was set), the values of the $n variables were not being returned to the
5302 main process; thus, they were not subsequently available in the $sn
5305 77. Added +return_path_on_delivery log selector.
5307 78. A connection timeout was being treated differently from recipients deferred
5308 when testing hosts_max_try with a message that was older than the host's
5309 retry timeout. (The host should not be counted, thus allowing all hosts to
5310 be tried at least once before bouncing.) This may have been the cause of an
5311 occasionally reported bug whereby a message would remain on the queue
5312 longer than the retry timeout, but would be bounced if a delivery was
5313 forced. I say "may" because I never totally pinned down the problem;
5314 setting up timeout/retry tests is difficult. See also the next item.
5316 79. The ultimate address timeout was not being applied to errors that involved
5317 a combination of host plus message (for example, a timeout on a MAIL
5318 command). When an address resolved to a number of possible hosts, and they
5319 were not all tried for each delivery (e.g. because of hosts_max_try), a
5320 message could remain on the queue longer than the retry timeout.
5322 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5325 81. Fixed an obscure SMTP outgoing bug which required at least the following
5326 conditions: (a) there was another message waiting for the same server;
5327 (b) the server returned 5xx to all RCPT commands in the first message so
5328 that the message was not completed; (c) the server dropped the connection
5329 or gave a negative response to the RSET that Exim sends to abort the
5330 transaction. The observed case was a dropped connection after DATA that had
5331 been sent in pipelining mode. That is, the server had advertised PIPELINING
5332 but was not implementing it correctly. The effect of the bug was incorrect
5333 behaviour, such as trying another host, and this could lead to a crash.
5339 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5340 and daemon.c were passed as pointers to ints; they should have been
5341 pointers to socklen_t variables (which are typically unsigned ints).
5343 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5346 3. Fixed a really odd bug that affected only the testing scheme; patching a
5347 certain fixed string in the binary changed the value of another string that
5348 happened to be identical to the end of the original first string.
5350 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5351 name", it returns that address as the IP address. On some operating
5352 systems (e.g. Solaris), it also passes back the IP address string as the
5353 "host name". However, on others (e.g. Linux), it passes back an empty
5354 string. Exim wasn't checking for this, and was changing the host name to an
5355 empty string, assuming it had been canonicized.
5357 5. Although rare, it is permitted to have more than one PTR record for a given
5358 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5359 all the names associated with an address, because they do in Solaris.
5360 However, it seems that they do not in Linux for data that comes from the
5361 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5362 I found this out when I moved to a new Linux workstation and tried to run
5363 the Exim test suite.
5365 To get round this problem I have changed the code so that it now does its
5366 own call to the DNS to look up PTR records when searching for a host name.
5367 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5368 addresses that are only in /etc/hosts are still found.
5370 This behaviour is, however, controlled by an option called host_lookup_
5371 order, which defaults to "bydns:byaddr". If people want to use the other
5372 order, or indeed, just use one or the other means of lookup, they can
5373 specify it in this variable.
5375 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5376 some operating systems, this comes back from gethostbyaddr() as an empty
5377 string, and this is what Exim used to test for. However, it seems that in
5378 other systems, "." is yielded. Exim now tests for this case too.
5380 7. The values of check_spool_space and check_log_space are now held internally
5381 as a number of kilobytes instead of an absolute number of bytes. If a
5382 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5383 kilobyte. This means that much larger values can be stored.
5385 8. Exim monitor: an attempt to get the action menu when not actually pointing
5386 at a message produces an empty menu entitled "No message selected". This
5387 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5388 no entries in it ("Shell widget menu has zero width and/or height"). So I
5389 have added a single, blank menu entry in this case.
5391 9. Added ${quote_local_part.
5393 10. MIME decoding is now applied to the contents of Subject: header lines when
5396 11. Now that a reference to $sender_host_address automatically causes a reverse
5397 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5398 host lookup before query-style lookups in lists that might use this
5399 variable. This has therefore been abolished, and the "net-" prefix is no
5400 longer necessary for query-style lookups.
5402 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5403 have been a typo for LFLAGS, so it has been changed.
5405 13. The install script calls Exim with "-C /dev/null" in order to find the
5406 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5407 to be output. Howeve, since Exim outputs its version number before the
5408 error, it didn't break the script. It just looked ugly. I fixed this by
5409 always allowing "-C /dev/null" if the caller is root.
5411 14. Ignore overlarge ACL variable number when reading spool file - insurance
5412 against a later release with more variables having written the file.
5414 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5415 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5416 this, as well as the form without the "IPv6" on the front.
5418 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5419 OS/Makefile-Darwin file.
5421 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5422 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5424 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5425 variable. However, some OS use ints for inodes. I've added cast to long int
5426 to get rid of the compiler warning.
5428 19. I had forgotten to lock out "/../" in configuration file names when
5429 ALT_CONFIG_PREFIX was set.
5431 20. Routers used for verification do not need to specify transports. However,
5432 if such a router generated a host list, and callout was configured, Exim
5433 crashed, because it could not find a port number from the (non-existent)
5434 transport. It now assumes port 25 in this circumstance.
5436 21. Added the -t option to exigrep.
5438 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5439 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5440 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5441 them is set, LOOKUP_LSEARCH is forced.
5443 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5445 24. Added sender and host information to the "rejected by local_scan()" log
5446 line; previously there was no indication of these.
5448 25. Added .include_if_exists.
5450 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5451 Linux. It turns out that not all file systems support this. Apparently some
5452 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5453 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5454 not supported on the file descriptor, is now ignored when Exim is trying to
5455 sync a directory. This applies only to Linux.
5457 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5459 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5460 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5463 29. Added support for maildirsize files from supplied patch (modified a bit).
5465 30. The use of :fail: followed by an empty string could lead Exim to respond to
5466 sender verification failures with (e.g.):
5468 550 Verification failed for <xxx>
5469 550 Sender verify failed
5471 where the first response line was missing the '-' that indicates it is not
5472 the final line of the response.
5474 31. The loop for finding the name of the user that called Exim had a hardwired
5475 limit of 10; it now uses the value of finduser_retries, which is used for
5476 all other user lookups.
5478 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5481 33. Exim was neglecting to zero errno before one call of strtol() when
5482 expanding a string and expecting an integer value. On some systems this
5483 resulted in spurious "integer overflow" errors. Also, it was casting the
5484 result into an int without checking.
5486 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5487 did not work. The code looks as if it has *never* worked, though it appears
5488 to have been documented since at least releast 1.62. I have made it work.
5490 35. The "timeout_DNS" error in retry rules, also documented since at least
5491 1.62, also never worked. As it isn't clear exactly what this means, and
5492 clearly it isn't a major issue, I have abolished the feature by treating it
5493 as "timeout", and writing a warning to the main and panic logs.
5495 36. The display of retry rules for -brt wasn't always showing the error code
5498 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5499 timeout_connect_A, timeout_connect_MX.
5501 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5502 to the empty sender.
5504 39. The daemon was not analysing the content of -oX till after it had closed
5505 stderr and disconnected from the controlling terminal. This meant that any
5506 syntax errors were only noted on the panic log, and the return code from
5507 the command was 0. By re-arranging the code a little, I've made the
5508 decoding happen first, so such errors now appear on stderr, and the return
5509 code is 1. However, the actual setting up of the sockets still happens in
5510 the disconnected process, so errors there are still only recorded on the
5513 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5514 connections (as happens on some IP stacks) was logged at start up time as
5515 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5516 it from "IPv6 and IPv4", which means that two separate sockets are being
5519 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5520 says whether AF_INET or AF_INET6 was passed as an argument.
5522 42. Exiwhat output was messed up when time zones were included in log
5525 43. Exiwhat now gives more information about the daemon's listening ports,
5526 and whether -tls-on-connect was used.
5528 44. The "port" option of the smtp transport is now expanded.
5530 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5531 silently ignored. Now an error message is written to the main and panic
5534 46. There's a new ACL modifier called "logwrite" which writes to a log file
5535 as soon as it is encountered.
5537 47. Added $local_user_uid and $local_user_gid at routing time.
5539 48. Exim crashed when trying to verify a sender address that was being
5542 49. Exim was recognizing only a space character after ".include". It now also
5543 recognizes a tab character.
5545 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5546 extracting the relevant information from the specification. The man page no
5547 longer contains scrambled data for the -d option, and I've added a section
5548 at the front about calling Exim under different names.
5550 51. Added "extra_headers" argument to the "mail" command in filter files.
5552 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5555 53. Installed eximstats 1.29.
5557 54. Added transport_filter_timeout as a generic transport option.
5559 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5560 Cc: header lines. This was required by RFC 822, but it not required by RFC
5563 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5564 incoming messages that did not have them. Now it does so only if the
5565 message originates locally, that is, if there is no associated remote host
5566 address. When Resent- header lines are present, this applies to the Resent-
5567 lines rather than the non-Resent- lines.
5569 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5570 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5572 58. Messages for configuration errors now include the name of the main
5573 configuration file - useful now that there may be more than one file in a
5574 list (.included file names were always shown).
5576 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5577 for those rare installations that do not start the daemon as root or run it
5578 setuid root. I've cut out the call to initgroups() if the daemon is not
5581 60. The Exim user and group can now be bound into the binary as text strings
5582 that are looked up at the start of Exim's processing.
5584 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5586 62. Added $mailstore_basename variable.
5588 63. Installed patch to sieve.c from Michael Haardt.
5590 64. When Exim failed to open the panic log after failing to open the main log,
5591 the original message it was trying to log was written to stderr and debug
5592 output, but if they were not available (the usual case in production), it
5593 was lost. Now it is written to syslog before the two lines that record the
5594 failures to open the logs.
5596 65. Users' Exim filters run in subprocesses under the user's uid. It is
5597 possible for a "deliver" command or an alias in a "personal" command to
5598 provoke an address rewrite. If logging of address rewriting is configured,
5599 this fails because the process is not running as root or exim. There may be
5600 a better way of dealing with this, but for the moment (because 4.30 needs
5601 to be released), I have disabled address rewrite logging when running a
5602 filter in a non-root, non-exim process.
5608 1. The buildconfig auxiliary program wasn't quoting the value set for
5609 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5610 not defined. This bug was masked in 4.22 by the effect that was fixed in
5613 2. Some messages that were rejected after a message id was allocated were
5614 shown as "incomplete" by exigrep. It no longer does this for messages that
5615 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5617 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5618 have allow_domain_literals set, the ID did not get logged in the <= line.
5619 Domain literals are now always recognized in Message-ID: header lines.
5621 4. The first argument for a ${extract expansion item is the key name or field
5622 number. Leading and trailing spaces in this item were not being ignored,
5623 causing some misleading effects.
5625 5. When deliver_drop_privilege was set, single queue runner processes started
5626 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5627 same command in the process it spins off) were not dropping privilege.
5629 6. When the daemon running as "exim" started a queue runner, it always
5630 re-executed Exim in the spun-off process. This is a waste of effort when
5631 deliver_drop_privilege is set. The new process now just calls the
5632 queue-runner function directly.
5638 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5641 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5642 ignored. Though the use of -oP was forcing the writing of a pid file, it
5643 was always written to the default place.
5645 3. If the message "no IP address found for host xxxx" is generated during
5646 incoming verification, it is now followed by identification of the incoming
5647 connection (so you can more easily find what provoked it).
5649 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5651 5. Added some features to "harden" Exim a bit more against certain attacks:
5653 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5654 be put in Local/Makefile. This is like the never_users runtime option,
5655 but it cannot be overridden. The default setting is "root".
5657 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5658 prefix string with which any file named in a -C command line option
5661 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5662 is retained for -C and -D only if the caller of Exim is root. Without
5663 it, the exim user may also use -C and -D and retain privilege.
5665 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5666 command line option is disabled.
5668 6. Macro names set by the -D option must start with an upper case letter, just
5669 like macro names defined in the configuration file.
5671 7. Added "dereference=" facility to LDAP.
5673 8. Two instances of the typo "uknown" in the source files are fixed.
5675 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5676 the Configure-Makefile script screwed up while processing it.
5678 10. Incorporated PCRE 4.4.
5680 11. The SMTP synchronization check was not operating right at the start of an
5681 SMTP session. For example, it could not catch a HELO sent before the client
5682 waited for the greeting. There is now a check for outstanding input at the
5683 point when the greeting is written. Because of the duplex, asynchronous
5684 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5685 way, but not yet received, when the check is performed.
5687 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5688 on TCP/IP sockets, because this apparently causes some broken clients to
5691 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5692 unchanged) from the Cygwin maintainer.
5694 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5695 was supported instead of testing for mbx. Effectively a typo.
5697 15. The spa authenticator server code was not checking that the input it
5698 received was valid base64.
5700 16. The debug output line for the "set" modifier in ACLs was not showing the
5701 name of the variable that was being set.
5703 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5705 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5706 lookup on its own. Something else was needed to trigger the lookup. For
5707 example, a match in host_lookup or the need for a host name in a host list.
5708 Now, if $sender_host_name is referenced and the host name has not yet been
5709 looked up, a lookup is performed. If the lookup fails, the variable remains
5710 empty, and $host_lookup_failed is set to "1".
5712 19. Added "eqi" as a case-independent comparison operator.
5714 20. The saslauthd authentication condition could segfault if neither service
5715 nor realm was specified.
5717 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5718 error message that was logged was misleading, and incoming SMTP
5719 connections were dropped. The message is now more accurate, and temporary
5720 errors are given to SMTP connections.
5722 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5723 (except RSET) with a 421 error, until QUIT is received. However, it was
5724 failing to send a response to QUIT.
5726 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5727 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5728 ACL is now run after the helo_try_verify_hosts code.
5730 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5731 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5732 case-independent, so other case variants are also recognized). Apparently
5733 some systems use these upper case variants.
5735 25. If more than two messages were waiting for the same host, and a transport
5736 filter was specified for the transport, Exim sent two messages over the
5737 same TCP/IP connection, and then failed with "socket operation on non-
5738 socket" when it tried to send the third.
5740 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5742 27. The extern definition of crypt16() in expand.c was not being excluded when
5743 the OS had its own crypt16() function.
5745 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5746 as a preferred synonym for return_size_limit, both as an option and as an
5749 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5751 30. Changed the default configuration ACL to relax the local part checking rule
5752 for addresses that are not in any local domains. For these addresses,
5753 slashes and pipe symbols are allowed within local parts, but the sequence
5754 /../ is explicitly forbidden.
5756 31. SPA server authentication was not clearing the challenge buffer before
5759 32. log_message in a "warn" ACL statement was writing to the reject log as
5760 well as to the main log, which contradicts the documentation and doesn't
5761 seem right (because no rejection is happening). So I have stopped it.
5763 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5764 However, I am unable to do any testing of this.
5766 34. Fixed an infelicity in the appendfile transport. When checking directories
5767 for a mailbox, to see if any needed to be created, it was accidentally
5768 using path names with one or more superfluous leading slashes; tracing
5769 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5771 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5772 contents are added to the log line that is written for every discarded
5773 recipient. (Previously a log_message setting was ignored.)
5775 36. The ${quote: operator now quotes the string if it is empty.
5777 37. The install script runs exim in order to find its version number. If for
5778 some reason other than non-existence or emptiness, which it checks, it
5779 could not run './exim', it was installing it with an empty version number,
5780 i.e. as "exim-". This error state is now caught, and the installation is
5783 38. An argument was missing from the function that creates an error message
5784 when Exim fails to connect to the socket for saslauthd authentication.
5785 This could cause Exim to crash, or give a corrupted message.
5787 39. Added isip, isip4, and isip6 to ${if conditions.
5789 40. The ACL variables $acl_xx are now saved with the message, and can be
5790 accessed later in routers, transports, and filters.
5792 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5793 strings in the file are not string-expanded.
5795 42. If a MAIL command specified a SIZE value that was too large to fit into an
5796 int variable, the check against message_size_limit failed. Such values are
5797 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5798 day this will have to be increased, but I don't think I want to be around
5799 when emails are that large.
5806 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5807 iconv() is not standard in FreeBSD.
5809 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5810 IPv6 enabled. The observed symptom was a segmentation fault on return from
5811 the function os_common_find_running_interfaces() in src/os.c.
5813 3. In the check_special_case() function in daemon.c I had used "errno" as an
5814 argument name, which causes warnings on some systems. This was basically a
5815 typo, since it was named "eno" in the comments!
5817 4. The code that waits for the clock to tick (at a resolution of some fraction
5818 of a second) so as to ensure message-id uniqueness was always waiting for
5819 at least one whole tick, when it could have waited for less. [This is
5820 almost certainly not relevant at current processor speeds, where it is
5821 unlikely to ever wait at all. But we try to future-proof.]
5823 5. The function that sleeps for a time interval that includes fractions of a
5824 second contained a race. It did not block SIGALRM between setting the
5825 timer, and suspending (a couple of lines later). If the interval was short
5826 and the sigsuspend() was delayed until after it had expired, the suspension
5827 never ended. On busy systems this could lead to processes getting stuck for
5830 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5831 process, before it forks any delivery processes. The open lookup caching
5832 mechanism meant that the open file or database connection was passed into
5833 the delivery process. The problem was that delivery processes always tidy
5834 up cached lookup data. This could cause a problem for the next delivery
5835 process started by the queue runner, because the external queue runner
5836 process does not know about the closure. So the next delivery process
5837 still has data in the lookup cache. In the case of a file lookup, there was
5838 no problem because closing a file descriptor in a subprocess doesn't affect
5839 the parent. However, if the lookup was caching a connection to a database,
5840 the connection was closed, and the second delivery process was likely to
5841 see errors such as "PGSQL: query failed: server closed the connection
5842 unexpectedly". The problem has been fixed by closing all cached lookups
5843 in a queue runner before running a delivery process.
5845 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5846 seem to have the "const" qualifier which it has on other OS. I've
5849 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5850 *of the same type* (client or server) with the same public name that an
5851 error should be diagnosed.
5853 9. When Exim looked up a host name for an IP address, but failed to find the
5854 original IP address when looking up the host name (a safety check), it
5855 output the message "<ip address> does not match any IP for NULL", which was
5856 confusing, to say the least. The bug was that the host name should have
5857 appeared instead of "NULL".
5859 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5860 user that is built into the binary, and the -C or -D options is used, root
5861 privilege is dropped before the configuration file is read. In addition,
5862 logging is switched to stderr instead of the normal log files. If the
5863 configuration then re-defines the Exim user, the unprivileged environment
5864 is probably not what is expected, so Exim logs a panic warning message (but
5867 However, if deliver_drop_privilege is set, the unprivileged state may well
5868 be exactly what is intended, so the warning has been cut out in that case,
5869 and Exim is allowed to try to write to its normal log files.
5875 1. smtp_return_error_details was not giving details for temporary sender
5876 or receiver verification errors.
5878 2. Diagnose a configuration error if two authenticators have the same public
5881 3. Exim used not to create the message log file for a message until the first
5882 delivery attempt. This could be confusing when incoming messages were held
5883 for policy or load reasons. The message log file is now created at the time
5884 the message is received, and an initial "Received" line is written to it.
5886 4. The automatically generated man page for command line options had a minor
5887 bug that caused no ill effects; however, a more serious problem was that
5888 the procedure for building the man page automatically didn't always
5889 operate. Consequently, release 4.20 contains an out-of-date version. This
5890 shouldn't happen again.
5892 5. When building Exim with embedded Perl support, the script that builds the
5893 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5894 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5896 6. The freeze_tell option was not being used for messages that were frozen on
5897 arrival, either by an ACL or by local_scan().
5899 7. Added the smtp_incomplete_transaction log selector.
5901 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5902 was accepting AUTH without a new EHLO.
5904 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5905 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5908 10. Logging of TCP/IP connections (when configured) now happens in the main
5909 daemon process instead of the child process, so that the TCP/IP connection
5910 count is more accurate (but it can never be perfect).
5912 11. The use of "drop" in a nested ACL was not being handled correctly in the
5913 outer ACL. Now, if condition failure induced by the nested "drop" causes
5914 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5915 or "require"), the connection is dropped.
5917 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5918 that yield "discard" can now be used with an "accept" or a "discard" verb,
5919 but an error is generated for any others (because I can't see a useful way
5920 to define what should happen).
5922 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5923 are now processed in the same way as lines in the Exim configuration file.
5924 In particular, continuation lines are supported.
5926 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5928 15. Added -ti meaning -t -i.
5930 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5931 domains, and warn by logging if others are found.
5933 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5934 returned from GIFCONF when Exim is trying to find the list of interfaces on
5935 a host. The code in os.c has been modified to copy each ifreq to an aligned
5936 structure in all cases.
5938 Also, in some cases, the returned ifreq's were being copied to a 'struct
5939 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5940 means the last couple of bytes of an IPv6 address could be chopped if the
5941 ifreq contained only a normal sockaddr (14 bytes storage).
5943 18. Named domain lists were not supported in the hosts_treat_as_local option.
5944 An entry such as +xxxx was not recognized, and was treated as a literal
5947 19. Ensure that header lines added by a DATA ACL are included in the reject log
5948 if the ACL subsequently rejects the message.
5950 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5951 MD5 (which is deprecated).
5953 21. When testing a filter file using -bf, Exim was writing a message when it
5954 took the sender from a "From " line in the message, but it was not doing so
5955 when it took $return_path from a Return-Path: header line. It now does.
5957 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5958 with a valid header line field name (a series of printing characters
5959 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5961 23. Changed "disc" in the source to "disk" to conform to the documentation and
5962 the book and for uniformity.
5964 24. Ignore Sendmail's -Ooption=value command line item.
5966 25. When execve() failed while trying to run a command in a pipe transport,
5967 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5968 could be confused with a return value of 69 from the command itself. This
5969 has been changed to 127, the value the shell returns if it is asked to run
5970 a non-existent command. The wording for the related log line suggests a
5971 non-existent command as the problem.
5973 26. If received_header_text expands to an empty string, do not add a Received:
5974 header line to the message. (Well, it adds a token one on the spool, but
5975 marks it "old" so that it doesn't get used or transmitted.)
5977 27. Installed eximstats 1.28 (addition of -nt option).
5979 28. There was no check for failure on the call to getsockname() in the daemon
5980 code. This can fail if there is a shortage of resources on the system, with
5981 ENOMEM, for example. A temporary error is now given on failure.
5983 29. Contrary to the C standard, it seems that in some environments, the
5984 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5985 program. Exim now does this explicitly; it affects the formatting of
5986 timestamps using strftime().
5988 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5989 complaints. I've now initialized all the variables, to avoid this.
5991 32. Header lines added by a system filter were not being "seen" during
5992 transport-time rewrites.
5994 33. The info_callback() function passed to OpenSSL is set up with type void
5995 (*)(SSL *, int, int), as described somewhere. However, when calling the
5996 function (actually a macro) that sets it up, the type void(*)() is
5997 expected. I've put in a cast to prevent warnings from picky compilers.
5999 34. If a DNS black list lookup found a CNAME record, but there were no A
6000 records associated with the domain it pointed at, Exim crashed.
6002 35. If a DNS black list lookup returned more than one A record, Exim ignored
6003 all but the first. It now scans all returned addresses if a particular IP
6004 value is being sought. In this situation, the contents of the
6005 $dnslist_value variable are a list of all the addresses, separated by a
6008 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6009 to accept a host name and all its aliases if the forward lookup for any of
6010 them yielded the IP address of the incoming connection. Now it accepts only
6011 those names whose forward lookup yields the correct IP address. Any other
6012 names are discarded. This closes a loophole whereby a rogue DNS
6013 administrator could create reverse DNS records to break through a
6014 wildcarded host restriction in an ACL.
6016 37. If a user filter or a system filter that ran in a subprocess used any of
6017 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6018 the wrong values were passed to the pipe command ($thisaddress had the
6019 value of $0, $0 had the value of $1, etc). This bug was introduced by
6020 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6022 38. Improved the line breaking for long SMTP error messages from ACLs.
6023 Previously, if there was no break point between 40 and 75 characters, Exim
6024 left the rest of the message alone. Two changes have been made: (a) I've
6025 reduced the minimum length to 35 characters; (b) if it can't find a break
6026 point between 35 and 75 characters, it looks ahead and uses the first one
6027 that it finds. This may give the occasional overlong line, but at least the
6028 remaining text gets split now.
6030 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6031 file descriptors might be low, and that setting 1000 would always raise it.
6032 It turns out that in some environments, the limit is already over 1000 and
6033 that lowering it causes trouble. So now Exim takes care not to decrease it.
6035 40. When delivering a message, the value of $return_path is set to $sender_
6036 address at the start of routing (routers may change the value). By an
6037 oversight, this default was not being set up when an address was tested by
6038 -bt or -bv, which affected the outcome if any router or filter referred to
6041 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6042 log only when "message" or "log_message" are set. However, if one of the
6043 conditions was an address verification, or a call to a nested ACL, the
6044 messages generated by the underlying test were being passed through. This
6045 no longer happens. The underlying message is available in $acl_verify_
6046 message for both "message" and "log_message" expansions, so it can be
6047 passed through if needed.
6049 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6050 new expansion $bh_ to give the encoded byte string without charset
6051 translation. Translation happens only if iconv() is available; HAVE_ICONV
6052 indicates this at build time. HEADERS_CHARSET gives the charset to
6053 translate to; headers_charset can change it in the configuration, and
6054 "headers charset" can change it in an individual filter file.
6056 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6057 that creates RFC 2047 encoded "words" labels them as that charset instead
6058 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6059 expansion operator; (ii) when Exim creates a From: line for a local
6060 message; (iii) when a header line is rewritten to include a "phrase" part.
6062 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6063 buggy, causing it to skip the first lines of messages whose message ID
6064 ended in 'D'. This would not have bitten before Exim release 4.14, because
6065 message IDs were unlikely to end in 'D' before then. The effect was to have
6066 incorrect size information for certain domains.
6068 45. #include "config.h" was missing at the start of the crypt16.c module. This
6069 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6072 46. If there was a timeout during a "random" callout check, Exim treated it as
6073 a failure of the random address, and carried on sending RSET and the real
6074 address. If the delay was just some slowness somewhere, the response to the
6075 original RCPT would be taken as a response to RSET and so on, causing
6076 mayhem of various kinds.
6078 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6079 when I implemented it. It didn't allow for the fact that some option values
6080 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6081 the right test for positive values.
6083 48. Domain names in DNS records are case-independent. Exim always looks them up
6084 in lower case. Some resolvers return domain names in exactly the case they
6085 appear in the zone file, that is, they may contain uppercase letters. Not
6086 all resolvers do this - some return always lower case. Exim was treating a
6087 change of case by a resolver as a change of domain, similar to a widening
6088 of a domain abbreviation. This triggered its re-routing code and so it was
6089 trying to route what was effectively the same domain again. This normally
6090 caused routing to fail (because the router wouldn't handle the domain
6091 twice). Now Exim checks for this case specially, and just changes the
6092 casing of the domain that it ultimately uses when it transmits the message
6095 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6098 50. If a filter generated a file delivery with a non-absolute name (possible if
6099 no home directory exists for the router), the forbid_file option was not
6102 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6103 the existing equality matching.
6105 52. Exim was using ints instead of ino_t variables in some places where it was
6106 dealing with inode numbers.
6108 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6109 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6110 called TMPDIR, and if it finds it is different, it changes its value.
6112 54. The smtp_printf() function is now made available to local_scan() so
6113 additional output lines can be written before returning. There is also an
6114 smtp_fflush() function to enable the detection of a dropped connection.
6115 The variables smtp_input and smtp_batched_input are exported to
6118 55. Changed the default runtime configuration: the message "Unknown user"
6119 has been removed from the ACL, and instead placed on the localuser router,
6120 using the cannot_route_message feature. This means that any verification
6121 failures that generate their own messages won't get overridden. Similarly,
6122 the "Unrouteable address" message that was in the ACL for unverifiable
6123 relay addresses has also been removed.
6125 56. Added hosts_avoid_esmtp to the smtp transport.
6127 57. The exicyclog script was not checking for the esoteric option
6128 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6129 will work only if exicyclog is run under the appropriate euid.
6131 58. Following a discussion on the list, the rules by which Exim recognises line
6132 endings on incoming messages have been changed. The -dropcr and drop_cr
6133 options are now no-ops, retained only for backwards compatibility. The
6134 following line terminators are recognized: LF CRLF CR. However, special
6135 processing applies to CR:
6137 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6138 nor a local message in the state where . is a terminator.
6140 (ii) If a bare CR is encountered in a header line, an extra space is added
6141 after the line terminator so as not to end the header. The reasoning
6142 behind this is that bare CRs in header lines are most likely either
6143 to be mistakes, or people trying to play silly games.
6145 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6146 was being incorrectly given as 18 bytes larger than it should have been.
6147 This is a VOB (very old bug).
6149 60. This may never have affected anything current, but just in case it has:
6150 When the local host is found other than at the start of a list of hosts,
6151 the local host, those with the same MX, and any that follow, are discarded.
6152 When the list in question was part of a longer list of hosts, the following
6153 hosts (not currently being processed) were also being discarded. This no
6154 longer happens. I'm not sure if this situation could ever has previously
6157 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6160 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6161 header. This is something that is recommended in a new Internet Draft, and
6162 is something that is documented as being done by Sendmail. There are two
6163 possible values. For messages generated by the autoreply transport, Exim
6166 Auto-Submitted: auto-replied
6168 whereas for all other generated messages (e.g. bounces) it adds
6170 Auto-Submitted: auto-generated
6172 63. The "personal" condition in filters now includes a test for the
6173 Auto-Submitted: header. If it contains the string "auto-" the message it
6174 not considered personal.
6176 64. Added rcpt_include_affixes as a generic transport option.
6178 65. Added queue_only_override (default true).
6180 66. Added the syslog_duplication option.
6182 67. If what should have been the first header line of a message consisted of
6183 a space followed by a colon, Exim was mis-interpreting it as a header line.
6184 It isn't of course - it is syntactically invalid and should therefore be
6185 treated as the start of the message body. The misbehaviour could have
6186 caused a number of strange effects, including loss of data in subsequent
6187 header lines, and spool format errors.
6189 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6190 client host had authenticated. This control can now be exercised by an ACL
6191 for more flexibility.
6193 69. By default, callouts do not happen when testing with -bh. There is now a
6194 variant, -bhc, which does actually run the callout code, including
6195 consulting and updating the callout cache.
6197 70. Added support for saslauthd authentication, courtesy of Alexander
6200 71. If statvfs() failed on the spool or log directories while checking their
6201 size for availability, Exim confusingly gave the error "space shortage".
6202 Furthermore, in debugging mode it crashed with a floating point exception.
6203 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6204 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6205 serious problem, Exim now writes to the main and panic logs when this
6206 happens, with details of the failure. It then refuses to accept the
6207 incoming message, giving the message "spool directory problem" or "log
6208 directory problem" with a 421 code for SMTP messages.
6210 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6211 0, 1, and 2 exist, because some OS complain for execs without them (see
6212 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6213 descriptors. However, the code omitted to check that the open succeeded,
6214 causing mysterious errors if for some reason the permissions on /dev/null
6215 got screwed. Now Exim writes a message to the main and panic logs, and
6216 bombs out if it can't open /dev/null.
6218 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6219 interact so that it is all more flexible. It is supposed to remain
6220 backwards compatible. Also added extra_local_interfaces.
6222 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6223 to bomb out with an assertion failure - to the client this appears as a
6224 connection drop. This problem occurs in the part of the code that was taken
6225 from the Samba project. Fortunately, the assertion is in a very simple
6226 function, so I have fixed this by reproducing the function inline in the
6227 one place where it is called, and arranging for authentication to fail
6228 instead of killing the process with assert().
6230 75. The SPA client code was not working when the server requested OEM rather
6231 than Unicode encoding.
6233 76. Added code to make require_files with a specific uid setting more usable in
6234 the case where statting the file as root fails - usually a non-root-mounted
6235 NFS file system. When this happens and the failure is EACCES, Exim now
6236 forks a subprocess and does the per-uid checking as the relevant uid.
6238 77. Added process_log_path.
6240 78. If log_file_path was not explicitly set, a setting of check_log_space or
6241 check_log_inodes was ignored.
6243 79. If a space check for the spool or log partitions fails, the incident is now
6244 logged. Of course, in the latter case the data may get lost...
6246 80. Added the %p formatting code to string_format() so that it can be used to
6247 print addresses in debug_print(). Adjusted all the address printing in the
6248 debugging in store.c to use %p rather than %d.
6250 81. There was a concern that a line of code in smtp_in.c could overflow a
6251 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6252 initially expressed, the concern was not well-founded, because trailing
6253 spaces are removed early. However, if the trailing spaces were followed by
6254 a NULL, they did not get removed, so the overflow was possible. Two fixes
6257 (a) I re-wrote the offending code in a cleaner fashion.
6258 (b) If an incoming SMTP command contains a NULL character, it is rejected
6261 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6262 runs initgroups(), so that if the Exim user is in any additional groups,
6263 they will be used during message reception.
6269 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob