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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
44 TL/01 When running the test suite, the README says that variables such as
45 no_msglog_check are global and can be placed anywhere in a specific
46 test's script, however it was observed that placement needed to be near
47 the beginning for it to behave that way. Changed the runtest perl
48 script to read through the entire script once to detect and set these
49 variables, reset to the beginning of the script, and then run through
50 the script parsing/test process like normal.
52 TL/02 The BSD's have an arc4random API. One of the functions to induce
53 adding randomness was arc4random_stir(), but it has been removed in
54 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
55 function when detected.
57 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
58 cause callback expansion.
60 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
61 syntax errors in an expansion can be treated as a string instead of
62 logging or causing an error, due to the internal use of bool_lax
63 instead of bool when processing it.
65 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
66 server certificates when making smtp deliveries.
68 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
70 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
72 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
74 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
75 Merged patch from Sebastian Wiedenroth.
77 JH/05 Fix results-pipe from transport process. Several recipients, combined
78 with certificate use, exposed issues where response data items split
79 over buffer boundaries were not parsed properly. This eventually
80 resulted in duplicates being sent. This issue only became common enough
81 to notice due to the introduction of conection certificate information,
82 the item size being so much larger. Found and fixed by Wolfgang Breyha.
84 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
85 size buffer was used, resulting in syntax errors when an expansion
88 JH/07 Add support for directories of certificates when compiled with a GnuTLS
89 version 3.3.6 or later.
91 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
92 is EXPERIMENTAL_EVENT, the main-configuration and transport options
93 both become "event_action", the variables become $event_name, $event_data
94 and $event_defer_errno. There is a new variable $verify_mode, usable in
95 routers, transports and related events. The tls:cert event is now also
96 raised for inbound connections, if the main configuration event_action
99 TL/06 In test suite, disable OCSP for old versions of openssl which contained
100 early OCSP support, but no stapling (appears to be less than 1.0.0).
102 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
103 server certificate names available under the smtp transport option
104 "tls_verify_cert_hostname" now do not permit multi-component wildcard
107 JH/10 Time-related extraction expansions from certificates now use the main
108 option "timezone" setting for output formatting, and are consistent
109 between OpenSSL and GnuTLS compilations. Bug 1541.
111 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
112 encoded parameter in the incoming message. Bug 1558.
114 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
115 include certificate info, eximon was claiming there were spoolfile
118 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
120 JH/14 Log delivery-related information more consistently, using the sequence
121 "H=<name> [<ip>]" wherever possible.
123 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
124 are problematic for Debian distribution, omit them from the release
127 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
129 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
132 JH/17 Fix a null-indirection in certextract expansions when a nondefault
133 output list separator was used.
138 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
139 checkers that were complaining about end of non-void function with no
142 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
143 This was a regression intruduced in 4.83 by another bugfix.
145 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
147 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
148 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
154 TF/01 Correctly close the server side of TLS when forking for delivery.
156 When a message was received over SMTP with TLS, Exim failed to clear up
157 the incoming connection properly after forking off the child process to
158 deliver the message. In some situations the subsequent outgoing
159 delivery connection happened to have the same fd number as the incoming
160 connection previously had. Exim would try to use TLS and fail, logging
161 a "Bad file descriptor" error.
163 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
164 utilities have not been installed.
166 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
167 temporary space as the ACL may create new global variables.
169 TL/01 LDAP support uses per connection or global context settings, depending
170 upon the detected version of the libraries at build time.
172 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
173 to extract and use the src ip:port in logging and expansions as if it
174 were a direct connection from the outside internet. PPv2 support was
175 updated based on HAProxy spec change in May 2014.
177 JH/02 Add ${listextract {number}{list}{success}{fail}}.
179 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
180 Properly escape header and check for NULL return.
182 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
185 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
187 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
188 characters in header names, implemented as a verify condition.
189 Contributed by Michael Fischer v. Mollard.
191 TL/05 Rename SPF condition results err_perm and err_temp to standardized
192 results permerror and temperror. Previous values are deprecated but
193 still accepted. In a future release, err_perm and err_temp will be
194 completely removed, which will be a backward incompatibility if the
195 ACL tests for either of these two old results. Patch contributed by
196 user bes-internal on the mailing list.
198 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
200 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
201 selectors, in both main and reject logs.
203 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
206 JH/07 Add malware type "sock" for talking to simple daemon.
208 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
210 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
211 routers/transports under cutthrough routing.
213 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
214 numbers. Touch up "bool" conditional to keep the same definition.
216 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
218 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
220 JH/12 Expand items in router/transport headers_add or headers_remove lists
221 individually rather than the list as a whole. Bug 1452.
223 Required for reasonable handling of multiple headers_ options when
224 they may be empty; requires that headers_remove items with embedded
225 colons must have them doubled (or the list-separator changed).
227 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
228 view the policy declared in the DMARC record. Currently, $dmarc_status
229 is a combined value of both the record presence and the result of the
232 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
234 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
235 dnslookup router and the smtp transport (applying to the forward
238 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
239 of ldap servers used for a specific lookup. Patch provided by Heiko
242 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
243 New variable $lookup_dnssec_authenticated for observability.
245 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
246 Patch submitted by Lars Timman.
248 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
250 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
251 Requires trusted mode and valid format message id, aborts otherwise.
252 Patch contributed by Heiko Schlichting.
254 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
255 certextract with support for various fields. Bug 1358.
257 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
258 is requested by default, modifiable by smtp transport option
261 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
262 operate on certificate variables to give certificate fingerprints
263 Also new ${sha256:cert_variable}.
265 JH/23 The PRDR feature is moved from being Experimental into the mainline.
267 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
268 Christian Aistleitner.
270 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
272 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
273 file. Patch from Wolfgang Breyha.
275 JH/25 Expand the coverage of the delivery $host and $host_address to
276 client authenticators run in verify callout. Bug 1476.
278 JH/26 Port service names are now accepted for tls_on_connect_ports, to
279 align with daemon_smtp_ports. Bug 72.
281 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
282 support and error reporting did not work properly.
284 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
285 and is readable. Patch from Andrew Colin Kissa.
287 TL/14 Enhance documentation of ${run expansion and how it parses the
288 commandline after expansion, particularly in the case when an
289 unquoted variable expansion results in an empty value.
291 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
293 PP/02 Fix internal collision of T_APL on systems which support RFC3123
294 by renaming away from it. Addresses GH issue 15, reported by
297 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
299 TL/15 SECURITY: prevent double expansion in math comparison functions
300 (can expand unsanitized data). Not remotely exploitable.
307 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
309 PP/02 Make -n do something, by making it not do something.
310 When combined with -bP, the name of an option is not output.
312 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
315 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
316 $sender_host_name and config options to manage this, and basic check
319 PP/05 DSCP support for outbound connections and control modifier for inbound.
321 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
322 (Only plugin which currently uses this is kerberos4, which nobody should
323 be using, but we should make it available and other future plugins might
324 conceivably use it, even though it would break NAT; stuff *should* be
325 using channel bindings instead).
327 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
328 name; added for Sendmail compatibility; requires admin caller.
329 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
330 just ignore it); requires trusted caller.
331 Also parse but ignore: -Ac -Am -X<logfile>
334 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
336 TL/02 Add +smtp_confirmation as a default logging option.
338 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
339 Patch by Magnus Holmgren from 2007-02-20.
341 TL/04 Bugzilla 1281 - Spec typo.
342 Bugzilla 1283 - Spec typo.
343 Bugzilla 1290 - Spec grammar fixes.
345 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
347 TL/06 Add Experimental DMARC support using libopendmarc libraries.
349 TL/07 Fix an out of order global option causing a segfault. Reported to dev
350 mailing list by by Dmitry Isaikin.
352 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
354 JH/02 Support "G" suffix to numbers in ${if comparisons.
356 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
358 NM/01 Bugzilla 1197 - Spec typo
359 Bugzilla 1196 - Spec examples corrections
361 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
363 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
364 gnutls_enable_pkcs11, but renamed to more accurately indicate its
367 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
368 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
370 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
371 "acl {{name}{arg}...}", and optional args on acl condition
374 JH/05 Permit multiple router/transport headers_add/remove lines.
376 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
378 JH/07 Avoid using a waiting database for a single-message-only transport.
379 Performance patch from Paul Fisher. Bugzilla 1262.
381 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
384 JH/09 Add $headers_added variable, with content from use of ACL modifier
385 add_header (but not yet added to the message). Bugzilla 199.
387 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
388 Pulled from Bugzilla 817 by Wolfgang Breyha.
390 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
392 (nb: this is the same fix as in Exim 4.80.1)
394 JH/11 Add A= logging on delivery lines, and a client_set_id option on
397 JH/12 Add optional authenticated_sender logging to A= and a log_selector
400 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
402 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
403 advertise SMTP AUTH mechanism to us, instead of a generic
404 protocol violation error. Also, make Exim more robust to bad
405 data from the Dovecot auth socket.
407 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
409 When a queue runner is handling a message, Exim first routes the
410 recipient addresses, during which it prunes them based on the retry
411 hints database. After that it attempts to deliver the message to
412 any remaining recipients. It then updates the hints database using
415 So if a recipient address works intermittently, it can get repeatedly
416 deferred at routing time. The retry hints record remains fresh so the
417 address never reaches the final cutoff time.
419 This is a fairly common occurrence when a user is bumping up against
420 their storage quota. Exim had some logic in its local delivery code
421 to deal with this. However it did not apply to per-recipient defers
422 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
424 This change adds a proper retry rule check during routing so that the
425 final cutoff time is checked against the message's age. We only do
426 this check if there is an address retry record and there is not a
427 domain retry record; this implies that previous attempts to handle
428 the address had the retry_use_local_parts option turned on. We use
429 this as an approximation for the destination being like a local
430 delivery, as in LMTP.
432 I suspect this new check makes the old local delivery cutoff check
433 redundant, but I have not verified this so I left the code in place.
435 TF/02 Correct gecos expansion when From: is a prefix of the username.
437 Test 0254 submits a message to Exim with the header
441 When I ran the test suite under the user fanf2, Exim expanded
442 the header to contain my full name, whereas it should have added
443 a Resent-Sender: header. It erroneously treats any prefix of the
444 username as equal to the username.
446 This change corrects that bug.
448 GF/01 DCC debug and logging tidyup
449 Error conditions log to paniclog rather than rejectlog.
450 Debug lines prefixed by "DCC: " to remove any ambiguity.
452 TF/03 Avoid unnecessary rebuilds of lookup-related code.
454 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
455 Bug spotted by Jeremy Harris; was flawed since initial commit.
456 Would have resulted in OCSP responses post-SNI triggering an Exim
457 NULL dereference and crash.
459 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
461 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
462 Bug detection, analysis and fix by Samuel Thibault.
463 Bugzilla 1331, Debian bug #698092.
465 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
467 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
468 Server implementation by Todd Lyons, client by JH.
469 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
470 config variable "prdr_enable" controls whether the server
471 advertises the facility. If the client requests PRDR a new
472 acl_data_smtp_prdr ACL is called once for each recipient, after
473 the body content is received and before the acl_smtp_data ACL.
474 The client is controlled by bolth of: a hosts_try_prdr option
475 on the smtp transport, and the server advertisement.
476 Default client logging of deliveries and rejections involving
477 PRDR are flagged with the string "PRDR".
479 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
480 fclose(). Diagnosis by Todd Lyons.
482 PP/17 Update configure.default to handle IPv6 localhost better.
483 Patch by Alain Williams (plus minor tweaks).
486 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
487 This is now consistent with GnuTLS, and is now documented: the
488 previous undocumented portable approach to treating the option as
489 unset was to force an expansion failure. That still works, and
490 an empty string is now equivalent.
492 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
493 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
494 not performing validation itself.
496 PP/20 Added force_command boolean option to pipe transport.
497 Patch from Nick Koston, of cPanel Inc.
499 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
502 TF/04 Added udpsend ACL modifer and hexquote expansion operator
504 PP/21 Fix eximon continuous updating with timestamped log-files.
505 Broken in a format-string cleanup in 4.80, missed when I repaired the
506 other false fix of the same issue.
507 Report and fix from Heiko Schlichting.
510 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
511 Report from Prashanth Katuri.
513 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
514 It's SecureTransport, so affects any MacOS clients which use the
515 system-integrated TLS libraries, including email clients.
517 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
518 using a MIME ACL for non-SMTP local injection.
519 Report and assistance in diagnosis by Warren Baker.
521 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
523 JH/16 Fix comparisons for 64b. Bugzilla 1385.
525 TL/09 Add expansion variable $authenticated_fail_id to keep track of
526 last id that failed so it may be referenced in subsequent ACL's.
528 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
531 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
532 ldap library initialization, allowing self-signed CA's to be
533 used. Also properly sets require_cert option later in code by
534 using NULL (global ldap config) instead of ldap handle (per
535 session). Bug diagnosis and testing by alxgomz.
537 TL/12 Enhanced documentation in the ratelimit.pl script provided in
538 the src/util/ subdirectory.
540 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
541 renamed to Transport Post Delivery Action by Jeremy Harris, as
544 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
545 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
546 redis_servers = needs to be configured which will be used by the redis
547 lookup. Patch from Warren Baker, of The Packet Hub.
549 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
551 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
552 hostname or reverse DNS when processing a host list. Used suggestions
553 from multiple comments on this bug.
555 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
557 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
558 Missed a few lines, added it to make the runtest require no keyboard
561 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
562 contains upper case chars. Make router use caseful_local_part.
564 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
565 support when GnuTLS has been built with p11-kit.
571 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
573 This, or similar/improved, will also be change PP/11 of 4.82.
579 PP/01 Handle short writes when writing local log-files.
580 In practice, only affects FreeBSD (8 onwards).
581 Bugzilla 1053, with thanks to Dmitry Isaikin.
583 NM/01 Bugzilla 949 - Documentation tweak
585 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
588 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
590 PP/02 Implemented gsasl authenticator.
592 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
594 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
595 `pkg-config foo` for cflags/libs.
597 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
598 with rest of GSASL and with heimdal_gssapi.
600 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
601 `pkg-config foo` for cflags/libs for the TLS implementation.
603 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
604 properties get this fed in as external SSF. A number of robustness
605 and debugging improvements to the cyrus_sasl authenticator.
607 PP/08 cyrus_sasl server now expands the server_realm option.
609 PP/09 Bugzilla 1214 - Log authentication information in reject log.
610 Patch by Jeremy Harris.
612 PP/10 Added dbmjz lookup type.
614 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
616 PP/12 MAIL args handles TAB as well as SP, for better interop with
617 non-compliant senders.
618 Analysis and variant patch by Todd Lyons.
620 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
621 Bug report from Lars Müller <lars@samba.org> (via SUSE),
622 Patch from Dirk Mueller <dmueller@suse.com>
624 PP/13 tls_peerdn now print-escaped for spool files.
625 Observed some $tls_peerdn in wild which contained \n, which resulted
626 in spool file corruption.
628 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
629 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
630 or write after TLS renegotiation, which otherwise led to messages
633 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
634 as a tracking header (ie: a signed header comes before the signature).
635 Patch from Wolfgang Breyha.
637 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
638 comma-sep list; embedded commas doubled.
640 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
642 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
644 Report and patch from Dmitry Banschikov.
646 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
647 Removed SSL_clear() after SSL_new() which led to protocol negotiation
648 failures. We appear to now support TLS1.1+ with Exim.
650 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
651 lets Exim select keys and certificates based upon TLS SNI from client.
652 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
653 before an outbound SMTP session. New log_selector, +tls_sni.
655 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
656 NULL dereference. Report and patch from Alun Jones.
658 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
659 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
660 Not seeing resolver debug output on NetBSD, but suspect this is a
661 resolver implementation change.
663 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
664 Left warnings. Added "eximon gdb" invocation mode.
666 PP/21 Defaulting "accept_8bitmime" to true, not false.
668 PP/22 Added -bw for inetd wait mode support.
670 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
671 locate the relevant includes and libraries. Made this the default.
673 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
674 Bugzilla 1246, report and most of solution from Tomasz Kusy.
676 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
677 This may cause build issues on older platforms.
679 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
680 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
681 gnutls_require_mac & gnutls_require_protocols (no longer supported).
682 Added SNI support via GnuTLS too.
683 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
685 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
687 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
688 Applied second patch from Janne, implementing suggestion to default
689 multiple-strings-in-record handling to match SPF spec.
691 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
693 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
694 read-only, out of scope).
695 Patch from Wolfgang Breyha, report from Stuart Northfield.
697 PP/29 Fix three issues highlighted by clang analyser static analysis.
698 Only crash-plausible issue would require the Cambridge-specific
699 iplookup router and a misconfiguration.
700 Report from Marcin Mirosław.
702 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
704 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
705 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
706 As part of this, removing so much warning spew let me fix some minor
707 real issues in debug logging.
709 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
710 assignment on my part. Fixed.
712 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
713 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
714 Janne Snabb (who went above and beyond: thank you).
716 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
717 string otherwise requires a connection and a bunch more work and it's
718 relatively easy to get wrong. Should also expose TLS library linkage
721 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
722 64-bit ${eval} (JH/03).
724 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
725 GNU libc to support some of the 64-bit stuff, should not lead to
726 conflicts. Defined before os.h is pulled in, so if a given platform
727 needs to override this, it can.
729 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
730 protection layer was required, which is not implemented.
731 Bugzilla 1254, patch from Wolfgang Breyha.
733 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
734 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
735 tls_dhparam take prime identifiers. Also unbreak combination of
736 OpenSSL+DH_params+TLSSNI.
738 PP/39 Disable SSLv2 by default in OpenSSL support.
744 PP/01 Solaris build fix for Oracle's LDAP libraries.
745 Bugzilla 1109, patch from Stephen Usher.
747 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
749 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
752 TF/02 Fix a couple more cases where we did not log the error message
753 when unlink() failed. See also change 4.74-TF/03.
755 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
756 lock up or crash if it happened to be inside a call to libc when it
757 got a SIGUSR1 from exiwhat.
759 The SIGUSR1 handler appends the current process status to the process
760 log which is later printed by exiwhat. It used to use the general
761 purpose logging code to do this, but several functions it calls are
762 not safe for signals.
764 The new output code in the SIGUSR1 handler is specific to the process
765 log, and simple enough that it's easy to inspect for signal safety.
766 Removing some special cases also simplifies the general logging code.
767 Removing the spurious timestamps from the process log simplifies
770 TF/04 Improved ratelimit ACL condition.
772 The /noupdate option has been deprecated in favour of /readonly which
773 has clearer semantics. The /leaky, /strict, and /readonly update modes
774 are mutually exclusive. The update mode is no longer included in the
775 database key; it just determines when the database is updated. (This
776 means that when you upgrde Exim will forget old rate measurements.)
778 Exim now checks that the per_* options are used with an update mode that
779 makes sense for the current ACL. For example, when Exim is processing a
780 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
781 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
782 must specify per_mail/readonly. If you omit the update mode it defaults to
783 /leaky where that makes sense (as before) or /readonly where required.
785 The /noupdate option is now undocumented but still supported for
786 backwards compatibility. It is equivalent to /readonly except that in
787 ACLs where /readonly is required you may specify /leaky/noupdate or
788 /strict/noupdate which are treated the same as /readonly.
790 A useful new feature is the /count= option. This is a generalization
791 of the per_byte option, so that you can measure the throughput of other
792 aggregate values. For example, the per_byte option is now equivalent
793 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
795 The per_rcpt option has been generalized using the /count= mechanism
796 (though it's more complicated than the per_byte equivalence). When it is
797 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
798 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
799 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
800 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
801 Note that using per_rcpt with a non-readonly update mode in more than
802 one ACL will cause the recipients to be double-counted. (The per_mail
803 and per_byte options don't have this problem.)
805 The handling of very low rates has changed slightly. If the computed rate
806 is less than the event's count (usually one) then this event is the first
807 after a long gap. In this case the rate is set to the same as this event's
808 count, so that the first message of a spam run is counted properly.
810 The major new feature is a mechanism for counting the rate of unique
811 events. The new per_addr option counts the number of different
812 recipients that someone has sent messages to in the last time period. It
813 behaves like per_rcpt if all the recipient addresses are different, but
814 duplicate recipient addresses do not increase the measured rate. Like
815 the /count= option this is a general mechanism, so the per_addr option
816 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
817 example, measure the rate that a client uses different sender addresses
818 with the options per_mail/unique=$sender_address. There are further
819 details in the main documentation.
821 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
823 TF/06 Removed a few PCRE remnants.
825 TF/07 Automatically extract Exim's version number from tags in the git
826 repository when doing development or release builds.
828 PP/02 Raise smtp_cmd_buffer_size to 16kB.
829 Bugzilla 879. Patch from Paul Fisher.
831 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
832 Heavily based on revision 40f9a89a from Simon Arlott's tree.
835 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
837 PP/05 Variable $av_failed, true if the AV scanner deferred.
838 Bugzilla 1078. Patch from John Horne.
840 PP/06 Stop make process more reliably on build failure.
841 Bugzilla 1087. Patch from Heiko Schlittermann.
843 PP/07 Make maildir_use_size_file an _expandable_ boolean.
844 Bugzilla 1089. Patch from Heiko Schlittermann.
846 PP/08 Handle ${run} returning more data than OS pipe buffer size.
847 Bugzilla 1131. Patch from Holger Weiß.
849 PP/09 Handle IPv6 addresses with SPF.
850 Bugzilla 860. Patch from Wolfgang Breyha.
852 PP/10 GnuTLS: support TLS 1.2 & 1.1.
854 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
857 PP/11 match_* no longer expand right-hand-side by default.
858 New compile-time build option, EXPAND_LISTMATCH_RHS.
859 New expansion conditions, "inlist", "inlisti".
861 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
863 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
865 PP/14 fix log_write() format string regression from TF/03.
866 Bugzilla 1152. Patch from Dmitry Isaikin.
872 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
874 PP/02 Harmonised TLS library version reporting; only show if debugging.
875 Layout now matches that introduced for other libraries in 4.74 PP/03.
877 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
879 PP/04 New "dns_use_edns0" global option.
881 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
884 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
885 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
887 TK/01 Updated PolarSSL code to 0.14.2.
888 Bugzilla 1097. Patch from Andreas Metzler.
890 PP/07 Catch divide-by-zero in ${eval:...}.
893 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
896 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
897 format-string attack -- SECURITY: remote arbitrary code execution.
899 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
900 time unintentionally subject to list matching rules, letting the header
901 cause arbitrary Exim lookups (of items which can occur in lists, *not*
902 arbitrary string expansion). This allowed for information disclosure.
904 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
905 INT_MIN/-1 -- value coerced to INT_MAX.
911 NM/01 Workround for PCRE version dependency in version reporting
914 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
915 This fixes portability to compilers other than gcc, notably
916 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
918 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
919 makefiles for portability to HP-UX and POSIX correctness.
921 PP/01 Permit LOOKUP_foo enabling on the make command-line.
922 Also via indented variable definition in the Makefile.
923 (Debugging by Oliver Heesakkers).
925 PP/02 Restore caching of spamd results with expanded spamd_address.
926 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
928 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
929 Improves build reliability. Fix from: Frank Elsner
931 NM/02 Fix wide character breakage in the rfc2047 coding
932 Fixes bug 1064. Patch from Andrey N. Oktyabrski
934 NM/03 Allow underscore in dnslist lookups
935 Fixes bug 1026. Patch from Graeme Fowler
937 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
938 Code patches from Adam Ciarcinski of NetBSD.
940 NM/04 Fixed exiqgrep to cope with mailq missing size issue
943 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
944 is logged, to avoid truncation. Patch from John Horne.
946 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
947 Patch from Jakob Hirsch.
949 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
950 SQL string expansion failure details.
951 Patch from Andrey Oktyabrski.
953 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
954 Patch from Simon Arlott.
956 PP/09 New lookups functionality failed to compile on old gcc which rejects
957 extern declarations in function scope.
958 Patch from Oliver Fleischmann
960 PP/10 Use sig_atomic_t for flags set from signal handlers.
961 Check getgroups() return and improve debugging.
962 Fixed developed for diagnosis in bug 927 (which turned out to be
965 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
966 Patch from Mark Zealey.
968 PP/12 Bugzilla 1056: Improved spamd server selection.
969 Patch from Mark Zealey.
971 PP/13 Bugzilla 1086: Deal with maildir quota file races.
972 Based on patch from Heiko Schlittermann.
974 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
975 Patch from Uwe Doering, sign-off by Michael Haardt.
977 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
978 variable declaration deep within a block. Bug and patch from
981 PP/15 lookups-Makefile IRIX compatibilty coercion.
983 PP/16 Make DISABLE_DKIM build knob functional.
985 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
986 Patch by Simon Arlott
988 TF/03 Fix valgrind.h portability to C89 compilers that do not support
989 variable argument macros. Our copy now differs from upstream.
995 TF/01 Failure to get a lock on a hints database can have serious
996 consequences so log it to the panic log.
998 TF/02 Log LMTP confirmation messages in the same way as SMTP,
999 controlled using the smtp_confirmation log selector.
1001 TF/03 Include the error message when we fail to unlink a spool file.
1003 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1004 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1005 for maintaining out-of-tree patches for some time.
1007 PP/01 Bugzilla 139: Documentation and portability issues.
1008 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1009 Handle per-OS dynamic-module compilation flags.
1011 PP/02 Let /dev/null have normal permissions.
1012 The 4.73 fixes were a little too stringent and complained about the
1013 permissions on /dev/null. Exempt it from some checks.
1014 Reported by Andreas M. Kirchwitz.
1016 PP/03 Report version information for many libraries, including
1017 Exim version information for dynamically loaded libraries. Created
1018 version.h, now support a version extension string for distributors
1019 who patch heavily. Dynamic module ABI change.
1021 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1022 privilege escalation vulnerability whereby the Exim run-time user
1023 can cause root to append content of the attacker's choosing to
1026 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1029 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1030 If dropping privileges for untrusted macros, we disabled normal logging
1031 on the basis that it would fail; for the Exim run-time user, this is not
1032 the case, and it resulted in successful deliveries going unlogged.
1033 Fixed. Reported by Andreas Metzler.
1039 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1040 only prepend for the Resent-* case. Fixes regression introduced in
1041 Exim 4.70 by NM/22 for Bugzilla 607.
1043 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1044 increasing numbers of administrators be bitten by this.
1046 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1048 PP/03 Bugzilla 994: added openssl_options main configuration option.
1050 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1052 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1054 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1056 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1057 without a peer certificate, leading to a segfault because of an
1058 assumption that peers always have certificates. Be a little more
1059 paranoid. Problem reported by Martin Tscholak.
1061 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1062 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1063 NB: ClamAV planning to remove STREAM in "middle of 2010".
1064 CL also introduces -bmalware, various -d+acl logging additions and
1065 more caution in buffer sizes.
1067 PP/09 Implemented reverse_ip expansion operator.
1069 PP/10 Bugzilla 937: provide a "debug" ACL control.
1071 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1073 PP/12 Bugzilla 973: Implement --version.
1075 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1077 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1079 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1081 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1082 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1083 ignore trailing whitespace.
1085 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1087 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1090 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1091 Notification from Dr Andrew Aitchison.
1093 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1094 ExtendedDetectionInfo response format.
1095 Notification from John Horne.
1097 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1100 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1101 XSL and documented dependency on system catalogs, with examples of how
1104 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1107 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1108 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1109 configuration file which is writeable by the Exim user or group.
1111 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1112 of configuration files to cover files specified with the -C option if
1113 they are going to be used with root privileges, not just the default
1116 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1117 option (effectively making it always true).
1119 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1120 files to be used while preserving root privileges.
1122 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1123 that rogue child processes cannot use them.
1125 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1126 run-time user, instead of root.
1128 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1129 Exim run-time user without dropping privileges.
1131 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1132 result string, instead of calling string_vformat() twice with the same
1135 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1136 for other users. Others should always drop root privileges if they use
1137 -C on the command line, even for a whitelisted configure file.
1139 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1141 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1147 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1148 $data_path, and $header_path variables; fixed documentation bugs and
1151 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1152 exipick to access non-standard spools, including the "frozen" queue
1155 NM/01 Bugzilla 965: Support mysql stored procedures.
1156 Patch from Alain Williams
1158 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1160 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1161 Patch from Andreas Metzler
1163 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1164 Patch from Kirill Miazine
1166 NM/05 Bugzilla 671: Added umask to procmail example.
1168 JJ/03 installed exipick 20100323.0, fixing doc bug
1170 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1171 directory. Notification and patch from Dan Rosenberg.
1173 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1175 TK/02 Improve log output when DKIM signing operation fails.
1177 MH/01 Treat the transport option dkim_domain as a colon separated
1178 list, not as a single string, and sign the message with each element,
1179 omitting multiple occurences of the same signer.
1181 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1182 Bugzilla 985, 986. Patch by Simon Arlott
1184 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1185 Patch by Simon Arlott
1187 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1188 MBX locking. Notification from Dan Rosenberg.
1194 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1196 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1198 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1200 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1202 NM/04 Bugzilla 915: Flag broken perl installation during build.
1208 TK/01 Added patch by Johannes Berg that expands the main option
1209 "spamd_address" if it starts with a dollar sign.
1211 TK/02 Write list of recipients to X-Envelope-Sender header when building
1212 the mbox-format spool file for content scanning (suggested by Jakob
1215 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1216 (http://www.dcc-servers.net/) support via dccifd. Activated by
1217 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1219 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1220 by Mark Daniel Reidel <mr@df.eu>.
1222 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1223 When building exim an external PCRE library is now needed -
1224 PCRE is a system library on the majority of modern systems.
1225 See entry on PCRE_LIBS in EDITME file.
1227 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1228 conversation. Added nologin parameter to request.
1229 Patch contributed by Kirill Miazine.
1231 TF/01 Do not log submission mode rewrites if they do not change the address.
1233 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1235 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1236 log files in place. Contributed by Roberto Lima.
1238 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1240 TF/03 Bugzilla 615: When checking the local_parts router precondition
1241 after a local_part_suffix or local_part_prefix option, Exim now
1242 does not use the address's named list lookup cache, since this
1243 contains cached lookups for the whole local part.
1245 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1246 Robert Millan. Documentation is in experimental-spec.txt.
1248 TF/04 Bugzilla 668: Fix parallel build (make -j).
1250 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1252 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1253 Patch provided by Jan Srzednicki.
1255 TF/05 Leading white space used to be stripped from $spam_report which
1256 wrecked the formatting. Now it is preserved.
1258 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1259 that they are available at delivery time.
1261 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1263 TF/08 TLS error reporting now respects the incoming_interface and
1264 incoming_port log selectors.
1266 TF/09 Produce a more useful error message if an SMTP transport's hosts
1267 setting expands to an empty string.
1269 NM/06 Bugzilla 744: EXPN did not work under TLS.
1270 Patch provided by Phil Pennock.
1272 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1273 Patch provided by Richard Godbee.
1275 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1276 acl_smtp_notquit, added index entry.
1278 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1279 Patch provided by Eugene Bujak.
1281 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1282 accept(). Patch provided by Maxim Dounin.
1284 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1285 Patch provided by Phil Pennock.
1287 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1289 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1290 Patch provided by Brad "anomie" Jorsch.
1292 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1293 Patch provided by Dean Brooks.
1295 TK/05 Add native DKIM support (does not depend on external libraries).
1297 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1298 Patch provided by Graeme Fowler.
1300 NM/16 Bugzilla 851: Documentation example syntax fix.
1302 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1304 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1307 NM/19 Bugzilla 745: TLS version reporting.
1308 Patch provided by Phil Pennock.
1310 NM/20 Bugzilla 167: bool: condition support.
1311 Patch provided by Phil Pennock.
1313 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1314 clients. Patch provided by Phil Pennock.
1316 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1317 Patch provided by Brad "anomie" Jorsch.
1319 NM/23 Bugzilla 687: Fix misparses in eximstats.
1320 Patch provided by Heiko Schlittermann.
1322 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1323 Patch provided by Heiko Schlittermann.
1325 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1326 plus update to original patch.
1328 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1330 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1331 Patch provided by David Brownlee.
1333 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1335 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1337 NM/30 Bugzilla 888: TLS documentation bugfixes.
1339 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1341 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1342 Unlike the original bugzilla I have changed all shell scripts in src tree.
1344 NM/33 Bugzilla 898: Transport filter timeout fix.
1345 Patch by Todd Rinaldo.
1347 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1348 Patch by Serge Demonchaux.
1350 NM/35 Bugzilla 39: Base64 decode bug fixes.
1351 Patch by Jakob Hirsch.
1353 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1355 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1357 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1359 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1365 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1366 ALT-N's libdkim that I have put here:
1367 http://duncanthrax.net/exim-experimental/
1369 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1370 were called 'true' and it seems that C99 defines that as a reserved
1371 keyword to be used with 'bool' variable types. That means you could
1372 not include C99-style headers which use bools without triggering
1373 build errors in sieve.c.
1375 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1376 as mailq or other aliases. Changed the --help handling significantly
1377 to do whats expected. exim_usage() emits usage/help information.
1379 SC/01 Added the -bylocaldomain option to eximstats.
1381 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1383 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1385 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1391 PH/01 Another patch from the Sieve maintainer.
1393 PH/02 When an IPv6 address is converted to a string for single-key lookup
1394 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1395 dots are used instead of colons so that keys in lsearch files need not
1396 contain colons. This was done some time before quoting was made available
1397 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1398 (notated using the quote facility) so as to distinguish them from IPv4
1399 keys. This meant that lookups for IP addresses in host lists did not work
1400 for iplsearch lookups.
1402 This has been fixed by arranging for IPv6 addresses to be expressed with
1403 colons if the lookup type is iplsearch. This is not incompatible, because
1404 previously such lookups could never work.
1406 The situation is now rather anomolous, since one *can* have colons in
1407 ordinary lsearch keys. However, making the change in all cases is
1408 incompatible and would probably break a number of configurations.
1410 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1413 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1414 conversion specification without a maximum field width, thereby enabling
1415 a rogue spamd server to cause a buffer overflow. While nobody in their
1416 right mind would setup Exim to query an untrusted spamd server, an
1417 attacker that gains access to a server running spamd could potentially
1418 exploit this vulnerability to run arbitrary code as the Exim user.
1420 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1421 $primary_hostname instead of what libspf2 thinks the hosts name is.
1423 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1424 a directory entry by the name of the lookup key. Previously, if a
1425 symlink pointed to a non-existing file or a file in a directory that
1426 Exim lacked permissions to read, a lookup for a key matching that
1427 symlink would fail. Now it is enough that a matching directory entry
1428 exists, symlink or not. (Bugzilla 503.)
1430 PH/03 The body_linecount and body_zerocount variables are now exported in the
1433 PH/04 Added the $dnslist_matched variable.
1435 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1436 This means they are set thereafter only if the connection becomes
1439 PH/06 Added the client_condition to authenticators so that some can be skipped
1440 by clients under certain conditions.
1442 PH/07 The error message for a badly-placed control=no_multiline_responses left
1443 "_responses" off the end of the name.
1445 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1447 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1448 (without spaces) instead of just copying the configuration text.
1450 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1452 PH/11 Added $max_received_linelength.
1454 PH/12 Added +ignore_defer and +include_defer to host lists.
1456 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1457 way in which PCRE > 7.0 is built.
1459 PH/14 Implemented queue_only_load_latch.
1461 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1462 MAIL command. The effect was to mangle the value on 64-bit systems.
1464 PH/16 Another patch from the Sieve maintainer.
1466 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1468 PH/18 If a system quota error occurred while trying to create the file for
1469 a maildir delivery, the message "Mailbox is full" was not appended to the
1470 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1471 only to a quota excession during the actual writing of the file.
1473 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1474 characters?) which causes problems in log lines. The DN values are now
1475 passed through string_printing() before being added to log lines.
1477 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1478 and InterBase are left for another time.)
1480 PH/21 Added message_body_newlines option.
1482 PH/22 Guard against possible overflow in moan_check_errorcopy().
1484 PH/23 POSIX allows open() to be a macro; guard against that.
1486 PH/24 If the recipient of an error message contained an @ in the local part
1487 (suitably quoted, of course), incorrect values were put in $domain and
1488 $local_part during the evaluation of errors_copy.
1494 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1495 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1498 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1499 issue a MAIL command.
1501 PH/02 In an ACL statement such as
1503 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1505 if a client was not listed at all, or was listed with a value other than
1506 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1507 the condition was not true (as it should be), so access was not denied.
1508 The bug was that the ! inversion was incorrectly passed on to the second
1509 item. This has been fixed.
1511 PH/03 Added additional dnslists conditions == and =& which are different from
1512 = and & when the dns lookup returns more than one IP address.
1514 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1515 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1517 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1518 FSYNC, which compiles an option called disable_fsync that allows for
1519 bypassing fsync(). The documentation is heavily laced with warnings.
1521 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1523 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1524 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1525 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1526 including adding "make clean"; (3) Added -fPIC when compiling the test
1527 dynamically loaded module, to get rid of a warning.
1529 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1530 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1531 = 0s. The bug is otherwise harmless.
1533 PH/07 There was a bug in the dovecot authenticator such that the value of
1534 $auth1 could be overwritten, and so not correctly preserved, after a
1535 successful authentication. This usually meant that the value preserved by
1536 the server_setid option was incorrect.
1538 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1540 PH/09 Installed PCRE release 7.0.
1542 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1543 run for batched SMTP input. It is now run at the start of every message
1544 in the batch. While fixing this I discovered that the process information
1545 (output by running exiwhat) was not always getting set for -bs and -bS
1546 input. This is fixed, and it now also says "batched" for BSMTP.
1548 PH/11 Added control=no_pipelining.
1550 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1551 patch, slightly modified), and move the expansion of helo_data till after
1552 the connection is made in the smtp transport (so it can use these
1555 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1557 PH/14 Added log_selector = +pid.
1559 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1561 PH/16 Add ${if forany and ${if forall.
1563 PH/17 Added dsn_from option to vary the From: line in DSNs.
1565 PH/18 Flush SMTP output before performing a callout, unless control =
1566 no_callout_flush is set.
1568 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1569 was true (the default) a successful delivery failed to delete the retry
1570 item, thus causing premature timeout of the address. The bug is now
1573 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1575 PH/21 Long custom messages for fakedefer and fakereject are now split up
1576 into multiline reponses in the same way that messages for "deny" and
1577 other ACL rejections are.
1579 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1580 with slight modification.
1582 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1583 draft, changing the syntax and factoring some duplicate code".
1585 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1586 for deliveries of the second and subsequent messages over the same SMTP
1589 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1590 ${reduce, with only minor "tidies".
1592 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1594 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1595 expansion side effects.
1597 PH/27 When a message times out after an over-quota error from an Exim-imposed
1598 quota, the bounce message says "mailbox is full". This message was not
1599 being given when it was a system quota that was exceeded. It now should
1602 MH/03 Made $recipients available in local_scan(). local_scan() already has
1603 better access to the recipient list through recipients_list[], but
1604 $recipients can be useful in postmaster-provided expansion strings.
1606 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1607 in the case of a MAIL command with additional options following the
1608 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1609 were accidentally chopped off.
1611 PH/29 SMTP synchronization checks are implemented when a command is read -
1612 there is a check that no more input is waiting when there shouldn't be
1613 any. However, for some commands, a delay in an ACL can mean that it is
1614 some time before the response is written. In this time, more input might
1615 arrive, invalidly. So now there are extra checks after an ACL has run for
1616 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1617 pipelining has not been advertised.
1619 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1621 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1622 returned in the SMTP response when smtp_return_error_details was false.
1623 This has been fixed.
1625 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1626 instead of the C I/O that was originally supplied, because problems were
1627 reported on Solaris.
1629 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1630 Exim which did not show up earlier: it was assuming that a call to
1631 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1632 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1633 was a macro that became an assignment, so it seemed to work. This has
1634 changed to a proper function call with a void return, hence the compile
1635 error. Exim's code has been fixed.
1637 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1640 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1642 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1644 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1645 because it turns out to be too big or there is a timeout) there is no
1646 "Completed" line in the log. When some messages of this type were
1647 selected by exigrep, they were listed as "not completed". Others were
1648 picked up by some special patterns. I have improved the selection
1649 criteria to be more general.
1651 PH/38 The host_find_failed option in the manualroute router can now be set
1652 to "ignore", to completely ignore a host whose IP address cannot be
1653 found. If all hosts are ignored, the behaviour is controlled by the new
1654 host_all_ignored option.
1656 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1657 homing or because of multiple MX records with /mx) generated more than
1658 one IP address, and the following item turned out to be the local host,
1659 all the secondary addresses of the first item were incorrectly removed
1660 from the list, along with the local host and any following hosts (which
1661 is what is supposed to happen).
1663 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1664 whoever called Exim into the -H file. In the case of the daemon it was
1665 behaving confusingly. When first started, it used values for whoever
1666 started the daemon, but after a SIGHUP it used the Exim user (because it
1667 calls itself on a restart). I have changed the code so that it now always
1670 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1671 message are rejected with the same error (e.g. no authentication or bad
1672 sender address), and a DATA command is nevertheless sent (as can happen
1673 with PIPELINING or a stupid MUA), the error message that was given to the
1674 RCPT commands is included in the rejection of the DATA command. This is
1675 intended to be helpful for MUAs that show only the final error to their
1678 PH/42 Another patch from the Sieve maintainer.
1680 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1681 Eximstats - Fixed some broken HTML links and added missing column headers
1683 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1684 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1686 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1688 PH/43 Yet another patch from the Sieve maintainer.
1690 PH/44 I found a way to check for a TCP/IP connection going away before sending
1691 the response to the final '.' that terminates a message, but only in the
1692 case where the client has not sent further data following the '.'
1693 (unfortunately, this is allowed). However, in many cases there won't be
1694 any further data because there won't be any more messages to send. A call
1695 to select() can be used: if it shows that the input is "ready", there is
1696 either input waiting, or the socket has been closed. An attempt to read
1697 the next input character can distinguish the two cases. Previously, Exim
1698 would have sent an OK response which the client would never have see.
1699 This could lead to message repetition. This fix should cure that, at
1700 least in a lot of common cases.
1702 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1703 advertised in response to EHLO.
1709 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1710 fixed by 4.65/MH/01 (is this a record?) are fixed:
1712 (i) An empty string was always treated as zero by the numeric comparison
1713 operators. This behaviour has been restored.
1715 (ii) It is documented that the numeric comparison operators always treat
1716 their arguments as decimal numbers. This was broken in that numbers
1717 starting with 0 were being interpreted as octal.
1719 While fixing these problems I realized that there was another issue that
1720 hadn't been noticed. Values of message_size_limit (both the global option
1721 and the transport option) were treated as octal if they started with 0.
1722 The documentation was vague. These values are now always treated as
1723 decimal, and I will make that clear in the documentation.
1729 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1730 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1733 MH/01 Don't check that the operands of numeric comparison operators are
1734 integers when their expansion is in "skipping" mode (fixes bug
1735 introduced by 4.64-PH/07).
1737 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1738 child addresses, Exim now panics and dies. Previously, because the count
1739 is held in a short int, deliveries were likely to be lost. As such a
1740 large number of recipients for a single message is ridiculous
1741 (performance will be very, very poor), I have chosen to impose a limit
1742 rather than extend the field.
1748 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1749 leftover -K file (the existence of which was triggered by #402).
1750 While we were at it, introduced process PID as part of the -K
1751 filename. This should rule out race conditions when creating
1754 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1755 processing considerably. Previous code took too long for large mails,
1756 triggering a timeout which in turn triggers #401.
1758 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1759 in the DK code in transports.c. sendfile() is not really portable,
1760 hence the _LINUX specificness.
1762 TF/01 In the add_headers option to the mail command in an Exim filter,
1763 there was a bug that Exim would claim a syntax error in any
1764 header after the first one which had an odd number of characters
1767 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1768 callout verification, Exim cached a "reject" for the entire domain. This
1769 is correct for most verifications, but it is not correct for a recipient
1770 verification with use_sender or use_postmaster set, because in that case
1771 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1772 case of MAIL FROM:<> rejection from other early rejections (e.g.
1773 rejection of HELO). When verifying a recipient using a non-null MAIL
1774 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1775 Whatever the result of the callout, the value of the domain cache is
1776 left unchanged (for any other kind of callout, getting as far as trying
1777 RCPT means that the domain itself is ok).
1779 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1782 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1783 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1784 session, the daemon ignores ECONNECT errors and logs others; it now
1785 ignores EPIPE as well.
1787 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1788 (quoted-printable decoding).
1790 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1791 later the small subsequent patch to fix an introduced bug.
1793 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1795 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1797 PH/08 An error is now given if message_size_limit is specified negative.
1799 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1800 to be given (somewhat) arbitrary names.
1802 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1805 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1806 miscellaneous code fixes
1808 PH/10 Added the log_reject_target ACL modifier to specify where to log
1811 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1812 hostname. This is wrong, because it relates to the incoming message (and
1813 probably the interface on which it is arriving) and not to the outgoing
1814 callout (which could be using a different interface). This has been
1815 changed to use the value of the helo_data option from the smtp transport
1816 instead - this is what is used when a message is actually being sent. If
1817 there is no remote transport (possible with a router that sets up host
1818 addresses), $smtp_active_hostname is used.
1820 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1821 tweaks were necessary in order to get it to work (see also 21 below):
1822 (a) The code assumed that strncpy() returns a negative number on buffer
1823 overflow, which isn't the case. Replaced with Exim's string_format()
1825 (b) There were several signed/unsigned issues. I just did the minimum
1826 hacking in of casts. There is scope for a larger refactoring.
1827 (c) The code used strcasecmp() which is not a standard C function.
1828 Replaced with Exim's strcmpic() function.
1829 (d) The code set only $1; it now sets $auth1 as well.
1830 (e) A simple test gave the error "authentication client didn't specify
1831 service in request". It would seem that Dovecot has changed its
1832 interface. Fortunately there's a specification; I followed it and
1833 changed what the client sends and it appears to be working now.
1835 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1838 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1839 address A is aliased to B and C, where B exists and C does not. Without
1840 -v the output is "A verified" because verification stops after a
1841 successful redirection if more than one address is generated. However,
1842 with -v the child addresses are also verified. Exim was outputting "A
1843 failed to verify" and then showing the successful verification for C,
1844 with its parentage. It now outputs "B failed to verify", showing B's
1845 parentage before showing the successful verification of C.
1847 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1848 look up a TXT record in a specific list after matching in a combined
1851 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1852 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1853 they consult the DNS. I had assumed they would set it the way they
1854 wanted; and indeed my experiments on Linux seem to show that in some
1855 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1856 To be on the safe side, however, I have now made the interface to
1857 host_find_byname() similar to host_find_bydns(), with an argument
1858 containing the DNS resolver options. The host_find_byname() function now
1859 sets these options at its start, just as host_find_bydns() does. The smtp
1860 transport options dns_qualify_single and dns_search_parents are passed to
1861 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1862 of host_find_byname() use the default settings of RES_DEFNAMES
1863 (qualify_single) but not RES_DNSRCH (search_parents).
1865 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1866 spool_read_header() do less string testing, by means of a preliminary
1867 switch on the second character of optional "-foo" lines. (This is
1868 overdue, caused by the large number of possibilities that now exist.
1869 Originally there were few.) While I was there, I also converted the
1870 str(n)cmp tests so they don't re-test the leading "-" and the first
1871 character, in the hope this might squeeze out yet more improvement.
1873 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1874 flag allowing group syntax was set by the header_syntax check but not
1875 turned off, possible causing trouble later; (2) The flag was not being
1876 set at all for the header_verify test, causing "group"-style headers to
1877 be rejected. I have now set it in this case, and also caused header_
1878 verify to ignore an empty address taken from a group. While doing this, I
1879 came across some other cases where the code for allowing group syntax
1880 while scanning a header line wasn't quite right (mostly, not resetting
1881 the flag correctly in the right place). These bugs could have caused
1882 trouble for malformed header lines. I hope it is now all correct.
1884 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1885 with the "reply" argument non-NULL. The code, however (which originally
1886 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1887 but it didn't always do it. This confused somebody who was copying the
1888 code for some other use. I have removed all the tests.
1890 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1891 feature that was used to support insecure browsers during the U.S. crypto
1892 embargo. It requires special client support, and Exim is probably the
1893 only MTA that supported it -- and would never use it because real RSA is
1894 always available. This code has been removed, because it had the bad
1895 effect of slowing Exim down by computing (never used) parameters for the
1896 RSA_EXPORT functionality.
1898 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1899 authenticator to fail if there's a tab character in the incoming data
1900 (there should never be unless someone is messing about, as it's supposed
1901 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1902 if the connection is using TLS or if the remote IP is the same as the
1903 local IP, and the "valid-client-cert option" if a client certificate has
1906 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1907 authenticators. This can be used for authorization after authentication
1908 succeeds. (In the case of plaintext, it servers for both authentication
1911 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1912 if any retry times were supplied.
1914 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1915 connection, where there is no client IP address to check. In this
1916 situation, the verify now always succeeds.
1918 PH/25 Applied John Jetmore's -Mset patch.
1920 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1922 PH/27 In a string expansion for a processed (not raw) header when multiple
1923 headers of the same name were present, leading whitespace was being
1924 removed from all of them, but trailing whitespace was being removed only
1925 from the last one. Now trailing whitespace is removed from each header
1926 before concatenation. Completely empty headers in a concatenation (as
1927 before) are ignored.
1929 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1930 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1932 PH/29 [Removed. This was a change that I later backed out, and forgot to
1933 correct the ChangeLog entry (that I had efficiently created) before
1934 committing the later change.]
1936 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1937 address errors (4xx response to RCPT) over the same connection as other
1938 messages routed to the same hosts. Such deliveries are always "forced",
1939 so retry times are not inspected. This resulted in far too many retries
1940 for the affected addresses. The effect occurred only when there were more
1941 hosts than the hosts_max_try setting in the smtp transport when it had
1942 the 4xx errors. Those hosts that it had tried were not added to the list
1943 of hosts for which the message was waiting, so if all were tried, there
1944 was no problem. Two fixes have been applied:
1946 (i) If there are any address or message errors in an SMTP delivery, none
1947 of the hosts (tried or untried) are now added to the list of hosts
1948 for which the message is waiting, so the message should not be a
1949 candidate for sending over the same connection that was used for a
1950 successful delivery of some other message. This seems entirely
1951 reasonable: after all the message is NOT "waiting for some host".
1952 This is so "obvious" that I'm not sure why it wasn't done
1953 previously. Hope I haven't missed anything, but it can't do any
1954 harm, as the worst effect is to miss an optimization.
1956 (ii) If, despite (i), such a delivery is accidentally attempted, the
1957 routing retry time is respected, so at least it doesn't keep
1958 hammering the server.
1960 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1961 in ${readsocket because some servers need this prod.
1963 PH/32 Added some extra debug output when updating a wait-xxx database.
1965 PH/33 The hint "could be header name not terminated by colon", which has been
1966 given for certain expansion errors for a long time, was not being given
1967 for the ${if def:h_colon_omitted{... case.
1969 PH/34 The spec says: "With one important exception, whenever a domain list is
1970 being scanned, $domain contains the subject domain." There was at least
1971 one case where this was not true.
1973 PH/35 The error "getsockname() failed: connection reset by peer" was being
1974 written to the panic log as well as the main log, but it isn't really
1975 panic-worthy as it just means the connection died rather early on. I have
1976 removed the panic log writing for the ECONNRESET error when getsockname()
1979 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1980 runs only) independently of the message's sender address. This meant
1981 that, if the 4xx error was in fact related to the sender, a different
1982 message to the same recipient with a different sender could confuse
1983 things. In particualar, this can happen when sending to a greylisting
1984 server, but other circumstances could also provoke similar problems.
1985 I have changed the default so that the retry time for these errors is now
1986 based a combination of the sender and recipient addresses. This change
1987 can be overridden by setting address_retry_include_sender=false in the
1990 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1991 remote server are returned as part of bounce messages. This was not
1992 happening for LMTP over a pipe (the lmtp transport), but now it is the
1993 same for both kinds of LMTP.
1995 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1996 in header lines that were in fact CNAMEs. This is no longer the case.
1998 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1999 and queue runs started by the daemon processed all messages. This has
2000 been fixed so that -R and -S can now usefully be given with -q<time>.
2002 PH/40 Import PCRE release 6.7 (fixes some bugs).
2004 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2006 PH/42 Give an error if -q is specified more than once.
2008 PH/43 Renamed the variables $interface_address and $interface_port as
2009 $received_ip_address and $received_port, to make it clear that these
2010 values apply to message reception, and not to the outgoing interface when
2011 a message is delivered. (The old names remain recognized, of course.)
2013 PH/44 There was no timeout on the connect() call when using a Unix domain
2014 socket in the ${readsocket expansion. There now is.
2016 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2017 be meaningful with "accept".
2019 SC/01 Eximstats V1.43
2020 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2022 SC/02 Eximstats V1.44
2023 Use a glob alias rather than an array ref in the generated
2024 parser. This improves both readability and performance.
2026 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2027 Collect SpamAssassin and rejection statistics.
2028 Don't display local sender or destination tables unless
2029 there is data to show.
2030 Added average volumes into the top table text output.
2032 SC/04 Eximstats V1.46
2033 Collect data on the number of addresses (recipients)
2034 as well as the number of messages.
2036 SC/05 Eximstats V1.47
2037 Added 'Message too big' to the list of mail rejection
2038 reasons (thanks to Marco Gaiarin).
2040 SC/06 Eximstats V1.48
2041 Mainlog lines which have GMT offsets and are too short to
2042 have a flag are now skipped.
2044 SC/07 Eximstats V1.49 (Alain Williams)
2045 Added the -emptyok flag.
2047 SC/08 Eximstats V1.50
2048 Fixes for obtaining the IP address from reject messages.
2050 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2051 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2052 whitesspace changes from 4.64-PH/27
2054 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2057 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2058 are found, allow negative numbers in numeric criteria)
2060 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2062 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2063 to match changes made in 4.64-PH/43
2065 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2067 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2068 "domain" is actually an IP address, and force a failure. This locks out
2069 those revolvers/nameservers that support "A-for-A" lookups, in
2070 contravention of the specifications.
2072 PH/48 When a host name was looked up from an IP address, and the subsequent
2073 forward lookup of the name timed out, the host name was left in
2074 $sender_host_name, contrary to the specification.
2076 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2077 restricted to single-key lookups, Exim was not diagnosing an error if
2078 * or *@ was used with a query-style lookup.
2080 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2082 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2083 long ago, but noone interested enough thought of it. Let's just say that
2084 the "1.1" means that there are some new functions that weren't there at
2085 some point in the past.
2087 PH/51 Error processing for expansion failure of helo_data from an smtp
2088 transport during callout processing was broken.
2090 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2091 tested/used via the -bh/-bhc/-bs options.
2093 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2094 bug, fixed in subsequent PCRE releases).
2096 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2097 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2099 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2105 SC/01 Use a glob alias rather than an array ref in eximstats generated
2106 parser. This improves both readability and performance.
2108 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2109 Don't display local sender or destination tables in eximstats unless
2110 there is data to show.
2111 Added average volumes into the eximstats top table text output.
2113 SC/03 Collect data on the number of addresses (recipients) as well
2114 as the number of messages in eximstats.
2116 TF/01 Correct an error in the documentation for the redirect router. Exim
2117 does (usually) call initgroups() when daemonizing.
2119 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2120 with consistent privilege compared to when running as a daemon.
2122 TF/03 Note in the spec that $authenticated_id is not set for local
2123 submissions from trusted users.
2125 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2126 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2128 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2129 by adding some example configuration directives to the default
2130 configuration file. A little bit of work is required to uncomment the
2131 directives and define how usernames and passwords are checked, but
2132 there is now a framework to start from.
2134 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2135 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2136 without this. I don't know how relevant this is to other LDAP libraries.
2138 PH/02 Add the verb name to the "unknown ACL verb" error.
2140 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2142 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2144 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2145 directory not expanded when it should be if an expanded home directory
2146 was set for the address (which is overridden by the transport).
2148 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2151 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2152 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2153 because it is too late at that time, and has no effect.
2155 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2156 security issue with \' (bugzilla #107). I could not use the
2157 PQescapeStringConn() function, because it needs a PGconn value as one of
2160 PH/08 When testing addresses using -bt, indicate those final addresses that
2161 are duplicates that would not cause an additional delivery. At least one
2162 person was confused, thinking that -bt output corresponded to deliveries.
2163 (Suppressing duplicates isn't a good idea as you lose the information
2164 about possibly different redirections that led to the duplicates.)
2166 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2167 systems where poll() doesn't work, in particular OS X.
2169 PH/10 Added more information to debugging output for retry time not reached.
2171 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2172 operations in malware.c.
2174 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2177 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2178 syslog_duplication set false, log lines that would normally be written
2179 both the the main log and to the reject log were not written to syslog at
2182 PH/14 In the default configuration, change the use of "message" in ACL warn
2183 statements to "add_header".
2185 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2186 not followed by a command (e.g. "seen endif").
2188 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2189 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2192 PH/17 Added extra conditions to the default value of delay_warning_condition
2196 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2197 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2198 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2201 The Auto-Submitted: and various List- headers are standardised, whereas I
2202 don't think Precedence: ever was.
2204 PH/18 Refactored debugging code in route_finduser() to show more information,
2205 in particular, the error code if getpwnam() issues one.
2207 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2208 This is apparently needed in addition to the PH/07 change above to avoid
2209 any possible encoding problems.
2211 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2212 but not after initializing Perl.
2214 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2215 output them only if debugging. By default they are written stderr,
2216 apparently, which is not desirable.
2218 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2221 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2224 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2226 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2227 authenticated or an ident call has been made. Suppress the default
2228 values for $authenticated_id and $authenticated_sender (but permit -oMai
2229 and -oMas) when testing with -bh.
2231 PH/24 Re-jigged the order of the tests in the default configuration so that the
2232 tests for valid domains and recipients precede the DNS black list and CSA
2233 tests, on the grounds that those ones are more expensive.
2235 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2236 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2237 This bug exists in every version of Exim that I still have, right back to
2240 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2241 However, an attempt to turn on case-sensitivity in a regex key by
2242 including (?-i) didn't work because the subject string was already
2243 lowercased, and the effects were non-intuitive. It turns out that a
2244 one-line patch can be used to allow (?-i) to work as expected.
2250 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2251 other effects) broke the use of negated acl sub-conditions.
2253 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2256 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2257 "Deny" causes Exim to reject the incoming connection with a 554 error.
2258 Unfortunately, if there is a major crisis, such as a disk failure,
2259 tcp-wrappers gives "deny", whereas what one would like would be some
2260 kind of temporary error. A kludge has been added to help with this.
2261 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2262 554 error is used if errno is still zero or contains ENOENT (which occurs
2263 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2266 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2268 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2269 errors. Otherwise a message that provokes a temporary error (when other
2270 messages do not) can cause a whole host to time out.
2272 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2273 addresses were routed directly to files or pipes from a redirect router.
2274 File deliveries just didn't batch; pipe deliveries might have suffered
2277 PH/06 A failure to get a lock for a hints database would erroneously always say
2278 "Failed to get write lock", even when it was really a read lock.
2280 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2281 of 0600. This has been changed to use the value of the lockfile_mode
2282 option (which defaults to 0600).
2284 PH/08 Applied small patch from the Sieve maintainer.
2286 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2287 folder from quota calculations, a direct delivery into this folder messed
2288 up the contents of the maildirsize file. This was because the regex was
2289 used only to exclude .Trash (or whatever) when the size of the mailbox
2290 was calculated. There was no check that a delivery was happening into an
2291 excluded directory. This bug has been fixed by ignoring all quota
2292 processing for deliveries into excluded directories.
2294 PH/10 Added the maildirfolder_create_regex option to appendfile.
2300 PH/01 The code for finding all the local interface addresses on a FreeBSD
2301 system running IPv6 was broken. This may well have applied to all BSD
2302 systems, as well as to others that have similar system calls. The broken
2303 code found IPv4 interfaces correctly, but gave incorrect values for the
2304 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2305 that it would not match correctly against @[] and not recognize the IPv6
2308 PH/02 The ipliteral router was not recognizing addresses of the form user@
2309 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2311 PH/03 Added disable_ipv6.
2313 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2314 decoded form, because it is most often used to construct To: headers
2315 lines in autoreplies, and the decoded form may well be syntactically
2316 invalid. However, $reply_address has leading white space removed, and all
2317 newlines turned into spaces so that the autoreply transport does not
2320 PH/05 If group was specified without a user on a router, and no group or user
2321 was specified on a transport, the group from the router was ignored.
2323 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2324 for visible compile-time settings that can be used to change these
2325 numbers, for those that want even more. Backwards compatibility with old
2326 spool files has been maintained. However, going back to a previous Exim
2327 release will lost any variables that are in spool files.
2329 PH/07 Two small changes when running in the test harness: increase delay when
2330 passing a TCP/IP connection to a new process, in case the original
2331 process has to generate a bounce, and remove special handling of
2332 127.0.0.2 (sic), which is no longer necessary.
2334 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2335 be the same on different OS.
2337 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2340 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2341 whether --show-vars was specified or not
2343 JJ/02 exipick: Added support for new ACL variable spool format introduced
2346 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2347 syntactically invalid From: or Reply-to: line, and a filter used this to
2348 generate an autoreply, and therefore failed to obtain an address for the
2349 autoreply, Exim could try to deliver to a non-existent relative file
2350 name, causing unrelated and misleading errors. What now happens is that
2351 it logs this as a hard delivery error, but does not attempt to create a
2354 PH/11 The exinext utility has a -C option for testing purposes, but although
2355 the given file was scanned by exinext itself; it wasn't being passed on
2356 when Exim was called.
2358 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2359 an end-of-file indication when reading a command response.
2361 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2362 compiled. In many other places in Exim, IPv6 addresses are always
2363 recognized, so I have changed this. It also means that IPv4 domain
2364 literals of the form [IPV4:n.n.n.n] are now always recognized.
2366 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2367 used if the router is not running as root, for example, when verifying at
2368 ACL time, or when using -bh. The debugging output from this situation was
2369 non-existent - all you got was a failure to exec. I have made two
2372 (a) Failures to set uid/gid, the current directory, or a process leader
2373 in a subprocess such as that created by queryprogram now generate
2374 suitable debugging ouput when -d is set.
2376 (b) The queryprogram router detects when it is not running as root,
2377 outputs suitable debugging information if -d is set, and then runs
2378 the subprocess without attempting to change uid/gid.
2380 PH/15 Minor change to Makefile for building test_host (undocumented testing
2383 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2384 additional section of a DNS packet that returns MX or SRV records.
2385 Instead, it always explicitly searches for A/AAAA records. This avoids
2386 major problems that occur when a DNS server includes only records of one
2387 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2388 fixed another bug: if SRV records were looked up and the corresponding
2389 address records were *not* found in the additional section, the port
2390 values from the SRV records were lost.
2392 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2393 using the correct key (the original address) when searching the retry
2394 rules in order to find which one to use for generating the retry hint.
2396 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2397 adding the default one. Similarly, if it contains a Reply-To: header, the
2398 errors_reply_to option, if set, is not used.
2400 PH/19 When calculating a retry time, Exim used to measure the "time since
2401 failure" by looking at the "first failed" field in the retry record. Now
2402 it does not use this if it is later than than the arrival time of the
2403 message. Instead it uses the arrival time. This makes for better
2404 behaviour in cases where some deliveries succeed, thus re-setting the
2405 "first failed" field. An example is a quota failure for a huge message
2406 when small messages continue to be delivered. Without this change, the
2407 "time since failure" will always be short, possible causing more frequent
2408 delivery attempts for the huge message than are intended.
2409 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2411 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2412 $1, $2, $3) because the numerical variables can be reset during some
2413 expansion items (e.g. "match"), thereby losing the authentication data.
2415 PH/21 Make -bV show the size of off_t variables so that the test suite can
2416 decide whether to run tests for quotas > 2G.
2418 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2419 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2420 filecount value is greater than 2G or if a quota value is greater than 2G
2421 on a system where the size of off_t is not greater than 4, a panic error
2424 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2425 never match. The debug and -bh output now contains an explicit error
2426 message indicating a malformed IPv4 address or mask.
2428 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2429 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2430 PH/23 above applies.
2432 PH/25 Do not write to syslog when running in the test harness. The only
2433 occasion when this arises is a failure to open the main or panic logs
2434 (for which there is an explicit test).
2436 PH/26 Added the /no_tell option to "control=freeze".
2438 PH/27 If a host name lookup failed very early in a connection, for example, if
2439 the IP address matched host_lookup and the reverse lookup yielded a name
2440 that did not have a forward lookup, an error message of the form "no IP
2441 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2442 could be logged. Now it outputs the IP address instead of "NULL".
2444 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2445 allows the sender and the authenticated sender to be set when
2446 submitting a message from within Exim. Since child_open_exim() is
2447 documented for local_scan(), the new function should be too.
2449 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2450 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2451 results in an empty string is now treated as unset.
2453 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2455 PH/31 Added sender_verify_fail logging option.
2457 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2458 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2459 tidied the source and removed it altogether.
2461 PH/33 When a queue run was abandoned because the load average was too high, a
2462 log line was always written; now it is written only if the queue_run log
2463 selector is set. In addition, the log line for abandonment now contains
2464 information about the queue run such as the pid. This is always present
2465 in "start" and "stop" lines but was omitted from the "abandon" line.
2467 PH/34 Omit spaces between a header name and the colon in the error message that
2468 is given when verify = headers_syntax fails (if there are lots of them,
2469 the message gets confusing).
2471 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2472 names, as there are now some PTR records that contain slashes. This check
2473 is only to protect against broken name servers that fall over on strange
2474 characters, so the fact that it applies to all lookups doesn't matter.
2476 PH/36 Now that the new test suite is complete, we can remove some of the
2477 special code in Exim that was needed for the old test suite. For example,
2478 sorting DNS records because real resolvers return them in an arbitrary
2479 order. The new test suite's fake resolver always returns records in the
2482 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2483 bounces) except when queue_only is set, to avoid logging races between
2484 the different processes.
2486 PH/38 Panic-die if .include specifies a non-absolute path.
2488 PH/39 A tweak to the "H" retry rule from its user.
2490 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2491 a label. They prevented compilation on older perls.
2493 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2494 a warning to be raised on newish perls.
2496 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2497 on queue. Changes to match documented behaviour of showing count of
2498 messages matching specified criteria.
2500 PH/40 Changed the default ident timeout from 30s to 5s.
2502 PH/41 Added support for the use of login_cap features, on those BSD systems
2503 that have them, for controlling the resources used by pipe deliveries.
2505 PH/42 The content-scanning code uses fopen() to create files in which to put
2506 message data. Previously it was not paying any attention to the mode of
2507 the files. Exim runs with umask(0) because the rest of the code creates
2508 files with open(), and sets the required mode explicitly. Thus, these
2509 files were ending up world-writeable. This was not a big issue, because,
2510 being within the spool directory, they were not world-accessible. I have
2511 created a function called modefopen, which takes an additional mode
2512 argument. It sets umask(777), creates the file, chmods it to the required
2513 mode, then resets the umask. All the relevant calls to fopen() in the
2514 content scanning code have been changed to use this function.
2516 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2517 to 24 hours. This avoids potential overflow problems when processing G
2518 and H retry rules. I suspect nobody ever tinkers with this value.
2520 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2522 PH/45 When the plaintext authenticator is running as a client, the server's
2523 challenges are checked to ensure they are valid base64 strings. By
2524 default, the authentication attempt is cancelled if an invalid string is
2525 received. Setting client_ignore_invalid_base64 true ignores these errors.
2526 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2527 they are received. Thus, the responses can be made to depend on the
2528 challenges. If an invalid string is ignored, an empty string is placed in
2531 PH/46 Messages that are created by the autoreply transport now contains a
2532 References: header, in accordance with RFCs 2822 and 3834.
2534 PH/47 Added authenticated_sender_force to the smtp transport.
2536 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2538 PH/49 Installed latest patch from the Sieve maintainer.
2540 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2541 was also set, the appendfile transport was unnecessarily scanning a
2542 directory of message files (e.g. for maildir delivery) to find the count
2543 of files (along with the size), even though it did not need this
2544 information. It now does the scan only if it needs to find either the
2545 size of the count of files.
2547 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2549 PH/52 Two bugs concerned with error handling when the smtp transport is
2552 (i) Exim was not creating retry information for temporary errors given
2553 for individual recipients after the DATA command when the smtp transport
2554 was used in LMTP mode. This meant that they could be retried too
2555 frequently, and not timed out correctly.
2557 (ii) Exim was setting the flag that allows error details to be returned
2558 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2559 recipients that were returned after the DATA command.
2561 PH/53 This is related to PH/52, but is more general: for any failing address,
2562 when detailed error information was permitted to be returned to the
2563 sender, but the error was temporary, then after the final timeout, only
2564 "retry timeout exceeded" was returned. Now it returns the full error as
2565 well as "retry timeout exceeded".
2567 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2568 do this, and (what is worse) MTAs that accept it.
2570 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2571 will now be deprecated.
2573 PH/56 New os.c-cygwin from the Cygwin maintainer.
2575 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2576 formats (previously only available in exim formats via -bpr, -bpru,
2577 and -bpra. Now also available in native and exiqgrep formats)
2579 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2580 with very large, slow to parse queues
2582 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2584 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2586 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2587 responses to authentication challenges, though it was showing the
2588 challenges; (ii) I've removed the CR characters from the debug output for
2591 PH/58 Allow for the insertion of a newline as well as a space when a string
2592 is turned into more than one encoded-word during RFC 2047 encoding. The
2593 Sieve code now uses this.
2595 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2596 data_4xx, lost_connection, tls_required.
2598 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2599 message was being sent as an SMTP response.
2601 PH/61 Add -l and -k options to exicyclog.
2603 PH/62 When verifying, if an address was redirected to one new address, so that
2604 verification continued, and the new address failed or deferred after
2605 having set something in $address_data, the value of $address_data was not
2606 passed back to the ACL. This was different to the case when no
2607 redirection occurred. The value is now passed back in both cases.
2609 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2610 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2611 use login_cap.h, so on its own it isn't the distinguishing feature. The
2612 new name refers directly to the setclassresources() function.
2614 PH/65 Added configuration files for NetBSD3.
2616 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2618 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2619 is preferred over IPv4.
2621 PH/68 The bounce_return_message and bounce_return_body options were not being
2622 honoured for bounces generated during the reception of non-SMTP messages.
2623 In particular, this applied to messages rejected by the ACL. This bug has
2624 been fixed. However, if bounce_return_message is true and bounce_return_
2625 body is false, the headers that are returned for a non-SMTP message
2626 include only those that have been read before the error was detected.
2627 (In the case of an ACL rejection, they have all been read.)
2629 PH/69 The HTML version of the specification is now built in a directory called
2630 spec_html instead of spec.html, because the latter looks like a path with
2631 a MIME-type, and this confuses some software.
2633 PH/70 Catch two compiler warnings in sieve.c.
2635 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2636 function verify_get_ident() calls ip_connect() to connect a socket, but
2637 if the "connect()" function timed out, ip_connect() used to close the
2638 socket. However, verify_get_ident() also closes the socket later, and in
2639 between Exim writes to the log, which may get opened at this point. When
2640 the socket was closed in ip_connect(), the log could get the same file
2641 descriptor number as the socket. This naturally causes chaos. The fix is
2642 not to close the socket in ip_connect(); the socket should be closed by
2643 the function that creates it. There was only one place in the code where
2644 this was missing, in the iplookup router, which I don't think anybody now
2645 uses, but I've fixed it anyway.
2647 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2648 well as to direct DNS lookups. Otherwise the handling of names in host
2649 lists is inconsistent and therefore confusing.
2655 PH/01 Two changes to the default runtime configuration:
2657 (1) Move the checks for relay_from_hosts and authenticated clients from
2658 after to before the (commented out) DNS black list checks.
2660 (2) Add control=submission to the relay_from_hosts and authenticated
2661 clients checks, on the grounds that messages accepted by these
2662 statements are most likely to be submissions.
2664 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2666 (1) Generate an error if the third argument for the ${prvs expansion is
2669 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2672 (3) Reset the variables that are obtained from the first argument of
2673 ${prvscheck and used in the second argument before leaving the code,
2674 because their memory is reclaimed, so using them afterwards may do
2677 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2678 one (it's much easier than Tom thought :-).
2680 (5) Because of (4), we can now allow for the use of $prvscheck_result
2681 inside the third argument.
2683 PH/03 For some reason, the default setting of PATH when running a command from
2684 a pipe transport was just "/usr/bin". I have changed it to
2687 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2688 anything to be listed in the output from -bV.
2690 PH/05 When a filter generated an autoreply, the entire To: header line was
2691 quoted in the delivery log line, like this:
2693 => >A.N.Other <ano@some.domain> <original@ddress> ...
2695 This has been changed so that it extracts the operative address. There
2696 may be more than one such address. If so, they are comma-separated, like
2699 => >ano@some.domain,ona@other.domain <original@ddress> ...
2701 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2702 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2703 not being looked up in the rDNS to get a host name, Exim was showing the
2704 IP address twice in Received: lines, even though the IP addresses were
2705 identical. For example:
2707 Received: from [1.2.3.4] (helo=[1.2.3.4])
2709 However, if the real host name was known, it was omitting the HELO data
2710 if it matched the actual IP address. This has been tidied up so that it
2711 doesn't show the same IP address twice.
2713 PH/07 When both +timestamp and +memory debugging was on, the value given by
2714 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2715 called by the debug printing, thereby overwriting the timestamp buffer.
2716 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2718 PH/08 When the original message was included in an autoreply transport, it
2719 always said "this is a copy of the message, including all the headers",
2720 even if body_only or headers_only was set. It now gives an appropriate
2723 PH/09 Applied a patch from the Sieve maintainer which:
2725 o fixes some comments
2726 o adds the (disabled) notify extension core
2727 o adds some debug output for the result of if/elsif tests
2728 o points to the current vacation draft in the documentation
2729 and documents the missing references header update
2733 o fixes a bug in processing the envelope test (when testing
2734 multiple envelope elements, the last element determinted the
2737 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2738 Electronic Mail") by including:
2740 Auto-submitted: auto-generated
2742 in the messages that it generates (bounce messages and others, such as
2743 warnings). In the case of bounce messages for non-SMTP mesages, there was
2744 also a typo: it was using "Auto_submitted" (underscore instead of
2745 hyphen). Since every message generated by Exim is necessarily in response
2746 to another message, thes have all been changed to:
2748 Auto-Submitted: auto-replied
2750 in accordance with these statements in the RFC:
2752 The auto-replied keyword:
2754 - SHOULD be used on messages sent in direct response to another
2755 message by an automatic process,
2757 - MUST NOT be used on manually-generated messages,
2759 - MAY be used on Delivery Status Notifications (DSNs) and Message
2760 Disposition Notifications (MDNs),
2762 - MUST NOT be used on messages generated by automatic or periodic
2763 processes, except for messages which are automatic responses to
2766 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2767 to the default Received: header definition.
2769 PH/12 Added log selector acl_warn_skipped (default on).
2771 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2772 variables because (a) they are in the wrong storage pool and (b) even if
2773 they were copied, it wouldn't work properly because of the caching.
2775 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2776 checking when decoding. Apparently there are clients that generate
2777 overlong encoded strings. Why am I not surprised?
2779 PH/15 If the first argument of "${if match_address" was not empty, but did not
2780 contain an "@" character, Exim crashed. Now it writes a panic log message
2781 and treats the condition as false.
2783 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2785 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2786 extension "envelope-auth". The code is finished and in agreement with
2787 other implementations, but there is no documentation so far and in fact,
2788 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2789 not changing the active code.
2791 Print executed "if" and "elsif" statements when debugging is used. This
2792 helps a great deal to understand what a filter does.
2794 Document more things not specified clearly in RFC3028. I had all this
2795 sorted out, when out of a sudden new issues came to my mind. Oops."
2797 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2800 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2801 canonical form (as documented). However, after a host name lookup from
2802 the IP address, check_host() was doing a simple string comparison with
2803 addresses acquired from the DNS when checking that the found name did
2804 have the original IP as one of its addresses. Since any found IPv6
2805 addresses are likely to be in abbreviated form, the comparison could
2806 fail. Luckily, there already exists a function for doing the comparison
2807 by converting both addresses to binary, so now that is used instead of
2808 the text comparison.
2810 PH/20 There was another similar case to PH/19, when a complete host name was
2811 given in a host list; looking up its IP address could give an abbreviated
2812 form, whereas the current host's name might or might not be abbreviated.
2813 The same fix has been applied.
2819 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2820 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2823 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2824 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2826 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2828 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2829 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2830 Also, for "Auto-Submitted", treat anything other than "no" as
2831 non-personal, in accordance with RFC 3834. (Previously it treated
2832 anything starting "auto-" as non-personal.)
2834 TF/01 The control=submission/name=... option had a problem with syntax
2835 errors if the name included a slash character. The /name= option
2836 now slurps the rest of the string, so it can include any characters
2837 but it must come last in the list of options (after /sender_retain
2840 PH/05 Some modifications to the interface to the fake nameserver for the new
2848 TK/01 Added the "success_on_redirect" address verification option. See
2849 NewStuff for rationale and an example.
2851 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2853 PH/02 Patch to exigrep to allow it to work on syslog lines.
2855 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2856 fread() to skip over the body file's header line, because in Cygwin the
2857 header line is locked and is inaccessible.
2859 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2860 co-exist for some time) to make it clear that it is the Exim ID that is
2861 referenced, not the Message-ID: header line.
2863 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2864 string_format() function, because snprintf() does not exist on all
2867 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2868 ${stat: expansion item.
2870 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2871 protocol synchronization error", to keep the pedants happy.
2873 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2874 well as for IRIX systems, when gcc is being used. See the host.c source
2877 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2879 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2882 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2883 transport and to the smtp transport in LMTP mode.
2885 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2887 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2888 run a filter in a subprocess. This could lead to confusion in subsequent
2889 lookups in the parent process. There should also be a search_tidyup() at
2890 the end of the subprocess.
2892 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2893 only if the host matched helo_try_verify_hosts, which caused the
2894 verification to occur when the EHLO/HELO command was issued. The ACL just
2895 tested the remembered result. Now, if a previous verification attempt has
2896 not happened, "verify = helo" does it there and then.
2898 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2900 TK/03 Fix log output including CR from clamd.
2902 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2903 exist provoked a memory error which could cause a segfault.
2905 PH/15 Installed PCRE 6.2
2907 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2909 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2910 of the problem. Specifically, suggested +O2 rather than +O1 for the
2913 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2915 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2916 record, the port was not correctly logged when the outgoing_port log
2917 selector was set (it logged the transort's default port).
2919 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2920 fallback_hosts, and "hosts" in the smtp transport.
2922 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2923 host errors such as "Connection refused".
2925 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2926 authentication with radiusclient 0.4.9:
2928 - Error returned from rc_read_config was caught wrongly
2929 - Username/password not passed on to radius server due to wrong length.
2931 The presumption is that some radiusclient API changes for 4.51/PH/17
2932 were not taken care of correctly. The code is still untested by me (my
2933 Linux distribution still has 0.3.2 of radiusclient), but it was
2934 contributed by a Radius user.
2936 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2937 expanding the "port" option of the smtp transport.
2939 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2940 while reading a MIME header. Thanks to Tom Hughes for a patch.
2942 PH/24 Include config.h inside local_scan.h so that configuration settings are
2945 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2946 that in an ACL for RCPT (for example), you can examine exactly what was
2949 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2950 commands, but it was not correctly comparing the address with the actual
2951 client host address. Thus, it would show the EHLO address in Received:
2952 header lines when this was not necessary.
2954 PH/27 Added the % operator to ${eval:}.
2956 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2957 it should be ignoring failures (because with -C, for example, it has lost
2958 privilege). It wasn't ignoring creation failures other than "already
2961 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2964 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2965 because an input line was too long, either on its own, or by virtue of
2966 too many continuations, the temporary file was not being removed, and the
2967 return code was incorrect.
2969 PH/31 Missing "BOOL" in function definition in filtertest.c.
2971 PH/32 Applied Sieve patches from the maintainer.
2973 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2975 PH/33 Added "verify = not_blind".
2977 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2978 Local/Makefile (with some defaults set). These are used in built scripts
2979 such as exicyclog, but they have never been used in the exim_install
2980 script (though there are many overriding facilities there). I have
2981 arranged that the exim_install script now takes note of these two
2984 PH/35 Installed configuration files for Dragonfly.
2986 PH/36 When a locally submitted message by a trusted user did not contain a
2987 From: header, and the sender address was obtained from -f or from an SMTP
2988 MAIL command, and the trusted user did not use -F to supply a sender
2989 name, $originator_name was incorrectly used when constructing a From:
2990 header. Furthermore, $originator_name was used for submission mode
2991 messages from external hosts without From: headers in a similar way,
2992 which is clearly wrong.
2994 PH/37 Added control=suppress_local_fixups.
2996 PH/38 When log_selector = +received_sender was set, and the addition of the
2997 sender made the log line's construction buffer exactly full, or one byte
2998 less than full, an overflow happened when the terminating "\n" was
3001 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3002 when the result of a list match is failure because a DNS lookup failed.
3004 PH/40 RM_COMMAND is now used in the building process.
3006 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3007 the "build-* directories that it finds.
3009 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3010 address in a domain literal was a prefix of an interface address.
3012 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3013 when verifying a sender address, unless rewrite_headers is false.
3015 PH/44 Wrote a long comment about why errors_to addresses are verified as
3016 recipients, not senders.
3018 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3019 the ratelimit ACL was added.
3021 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3023 PH/46 Added extra information about PostgreSQL errors to the error string.
3025 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3026 avoiding the need to install special zones in a real server. This is
3027 backwards compatible; if it can't find the fake resolver, it drops back.
3028 Thus, both old and new test suites can be run.
3030 TF/02 Added util/ratelimit.pl
3032 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3033 clock is set back in time.
3035 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3036 Candler <B.Candler@pobox.com>.
3038 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3039 OK for addresses that are the result of redirections.
3041 PH/48 A number of further additions for the benefit of the new test suite,
3042 including a fake gethostbyname() that interfaces to the fake DNS resolver
3045 TF/06 The fix for widen_domains has also been applied to qualify_single and
3046 search_parents which are the other dnslookup options that can cause
3049 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3052 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3054 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3055 service" if it encountered an MX record with an empty target hostname.
3056 The message is now "an MX or SRV record indicated no SMTP service".
3058 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3059 if the DNS of the sending site is misconfigured. This is quite a
3060 common situation. This change restores the behaviour of treating a
3061 helo verification defer as a failure.
3063 PH/51 If self=fail was set on a router, the bounce message did not include the
3064 actual error message.
3070 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3072 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3073 command itself ended in error, the underlying message about the transport
3074 filter timeout was being overwritten with the pipe command error. Now the
3075 underlying error message should be appended to the second error message.
3077 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3079 PH/02 Reduce the amount of output that "make" produces by default. Full output
3080 can still be requested.
3082 PH/03 The warning log line about a condition test deferring for a "warn" verb
3083 was being output only once per connection, rather than after each
3084 occurrence (because it was using the same function as for successful
3085 "warn" verbs). This seems wrong, so I have changed it.
3087 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3088 it should not have, which might have caused a crash in the right
3089 circumstances, but probably never did.
3091 PH/04 Installed a modified version of Tony Finch's patch to make submission
3092 mode fix the return path as well as the Sender: header line, and to
3093 add a /name= option so that you can make the user's friendly name appear
3096 TF/03 Added the control = fakedefer ACL modifier.
3098 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3099 Mark Lowes for thorough testing.
3101 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3103 TK/03 Merged latest SRS patch from Miles Wilton.
3105 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3106 with the definition in sysexits.h (which is #included earlier).
3107 Fortunately, Exim does not actually use EX_OK. The code used to try to
3108 preserve the sysexits.h value, by assumimg that macro definitions were
3109 scanned for macro replacements. I have been disabused of this notion,
3110 so now the code just undefines EX_OK before #including unistd.h.
3112 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3113 in the smtp transport. When a block could not be written in a single
3114 write() function, the timeout was being re-applied to each part-write.
3115 This seems wrong - if the receiver was accepting one byte at a time it
3116 would take for ever. The timeout is now adjusted when this happens. It
3117 doesn't have to be particularly precise.
3119 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3120 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3122 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3123 without a domain if the check to <postmaster@domain> fails.
3125 SC/01 Eximstats: added -xls and the ability to specify output files
3126 (patch written by Frank Heydlauf).
3128 SC/02 Eximstats: use FileHandles for outputing results.
3130 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3132 SC/04 Eximstats: fixed display of large numbers with -nvr option
3134 SC/05 Eximstats: fixed merging of reports with empty tables.
3136 SC/06 Eximstats: added the -include_original_destination flag
3138 SC/07 Eximstats: removed tabs and trailing whitespace.
3140 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3142 TK/06 MBOX spool code: Add real "From " MBOX separator line
3143 so the .eml file is really in mbox format (even though
3144 most programs do not really care). Patch from Alex Miller.
3146 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3147 The latter is generated from $received_to and is only set if the
3148 message has one envelope recipient. SA can use these headers,
3149 obviously out-of-the-box. Patch from Alex Miller.
3151 PH/08 The ${def test on a variable was returning false if the variable's
3152 value was "0", contrary to what the specification has always said!
3153 The result should be true unless the variable is empty.
3155 PH/09 The syntax error of a character other than { following "${if
3156 def:variable_name" (after optional whitespace) was not being diagnosed.
3157 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3158 accidental colon was present, for example, could give incorrect results.
3160 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3161 result is used (not including appendfile, where other changes are about
3164 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3165 This involved changing a lot of size variables from int to off_t. It
3166 should work with maildirs and everything.
3168 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3169 spamd dying while we are connected to it.
3171 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3174 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3175 function for BDB 4.3.
3177 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3179 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3180 the sources for PCRE have been re-arranged and more files are now
3183 PH/15 The code I had for printing potentially long long variables in PH/11
3184 above was not the best (it lost precision). The length of off_t variables
3185 is now inspected at build time, and an appropriate printing format (%ld
3186 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3187 to be "long long int" or "long int". This is needed for the internal
3188 formatting function string_vformat().
3190 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3191 the configuration file to be ":syslog", then the script "guesses" where
3192 the logs files are, rather than using the compiled in default. In our
3193 case the guess is not the same as the compiled default, so the script
3194 suddenly stopped working when I started to use syslog. The patch checks
3195 to see if log_file_path is "". If so, it attempts to read it from exim
3196 with no configuration file to get the compiled in version, before it
3197 falls back to the previous guessing code."
3199 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3200 implementing BATV in an Exim configuration. See NewStuff for the gory
3203 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3204 Makefile that are specific to HP-UX.
3206 PH/18 If the "use_postmaster" option was set for a recipient callout together
3207 with the "random" option, the postmaster address was used as the MAIL
3208 FROM address for the random test, but not for the subsequent recipient
3209 test. It is now used for both.
3211 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3212 patch removes a few documentation additions to RFC 3028, because the
3213 latest draft now contains them. It adds the new en;ascii-case comparator
3214 and a new error check for 8bit text in MIME parts. Comparator and
3215 require names are now matched exactly. I enabled the subaddress
3216 extension, but it is not well tested yet (read: it works for me)."
3218 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3219 rework some of the code of TK/09 above to avoid the hardwired use of
3220 "%lld" and "long long". Replaced the call to snprintf() with a call to
3223 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3224 records point to non-existent hosts", "retry timeout exceeded", and
3225 "retry time not reached for any host after a long failure period".
3227 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3228 experimental DomainKeys support:
3230 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3231 (2) On an error such as an illegally used "control", the wrong name for
3232 the control was given.
3234 These problems did NOT occur unless DomainKeys support was compiled.
3236 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3238 PH/24 Added ${if match_ip condition.
3240 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3241 databases so that it will be absolutely obvious if a crash occurs in the
3242 DB library. This is a regular occurrence (often caused by mis-matched
3245 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3246 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3247 picked up on a user's system that detects such things. There doesn't seem
3248 to be a gcc warning option for this - only an attribute that has to be
3249 put on the function's prototype. It seems that in Fedora Core 4 they have
3250 set this on a number of new functions. No doubt there will be more in due
3253 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3254 specify a transport. However, if an address that was verified by such a
3255 router was the subject of a callout, Exim crashed because it tried to
3256 read the rcpt_include_affixes from the non-existent transport. Now it
3257 just assumes that the setting of that option is false. This bug was
3258 introduced by 4.51/PH/31.
3260 PH/28 Changed -d+all to exclude +memory, because that information is very
3261 rarely of interest, but it makes the output a lot bigger. People tend to
3262 do -d+all out of habit.
3264 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3265 code in os-type was giving problems when libc.so lives in lib64, like on
3268 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3269 aren't the modern standard, and it seems that some systems' include files
3270 don't always have them. Exim was already checking for some of the newer
3271 ones like T_AAAA, and defining it itself. I've added checks for all the
3272 record types that Exim uses.
3274 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3275 not automatically generating a new one, as it is supposed to. This
3276 prevented TLS from working. If the file did exist, but contained invalid
3277 data, a new version was generated, as expected. It was only the case of a
3278 non-existent file that was broken.
3280 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3281 with a change in libdomainkeys > 0.64.
3283 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3284 from DNS. If the selector record carries the flag, it now has
3285 precedence over the domain-wide flag.
3287 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3289 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3290 the use of a port name that isn't defined in /etc/services) occurred, the
3291 message was deferred as in a normal delivery, and thus remained on the
3292 spool, instead of being failed because of the mua_wrapper setting. This
3293 is now fixed, and I tidied up some of the mua_wrapper messages at the
3296 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3297 the messages in a hash of arrays rather than using individual hashes.
3298 This is a bit cleaner and results in dramatic memory savings, albeit
3299 at a slight CPU cost.
3301 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3302 as requested by Marc Sherman.
3304 SC/10 Eximstats: added histograms for user specified patterns as requested
3307 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3309 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3310 fopen() in the content-scanning modules that did not already have it.
3316 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3317 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3319 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3321 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3323 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3324 to test Sieve filters that use "vacation".
3326 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3327 that changes the way the GnuTLS parameters are stored in the cache file.
3328 The new format can be generated externally. For backward compatibility,
3329 if the data in the cache doesn't make sense, Exim assumes it has read an
3330 old-format file, and it generates new data and writes a new file. This
3331 means that you can't go back to an older release without removing the
3334 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3335 work if there are any delivery delays because "one_time" forces the
3336 parent to be marked "delivered", so its unseen clone is never tried
3337 again. For this reason, Exim now forbids the simultaneous setting of
3340 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3341 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3342 entry if you want to know the details. The fix, however, neglected to
3343 consider the case when local delivery batching is involved. The test for
3344 "previously delivered" was not happening when checking to see if an
3345 address could be batched with a previous (undelivered) one; under
3346 certain circumstances this could lead to multiple deliveries to the same
3349 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3350 in its include files, and this causes problems building Exim.
3352 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3353 header_syntax) but Exim was just ignoring anything given after a slash.
3354 In particular, this caused confusion with an attempt to use "verify =
3355 reverse_host_lookup/defer_ok". An error is now given when options are
3356 supplied for verify items that do not have them. (Maybe reverse_host_
3357 lookup should have a defer_ok option, but that's a different point.)
3359 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3360 defined by RFC 821) to 2048, because there were problems with some AUTH
3361 commands, and RFC 1869 says the size should be increased for extended
3362 SMTP commands that take arguments.
3364 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3367 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3368 "unknown" error; now it says that the functionality isn't in the binary.
3370 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3371 an address' error message when a string expansion fails (syntax or
3372 whatever). Otherwise the password may appear in the log. Following change
3373 PH/42 below, there is no longer a chance of it appearing in a bounce
3376 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3378 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3379 of its addresses was ever tried. (Bugzilla bug #2.)
3381 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3382 the result incorrectly in the debug output. (It correctly added a newline
3383 to what was transported.)
3385 TF/01 Added $received_time.
3387 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3388 commented out examples of how to interface to a virus scanner and to
3389 SpamAssassin. Also added commented examples of av_scanner and
3390 spamd_address settings.
3392 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3393 and controls are allowed in which ACLs. There were a couple of minor
3394 errors. Some of the entries in the conditions table (which is a table of
3395 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3396 negation of where the condition IS allowed.
3398 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3400 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3401 header file does not have a version number, so I've had to invent a new
3402 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3403 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3404 radiusclient), but it was contributed by a Radius user.
3406 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3407 files or directories, for OpenSSL.
3409 PH/19 When an Exim process that is running as root has to create an Exim log
3410 file, it does so in a subprocess that runs as exim:exim so as to get the
3411 ownership right at creation (otherwise, other Exim processes might see
3412 the file with the wrong ownership). There was no test for failure of this
3413 fork() call, which would lead to the process getting stuck as it waited
3414 for a non-existent subprocess. Forks do occasionally fail when resources
3415 run out. I reviewed all the other calls to fork(); they all seem to check
3418 PH/20 When checking for unexpected SMTP input at connect time (before writing
3419 the banner), Exim was not dealing correctly with a non-positive return
3420 from the read() function. If the client had disconnected by this time,
3421 the result was a log entry for a synchronization error with an empty
3422 string after "input=" when read() returned zero. If read() returned -1
3423 (an event I could not check), uninitialized data bytes were printed.
3424 There were reports of junk text (parts of files, etc) appearing after
3427 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3429 PH/22 Added support for macro redefinition, and (re)definition in between
3430 driver and ACL definitions.
3432 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3433 forgetting to use the resulting value; it was using the unexpanded value.
3435 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3436 hadn't been configured. The fix is from Juergen Kreileder, who
3437 understands it better than I do:
3439 "Here's what I see happening with three configured cyrus_sasl
3440 authenticators configured (plain, login, cram-md5):
3442 On startup auth_cyrus_sasl_init() gets called for each of these.
3443 This means three calls to sasl_listmech() without a specified mech_list.
3444 => SASL tests which mechs of all available mechs actually work
3445 => three warnings about OTP not working
3446 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3448 With the patch, sasl_listmech() also gets called three times. But now
3449 SASL's mech_list option is set to the server_mech specified in the the
3450 authenticator. Or in other words, the answer from sasl_listmech()
3451 gets limited to just the mech you're testing for (which is different
3453 => the return list contains just 'plain' or 'login', 'cram-md5' or
3454 nothing depending on the value of ob->server_mech.
3456 I've just tested the patch: Authentication still works fine,
3457 unavailable mechs specified in the exim configuration are still
3458 caught, and the auth.log warnings about OTP are gone."
3460 PH/25 When debugging is enabled, the contents of the command line are added
3461 to the debugging output, even when log_selector=+arguments is not
3464 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3465 answer is "GNU", and only if the return is "GNU/something" is the answer
3468 PH/27 $acl_verify_message is now set immediately after the failure of a
3469 verification in an ACL, and so is available in subsequent modifiers. In
3470 particular, the message can be preserved by coding like this:
3472 warn !verify = sender
3473 set acl_m0 = $acl_verify_message
3475 Previously, $acl_verify_message was set only while expanding "message"
3476 and "log_message" when a very denied access.
3478 PH/28 Modified OS/os.c-Linux with
3480 -#ifndef OS_LOAD_AVERAGE
3481 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3483 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3484 nomenclature these days.)
3486 PH/29 Installed patch from the Sieve maintainer that adds the options
3487 sieve_useraddress and sieve_subaddress to the redirect router.
3489 PH/30 In these circumstances:
3490 . Two addresses routed to the same list of hosts;
3491 . First host does not offer TLS;
3492 . First host accepts first address;
3493 . First host gives temporary error to second address;
3494 . Second host offers TLS and a TLS session is established;
3495 . Second host accepts second address.
3496 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3497 and peerdn, if requested) that were in fact used only for the second
3500 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3501 attention to any local part prefix or suffix that was matched by the
3502 router that accepted the address. It now behaves in the same way as it
3503 does for delivery: the affixes are removed from the local part unless
3504 rcpt_include_affixes is set on the transport.
3506 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3507 timeout during the DATA phase of an incoming message.
3509 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3510 applied a patch sanctioned by the Sieve maintainer.
3512 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3513 the uid or gid is negative. A case of a negative gid caused this to be
3514 noticed. The fix allows for either to be negative.
3516 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3517 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3520 PH/36 The change PH/12 above was broken. Fixed it.
3522 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3523 the grounds that routing the same address twice would always produce the
3524 same answer. This might have been true once, but it is certainly no
3525 longer true now. Routing a child address may depend on the previous
3526 routing that produced that child. Some complicated redirection strategies
3527 went wrong when messages had multiple recipients, and made Exim's
3528 behaviour dependent on the order in which the addresses were given.
3530 I have moved the duplicate checking until after the routing is complete.
3531 Exim scans the addresses that are assigned to local and remote
3532 transports, and removes any duplicates. This means that more work will be
3533 done, as duplicates will always all be routed, but duplicates are
3534 presumably rare, so I don't expect this is of any significance.
3536 For deliveries to pipes, files, and autoreplies, the duplicate checking
3537 still happens during the routing process, since they are not going to be
3540 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3541 It corrects a timeout issue with spamd. This is Ian's comment: "The
3542 background is that sometimes spamd either never reads data from a
3543 connection it has accepted, or it never writes response data. The exiscan
3544 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3545 blindly assumes that writes won't block so it may never time out."
3547 PH/39 Allow G after quota size as well as K and M.
3549 PH/40 The value set for $authenticated_id in an authenticator may not contain
3550 binary zeroes or newlines because the value is written to log lines and
3551 to spool files. There was no check on this. Now the value is run through
3552 the string_printing() function so that such characters are converted to
3553 printable escape sequences.
3555 PH/41 $message_linecount is a new variable that contains the total number of
3556 lines in the message. Compare $body_linecount, which is the count for the
3559 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3560 bounce and delay warning messages, except in certain special cases, which
3563 (a) An SMTP error message from a remote host;
3564 (b) A message specified in a :fail: redirection;
3565 (c) A message specified in a "fail" command in a system filter;
3566 (d) A message specified in a FAIL return from the queryprogram router;
3567 (e) A message specified by the cannot_route_message router option.
3569 In these cases only, Exim does include the error details in bounce and
3570 warning messages. There are also a few cases where bland messages such
3571 as "unrouteable address" or "local delivery error" are given.
3573 PH/43 $value is now also set for the "else" part of a ${run expansion.
3575 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3576 being worked on, but at least Exim now implements the latest version to
3579 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3580 process to complete was treated as a delivery failure, a timeout while
3581 writing the message to the pipe was logged, but erroneously treated as a
3582 successful delivery. Such timeouts include transport filter timeouts. For
3583 consistency with the overall process timeout, these timeouts are now
3584 treated as errors, giving rise to delivery failures by default. However,
3585 there is now a new Boolean option for the pipe transport called
3586 timeout_defer, which, if set TRUE, converts the failures into defers for
3587 both kinds of timeout. A transport filter timeout is now identified in
3590 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3591 systems where "make" and "gmake" are different, calling "gmake" at top
3592 level broke things. I've arranged for the value of $(MAKE) to be passed
3593 from the Makefile to this script so that it can call the same version of
3597 A note about Exim versions 4.44 and 4.50
3598 ----------------------------------------
3600 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3601 changes of various kinds. As a consequence, a big documentation update was
3602 needed. This delayed the release for rather longer than seemed good, especially
3603 in the light of a couple of (minor) security issues. Therefore, the changes
3604 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3605 release. So 4.44 and 4.50 are in effect two different branches that both start
3608 I have left the 4.50 change log unchanged; it contains all the changes since
3609 4.43. The change log for 4.44 is below; many of its items are identical to
3610 those for 4.50. This seems to be the most sensible way to preserve the
3611 historical information.
3617 1. Minor wording change to the doc/README.SIEVE file.
3619 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3620 computation of the current number of files was incorrect.
3622 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3623 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3626 4. Give more explanation in the error message when the command for a transport
3627 filter fails to execute.
3629 5. There are several places where Exim runs a non-Exim command in a
3630 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3631 was being done only for the command run by the queryprogram router. It is
3632 now done for all such subprocesses. The other cases are: ${run, transport
3633 filters, and the commands run by the lmtp and pipe transports.
3635 6. Added CONFIGURE_GROUP build-time option.
3637 7. Some older OS have a limit of 256 on the maximum number of file
3638 descriptors. Exim was using setrlimit() to set 1000 as a large value
3639 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3640 systems. I've change it so that if it can't get 1000, it tries for 256.
3642 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3643 was an oversight, and furthermore, ever since the addition of extra
3644 controls (e.g. 4.43/32), the checks on when to allow different forms of
3645 "control" were broken. There should now be diagnostics for all cases when a
3646 control that does not make sense is encountered.
3648 9. Added the /retain_sender option to "control=submission".
3650 10. $recipients is now available in the predata ACL (oversight).
3652 11. Tidy the search cache before the fork to do a delivery from a message
3653 received from the command line. Otherwise the child will trigger a lookup
3654 failure and thereby defer the delivery if it tries to use (for example) a
3655 cached ldap connection that the parent has called unbind on.
3657 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3658 of $address_data from the recipient verification was clobbered by the
3659 sender verification.
3661 13. The value of address_data from a sender verification is now available in
3662 $sender_address_data in subsequent conditions in the ACL statement.
3664 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3666 15. Added a new option "connect=<time>" to callout options, to set a different
3669 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3670 was its contents. (It was OK if the option was not defined at all.)
3672 17. A "Completed" log line is now written for messages that are removed from
3673 the spool by the -Mrm option.
3675 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3676 information about exactly what failed.
3678 19. Added -dd to debug only the daemon process.
3680 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3681 handles timeouts, both on the server side and network timeouts. Renamed the
3682 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3684 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3685 It is now set to "smtps".
3687 22. $host_address is now set to the target address during the checking of
3688 ignore_target_hosts.
3690 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3691 being passed; this would have caused $sender_host_name to have been used if
3692 matching the list had actually called for a host name (not very likely,
3693 since this list is usually IP addresses). A host name is now passed as
3696 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3697 code that specifies a non-restarting handler (typically sigaction() in
3698 modern systems) in an attempt to fix a rare and obscure crash bug.
3700 25. Narrowed the window for a race in the daemon that could cause it to ignore
3701 SIGCHLD signals. This is not a major problem, because they are used only to
3702 wake it up if nothing else does.
3704 26. A malformed maildirsize file could cause Exim to calculate negative values
3705 for the mailbox size or file count. Odd effects could occur as a result.
3706 The maildirsize information is now recalculated if the size or filecount
3709 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3710 support for a long time. Removed HAVE_SYS_VFS_H.
3712 28. Installed the latest version of exipick from John Jetmore.
3714 29. In an address list, if the pattern was not a regular expression, an empty
3715 subject address (from a bounce message) matched only if the pattern was an
3716 empty string. Non-empty patterns were not even tested. This was the wrong
3717 because it is perfectly reasonable to use an empty address as part of a
3718 database query. An empty address is now tested by patterns that are
3719 lookups. However, all the other forms of pattern expect the subject to
3720 contain a local part and a domain, and therefore, for them, an empty
3721 address still always fails if the pattern is not itself empty.
3723 30. Exim went into a mad DNS loop when attempting to do a callout where the
3724 host was specified on an smtp transport, and looking it up yielded more
3725 than one IP address.
3727 31. Re-factored the code for checking spool and log partition space into a
3728 function that finds that data and another that does the check. The former
3729 is then used to implement four new variables: $spool_space, $log_space,
3730 $spool_inodes, and $log_inodes.
3732 32. The RFC2047 encoding function was originally intended for short strings
3733 such as real names; it was not keeping to the 75-character limit for
3734 encoded words that the RFC imposes. It now respects the limit, and
3735 generates multiple encoded words if necessary. To be on the safe side, I
3736 have increased the buffer size for the ${rfc2047: expansion operator from
3739 33. It is now permitted to omit both strings after an "if" condition; if the
3740 condition is true, the result is "true". As before, when the second string
3741 is omitted, a false condition yields an empty string. This makes it less
3742 cumbersome to write custom ACL and router conditions.
3744 34. Failure to deliver a bounce message always caused it to be frozen, even if
3745 there was an errors_to setting on the router. The errors_to setting is now
3748 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3749 canonical form (fully expanded) before being placed in
3750 $sender_host_address.
3752 36. The table in the code that translates DNS record types into text (T_A to
3753 "A" for instance) was missing entries for NS and CNAME. It is just possible
3754 that this could have caused confusion if both these types were looked up
3755 for the same domain, because the text type is used as part of Exim's
3756 per-process caching. But the chance of anyone hitting this buglet seems
3759 37. The dnsdb lookup has been extended in a number of ways.
3761 (1) There is a new type, "zns", which walks up the domain tree until it
3762 finds some nameserver records. It should be used with care.
3764 (2) There is a new type, "mxh", which is like "mx" except that it returns
3765 just the host names, not the priorities.
3767 (3) It is now possible to give a list of domains (or IP addresses) to be
3768 looked up. The behaviour when one of the lookups defers can be
3769 controlled by a keyword.
3771 (4) It is now possible to specify the separator character for use when
3772 multiple records are returned.
3774 38. The dnslists ACL condition has been extended: it is now possible to supply
3775 a list of IP addresses and/or domains to be looked up in a particular DNS
3778 39. Added log_selector=+queue_time_overall.
3780 40. When running the queue in the test harness, wait just a tad after forking a
3781 delivery process, to get repeatability of debugging output.
3783 41. Include certificate and key file names in error message when GnuTLS fails
3784 to set them up, because the GnuTLS error message doesn't include the name
3785 of the failing file when there is a problem reading it.
3787 42. Allow both -bf and -bF in the same test run.
3789 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3791 44. The "Exiscan patch" is now merged into the mainline Exim source.
3793 45. Sometimes the final signoff response after QUIT could fail to get
3794 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3795 < 0 before doing a fflush(). This bug looks as though it goes back to the
3796 introduction of TLS in release 3.20, but "sometimes" must have been rare
3797 because the tests only now provoked it.
3799 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3800 (this can affect the format of dates).
3802 47. exim_tidydb, when checking for the continued existence of a message for
3803 which it has found a message-specific retry record, was not finding
3804 messages that were in split spool directories. Consequently, it was
3805 deleting retry records that should have stayed in existence.
3807 48. Steve fixed some bugs in eximstats.
3809 49. The SPA authentication driver was not abandoning authentication and moving
3810 on to the next authenticator when an expansion was forced to fail,
3811 contradicting the general specification for all authenticators. Instead it
3812 was generating a temporary error. It now behaves as specified.
3814 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3815 (the order specifies the preference for clients). The order is now AES256,
3816 AES128, 3DES, ARCFOUR128.
3818 51. Small patch to Sieve code - explicitly set From: when generating an
3821 52. Exim crashed if a remote delivery caused a very long error message to be
3822 recorded - for instance if somebody sent an entire SpamAssassin report back
3823 as a large number of 550 error lines. This bug was coincidentally fixed by
3824 increasing the size of one of Exim's internal buffers (big_buffer) that
3825 happened as part of the Exiscan merge. However, to be on the safe side, I
3826 have made the code more robust (and fixed the comments that describe what
3829 53. Now that there can be additional text after "Completed" in log lines (if
3830 the queue_time_overall log selector is set), a one-byte patch to exigrep
3831 was needed to allow it to recognize "Completed" as not the last thing in
3834 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3835 patch that reportedly fixes this has been added. I am not expert enough to
3836 create a test for it. This is what the patch creator wrote:
3838 "I found a little strange behaviour of ldap code when working with
3839 Windows 2003 AD Domain, where users was placed in more than one
3840 Organization Units. When I tried to give exim partial DN, the exit code
3841 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3842 But simultaneously result of request was absolutely normal ldap result,
3843 so I produce this patch..."
3845 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3846 REFERENCE, so I have modified the code to exclude the patch when that macro
3849 55. Some experimental protocols are using DNS PTR records for new purposes. The
3850 keys for these records are domain names, not reversed IP addresses. The
3851 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3852 leaves it alone. Component reversal etc. now happens only for IP addresses.
3855 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3857 57. Double the size of the debug message buffer (to 2048) so that more of very
3858 long debug lines gets shown.
3860 58. The exicyclog utility now does better if the number of log files to keep
3861 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3863 59. Two changes related to the smtp_active_hostname option:
3865 (1) $smtp_active_hostname is now available as a variable.
3866 (2) The default for smtp_banner uses $smtp_active_hostname instead
3867 of $primary_hostname.
3869 60. The host_aton() function is supposed to be passed a string that is known
3870 to be a valid IP address. However, in the case of IPv6 addresses, it was
3871 not checking this. This is a hostage to fortune. Exim now panics and dies
3872 if the condition is not met. A case was found where this could be provoked
3873 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3874 components; fortuitously, this particular loophole had already been fixed
3875 by change 4.50/55 above.
3877 If there are any other similar loopholes, the new check in host_aton()
3878 itself should stop them being exploited. The report I received stated that
3879 data on the command line could provoke the exploit when Exim was running as
3880 exim, but did not say which command line option was involved. All I could
3881 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3882 running as the user.
3885 61. There was a buffer overflow vulnerability in the SPA authentication code
3886 (which came originally from the Samba project). I have added a test to the
3887 spa_base64_to_bits() function which I hope fixes it.
3890 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3891 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3893 63. The daemon start-up calls getloadavg() while still root for those OS that
3894 need the first call to be done as root, but it missed one case: when
3895 deliver_queue_load_max is set with deliver_drop_privilege. This is
3896 necessary for the benefit of the queue runner, because there is no re-exec
3897 when deliver_drop_privilege is set.
3899 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3900 This has been fixed.
3902 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3903 was in use, was not putting the data itself into the right store pool;
3904 consequently, it could be overwritten for a subsequent message in the same
3905 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3908 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3910 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3911 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3912 the function were treating the return as a boolean value, which happened to
3913 work because 0=false and not-0=true, but is not correct code.
3915 68. The host_aton() function was not handling scoped IPv6 addresses (those
3916 with, for example, "%eth0" on the end) correctly.
3918 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3919 negated items (that is, ~something) in unsigned ints. Some compilers
3920 apparently mutter when there is no cast.
3922 70. If an address verification called from an ACL failed, and did not produce a
3923 user-specific message (i.e. there was only a "system" message), nothing was
3924 put in $acl_verify_message. In this situation, it now puts the system
3927 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3928 session; change 4.31/43 added the unwanted input to the log line - except
3929 that it did not do this in the start of session case. It now does.
3931 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3932 This is wrong and can cause the other end to generate a synchronization
3933 error if it is another Exim or anything else that does the synchronization
3934 check. A QUIT command is no longer sent after a timeout.
3936 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3937 during host lookups.
3939 74. The defer_ok option of callout verification was not working if it was used
3940 when verifying addresses in header lines, that is, for this case:
3942 verify = header_sender/callout=defer_ok
3944 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3945 those file descriptors could be used for SMTP connections. If anything
3946 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3947 could be sent to the SMTP client, causing chaos. The daemon now opens
3948 stdin, stdout, and stderr to /dev/null when it puts itself into the
3951 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3952 log by default. The user can override this with suitable Perl magic.
3954 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3955 the log message when discard triggers, was not working for the DATA ACL or
3956 for the non-SMTP ACL.
3958 78. Error message wording change in sieve.c.
3960 79. If smtp_accept_max_per_host was set, the number of connections could be
3961 restricted to fewer than expected, because the daemon was trying to set up
3962 a new connection before checking whether the processes handling previous
3963 connections had finished. The check for completed processes is now done
3964 earlier. On busy systems, this bug wouldn't be noticed because something
3965 else would have woken the daemon, and it would have reaped the completed
3968 80. If a message was submitted locally by a user whose login name contained one
3969 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3970 It caused a spool format error. I have fixed the spool reading code. A
3971 related problem was that the "from" clause in the Received: line became
3972 illegal because of the space(s). It is now covered by ${quote_local_part.
3974 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3977 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3979 83. Patch from Sieve maintainer to fix unterminated string problem in
3980 "vacation" handling.
3982 84. Some minor changes to the Linux configuration files to help with other
3983 OS variants using glibc.
3985 85. One more patch for Sieve to update vacation handling to latest spec.
3988 ----------------------------------------------------
3989 See the note above about the 4.44 and 4.50 releases.
3990 ----------------------------------------------------
3996 1. Change 4.43/35 introduced a bug that caused file counts to be
3997 incorrectly computed when quota_filecount was set in an appendfile
4000 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4001 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4004 3. Give more explanation in the error message when the command for a transport
4005 filter fails to execute.
4007 4. There are several places where Exim runs a non-Exim command in a
4008 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4009 was being done only for the command run by the queryprogram router. It is
4010 now done for all such subprocesses. The other cases are: ${run, transport
4011 filters, and the commands run by the lmtp and pipe transports.
4013 5. Some older OS have a limit of 256 on the maximum number of file
4014 descriptors. Exim was using setrlimit() to set 1000 as a large value
4015 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4016 systems. I've change it so that if it can't get 1000, it tries for 256.
4018 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4019 was an oversight, and furthermore, ever since the addition of extra
4020 controls (e.g. 4.43/32), the checks on when to allow different forms of
4021 "control" were broken. There should now be diagnostics for all cases when a
4022 control that does not make sense is encountered.
4024 7. $recipients is now available in the predata ACL (oversight).
4026 8. Tidy the search cache before the fork to do a delivery from a message
4027 received from the command line. Otherwise the child will trigger a lookup
4028 failure and thereby defer the delivery if it tries to use (for example) a
4029 cached ldap connection that the parent has called unbind on.
4031 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4032 of $address_data from the recipient verification was clobbered by the
4033 sender verification.
4035 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4036 was its contents. (It was OK if the option was not defined at all.)
4038 11. A "Completed" log line is now written for messages that are removed from
4039 the spool by the -Mrm option.
4041 12. $host_address is now set to the target address during the checking of
4042 ignore_target_hosts.
4044 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4045 being passed; this would have caused $sender_host_name to have been used if
4046 matching the list had actually called for a host name (not very likely,
4047 since this list is usually IP addresses). A host name is now passed as
4050 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4051 code that specifies a non-restarting handler (typically sigaction() in
4052 modern systems) in an attempt to fix a rare and obscure crash bug.
4054 15. Narrowed the window for a race in the daemon that could cause it to ignore
4055 SIGCHLD signals. This is not a major problem, because they are used only to
4056 wake it up if nothing else does.
4058 16. A malformed maildirsize file could cause Exim to calculate negative values
4059 for the mailbox size or file count. Odd effects could occur as a result.
4060 The maildirsize information is now recalculated if the size or filecount
4063 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4064 support for a long time. Removed HAVE_SYS_VFS_H.
4066 18. Updated exipick to current release from John Jetmore.
4068 19. Allow an empty sender to be matched against a lookup in an address list.
4069 Previously the only cases considered were a regular expression, or an
4072 20. Exim went into a mad DNS lookup loop when doing a callout where the
4073 host was specified on the transport, if the DNS lookup yielded more than
4076 21. The RFC2047 encoding function was originally intended for short strings
4077 such as real names; it was not keeping to the 75-character limit for
4078 encoded words that the RFC imposes. It now respects the limit, and
4079 generates multiple encoded words if necessary. To be on the safe side, I
4080 have increased the buffer size for the ${rfc2047: expansion operator from
4083 22. Failure to deliver a bounce message always caused it to be frozen, even if
4084 there was an errors_to setting on the router. The errors_to setting is now
4087 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4088 canonical form (fully expanded) before being placed in
4089 $sender_host_address.
4091 24. Updated eximstats to version 1.33
4093 25. Include certificate and key file names in error message when GnuTLS fails
4094 to set them up, because the GnuTLS error message doesn't include the name
4095 of the failing file when there is a problem reading it.
4097 26. Expand error message when OpenSSL has problems setting up cert/key files.
4100 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4101 (this can affect the format of dates).
4103 28. exim_tidydb, when checking for the continued existence of a message for
4104 which it has found a message-specific retry record, was not finding
4105 messages that were in split spool directories. Consequently, it was
4106 deleting retry records that should have stayed in existence.
4108 29. eximstats updated to version 1.35
4109 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4110 1.35 - bugfix such that pie charts by volume are generated correctly
4112 30. The SPA authentication driver was not abandoning authentication and moving
4113 on to the next authenticator when an expansion was forced to fail,
4114 contradicting the general specification for all authenticators. Instead it
4115 was generating a temporary error. It now behaves as specified.
4117 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4118 (the order specifies the preference for clients). The order is now AES256,
4119 AES128, 3DES, ARCFOUR128.
4121 31. Small patch to Sieve code - explicitly set From: when generating an
4124 32. Exim crashed if a remote delivery caused a very long error message to be
4125 recorded - for instance if somebody sent an entire SpamAssassin report back
4126 as a large number of 550 error lines. This bug was coincidentally fixed by
4127 increasing the size of one of Exim's internal buffers (big_buffer) that
4128 happened as part of the Exiscan merge. However, to be on the safe side, I
4129 have made the code more robust (and fixed the comments that describe what
4132 33. Some experimental protocols are using DNS PTR records for new purposes. The
4133 keys for these records are domain names, not reversed IP addresses. The
4134 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4135 leaves it alone. Component reversal etc. now happens only for IP addresses.
4138 34. The host_aton() function is supposed to be passed a string that is known
4139 to be a valid IP address. However, in the case of IPv6 addresses, it was
4140 not checking this. This is a hostage to fortune. Exim now panics and dies
4141 if the condition is not met. A case was found where this could be provoked
4142 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4143 components; fortuitously, this particular loophole had already been fixed
4144 by change 4.50/55 or 4.44/33 above.
4146 If there are any other similar loopholes, the new check in host_aton()
4147 itself should stop them being exploited. The report I received stated that
4148 data on the command line could provoke the exploit when Exim was running as
4149 exim, but did not say which command line option was involved. All I could
4150 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4151 running as the user.
4154 35. There was a buffer overflow vulnerability in the SPA authentication code
4155 (which came originally from the Samba project). I have added a test to the
4156 spa_base64_to_bits() function which I hope fixes it.
4159 36. The daemon start-up calls getloadavg() while still root for those OS that
4160 need the first call to be done as root, but it missed one case: when
4161 deliver_queue_load_max is set with deliver_drop_privilege. This is
4162 necessary for the benefit of the queue runner, because there is no re-exec
4163 when deliver_drop_privilege is set.
4165 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4166 was in use, was not putting the data itself into the right store pool;
4167 consequently, it could be overwritten for a subsequent message in the same
4168 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4171 38. Sometimes the final signoff response after QUIT could fail to get
4172 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4173 < 0 before doing a fflush(). This bug looks as though it goes back to the
4174 introduction of TLS in release 3.20, but "sometimes" must have been rare
4175 because the tests only now provoked it.
4181 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4182 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4183 Now it returns an integer. A number of calls were still expecting a T/F
4184 return. Fortuitously, in all cases, the tests worked in OK situations,
4185 which is the norm. However, things would have gone wrong on any write
4186 failures on the smtp file descriptor. This function is used when sending
4187 messages over SMTP and also when doing verify callouts.
4189 2. When Exim is called to do synchronous delivery of a locally submitted
4190 message (the -odf or -odi options), it no longer closes stderr before doing
4193 3. Implemented the mua_wrapper option.
4195 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4197 5. Implemented the functions header_remove(), header_testname(),
4198 header_add_at_position(), and receive_remove_recipient(), and exported them
4201 6. If an ACL "warn" statement specified the addition of headers, Exim already
4202 inserted X-ACL-Warn: at the start if there was no header name. However, it
4203 was not making this test for the second and subsequent header lines if
4204 there were newlines in the string. This meant that an invalid header could
4205 be inserted if Exim was badly configured.
4207 7. Allow an ACL "warn" statement to add header lines at the start or after all
4208 the Received: headers, as well as at the end.
4210 8. Added the rcpt_4xx retry error code.
4212 9. Added postmaster_mailfrom=xxx to callout verification option.
4214 10. Added mailfrom=xxxx to the callout verification option, for verify=
4217 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4218 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4220 12. Inserted some casts to stop certain compilers warning when using pointer
4221 differences as field lengths or precisions in printf-type calls (mostly
4222 affecting debugging statements).
4224 13. Added optional readline() support for -be (dynamically loaded).
4226 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4227 same clock tick as a message's arrival, so that its received time was the
4228 same as the "first fail" time on the retry record, and that message
4229 remained on the queue past the ultimate address timeout, every queue runner
4230 would try a delivery (because it was past the ultimate address timeout) but
4231 after another failure, the ultimate address timeout, which should have then
4232 bounced the address, did not kick in. This was a "< instead of <=" error;
4233 in most cases the first failure would have been in the next clock tick
4234 after the received time, and all would be well.
4236 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4237 being recognized when the domain list was tested by the match_domain
4238 condition in an expansion string.
4240 16. Added the ${str2b64: operator.
4242 17. Exim was always calling setrlimit() to set a large limit for the number of
4243 processes, without checking whether the existing limit was already
4244 adequate. (It did check for the limit on file descriptors.) Furthermore,
4245 errors from getrlimit() and setrlimit() were being ignored. Now they are
4246 logged to the main and panic logs, but Exim does carry on, to try to do its
4247 job under whatever limits there are.
4249 18. Imported PCRE 5.0.
4251 19. Trivial typo in log message " temporarily refused connection" (the leading
4254 20. If the log selector return_path_on_delivery was set and an address was
4255 redirected to /dev/null, the delivery process crashed because it assumed
4256 that a return path would always be set for a "successful" delivery. In this
4257 case, the whole delivery is bypassed as an optimization, and therefore no
4260 21. Internal re-arrangement: the function for sending a challenge and reading
4261 a response while authentication was assuming a zero-terminated challenge
4262 string. It's now changed to take a pointer and a length, to allow for
4263 binary data in such strings.
4265 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4267 23. Exim was not respecting finduser_retries when seeking the login of the
4268 uid under which it was called; it was always trying 10 times. (The default
4269 setting of finduser_retries is zero.) Also, it was sleeping after the final
4270 failure, which is pointless.
4272 24. Implemented tls_on_connect_ports.
4274 25. Implemented acl_smtp_predata.
4276 26. If the domain in control=submission is set empty, Exim assumes that the
4277 authenticated id is a complete email address when it generates From: or
4278 Sender: header lines.
4280 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4281 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4282 chown and chgrp in /bin and hostname in /usr/bin.
4284 28. Exim was keeping the "process log" file open after each use, just as it
4285 does for the main log. This opens the possibility of it remaining open for
4286 long periods when the USR1 signal hits a daemon. Occasional processlog
4287 errors were reported, that could have been caused by this. Anyway, it seems
4288 much more sensible not to leave this file open at all, so that is what now
4291 29. The long-running daemon process does not normally write to the log once it
4292 has entered its main loop, and it closes the log before doing so. This is
4293 so that log files can straightforwardly be renamed and moved. However,
4294 there are a couple of unusual error situations where the daemon does write
4295 log entries, and I had neglected to close the log afterwards.
4297 30. The text of an SMTP error response that was received during a remote
4298 delivery was being truncated at 512 bytes. This is too short for some of
4299 the long messages that one sometimes sees. I've increased the limit to
4302 31. It is now possible to make retry rules that apply only when a message has a
4303 specific sender, in particular, an empty sender.
4305 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4306 it possible to be selective about when SMTP synchronization is enforced.
4308 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4310 32. Implemented hosts_connection_nolog.
4312 33. Added an ACL for QUIT.
4314 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4317 35. Added mailbox_size and mailbox_filecount to appendfile.
4319 36. Added control = no_multiline_responses to ACLs.
4321 37. There was a bug in the logic of the code that waits for the clock to tick
4322 in the case where the clock went backwards by a substantial amount such
4323 that the microsecond fraction of "now" was more than the microsecond
4324 fraction of "then" (but the whole seconds number was less).
4326 38. Added support for the libradius Radius client library this is found on
4327 FreeBSD (previously only the radiusclient library was supported).
4333 1. When certain lookups returned multiple values in the form name=value, the
4334 quoting of the values was not always being done properly. Specifically:
4335 (a) If the value started with a double quote, but contained no whitespace,
4337 (b) If the value contained whitespace other than a space character (i.e.
4338 tabs or newlines or carriage returns) it was not quoted.
4339 This fix has been applied to the mysql and pgsql lookups by writing a
4340 separate quoting function and calling it from the lookup code. The fix
4341 should probably also be applied to nisplus, ibase and oracle lookups, but
4342 since I cannot test any of those, I have not disturbed their existing code.
4344 2. A hit in the callout cache for a specific address caused a log line with no
4345 reason for rejecting RCPT. Now it says "Previous (cached) callout
4346 verification failure".
4348 3. There was an off-by-one bug in the queryprogram router. An over-long
4349 return line was truncated at 256 instead of 255 characters, thereby
4350 overflowing its buffer with the terminating zero. As well as fixing this, I
4351 have increased the buffer size to 1024 (and made a note to document this).
4353 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4354 when Exim is waiting for an SMTP response from a remote server, Exim
4355 restarts its select() call on the socket, thereby resetting its timeout.
4356 This is not a problem when such interrupts are rare. Somebody set up a cron
4357 job to run exiwhat every 2 minutes, which is less than the normal select()
4358 timeout (5 or 10 minutes). This meant that the select() timeout never
4359 kicked in because it was always reset. I have fixed this by comparing the
4360 time when an interrupt arrives with the time at the start of the first call
4361 to select(). If more time than the timeout has elapsed, the interrupt is
4362 treated as a timeout.
4364 5. Some internal re-factoring in preparation for the addition of Sieve
4365 extensions (by MH). In particular, the "personal" test is moved to a
4366 separate function, and given an option for scanning Cc: and Bcc: (which is
4367 not set for Exim filters).
4369 6. When Exim created an email address using the login of the caller as the
4370 local part (e.g. when creating a From: or Sender: header line), it was not
4371 quoting the local part when it contained special characters such as @.
4373 7. Installed new OpenBSD configuration files.
4375 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4376 try to make them clearer.
4378 9. Callout options, other than the timeout value, were being ignored when
4379 verifying sender addresses in header lines. For example, when using
4381 verify = header_sender/callout=no_cache
4383 the cache was (incorrectly) being used.
4385 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4386 only the Cygwin environment.
4388 11. When return_path_on_delivery was set as a log selector, if different remote
4389 addresses in the same message used different return paths and parallel
4390 remote delivery occurred, the wrong values would sometimes be logged.
4391 (Whenever a remote delivery process finished, the return path value from
4392 the most recently started remote delivery process was logged.)
4394 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4395 lines when AUTH and/or TLS are in use. This is the "received protocol"
4396 field. Exim used to use "asmtp" for authenticated SMTP, without any
4397 indication (in the protocol name) for TLS use. Now it follows the RFC and
4398 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4399 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4400 names appear in log lines as well as in Received: header lines.
4402 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4403 extensions, and comparison tests, and to fix some bugs.
4405 14. Changes to the "personal" filter test:
4407 (1) The test was buggy in that it was just doing the equivalent of
4408 "contains" tests on header lines. For example, if a user's address was
4409 anne@some.where, the "personal" test would incorrectly be true for
4411 To: susanne@some.where
4413 This test is now done by extracting each address from the header in turn,
4414 and checking the entire address. Other tests that are part of "personal"
4415 are now done using regular expressions (for example, to check local parts
4416 of addresses in From: header lines).
4418 (2) The list of non-personal local parts in From: addresses has been
4419 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4420 taken from the Sieve specification recommendations.
4422 (3) If the message contains any header line starting with "List-" it is
4423 treated as non-personal.
4425 (4) The test for "circular" in the Subject: header line has been removed
4426 because it now seems ill-conceived.
4428 15. Minor typos in src/EDITME comments corrected.
4430 16. Installed latest exipick from John Jetmore.
4432 17. If headers_add on a router specified a text string that was too long for
4433 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4434 of string_sprintf() is now avoided.
4436 18. $message_body_size was not set (it was always zero) when running the DATA
4437 ACL and the local_scan() function.
4439 19. For the "mail" command in an Exim filter, no default was being set for
4440 the once_repeat time, causing a random time value to be used if "once" was
4441 specified. (If the value happened to be <= 0, no repeat happened.) The
4442 default is now 0s, meaning "never repeat". The "vacation" command was OK
4443 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4444 (I found it when inspecting the code).
4446 20. There is now an overall timeout for performing a callout verification. It
4447 defaults to 4 times the callout timeout, which applies to individual SMTP
4448 commands during the callout. The overall timeout applies when there is more
4449 than one host that can be tried. The timeout is checked before trying the
4450 next host. This prevents very long delays if there are a large number of
4451 hosts and all are timing out (e.g. when the network connections are timing
4452 out). The value of the overall timeout can be changed by specifying an
4453 additional sub-option for "callout", called "maxwait". For example:
4455 verify = sender/callout=5s,maxwait=20s
4457 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4458 to the end before writing, but this should make it even safer).
4460 22. Exim was forgetting that it had advertised PIPELINING for the second and
4461 subsequent messages on an SMTP connection. It was also not resetting its
4462 memory on STARTTLS and an internal HELO.
4464 23. When Exim logs an SMTP synchronization error within a session, it now
4465 records whether PIPELINING has been advertised or not.
4467 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4468 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4469 rather than long int.
4471 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4473 26. Added the never_mail option to autoreply.
4479 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4480 crash if the getsockname() call failed; this can happen if a connection is
4481 closed very soon after it is established. The problem was simply in the
4482 order in which certain operations were done, causing Exim to try to write
4483 to the SMTP stream before it had set up the file descriptor. The bug has
4484 been fixed by making things happen in the correct order.
4490 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4491 before the connection was closed, thus losing the rejection response.
4493 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4494 some early Solaris releases, but causes trouble in current releases where
4495 socklen_t is defined.
4497 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4500 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4503 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4504 that releasing the top of it at the end releases what was used for sub-
4505 expansions (unless the block got too big). However, discard this block if
4506 the first thing is a variable or header, so that we can use its block when
4507 it is dynamic (useful for very large $message_headers, for example).
4509 6. Lookups now cache *every* query, not just the most recent. A new, separate
4510 store pool is used for this. It can be recovered when all lookup caches are
4511 flushed. Lookups now release memory at the end of their result strings.
4512 This has involved some general refactoring of the lookup sources.
4514 7. Some code has been added to the store_xxx() functions to reduce the amount
4515 of flapping under certain conditions.
4517 8. log_incoming_interface used to affect only the <= reception log lines. Now
4518 it causes the local interface and port to be added to several more SMTP log
4519 lines, for example "SMTP connection from", and rejection lines.
4521 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4523 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4525 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4526 could be overwritten at the end of the current message (or the start of a
4527 new message if it was set in a HELO ACL). The value is now preserved for
4528 the duration of the SMTP connection.
4530 12. If a transport had a headers_rewrite setting, and a matching header line
4531 contained an unqualified address, that address was qualified, even if it
4532 did not match any rewriting rules. The underlying bug was that the values
4533 of the flags that permit the existence of unqualified sender and recipient
4534 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4535 non-local messages, and by -bnq for local messages) were not being
4536 preserved with the message after it was received.
4538 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4539 "next input=" as part of the text comprising the host identity instead of
4540 the correct text. The code was using the same buffer for two different
4541 strings. However, depending on which order the printing function evaluated
4542 its arguments, the bug did not always show up. Under Linux, for example, my
4543 test suite worked just fine.
4545 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4546 doesn't work with some older versions of Perl. It has been changed to "my",
4547 which in any case is probably the better facility to use.
4549 15. A really picky compiler found some instances of statements for creating
4550 error messages that either had too many or two few arguments for the format
4553 16. The size of the buffer for calls to the DNS resolver has been increased
4554 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4555 for addresses that have a lot of PTR records. This alleviates a problem; it
4556 does not fully solve it.
4558 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4559 buffer now truncates the list and logs the incident, which is the same
4560 action as happens when Exim is looking up a host name and its aliases.
4561 Previously in this situation something unpredictable would happen;
4562 sometimes it was "internal error: store_reset failed".
4564 18. If a server dropped the connection unexpectedly when an Exim client was
4565 using GnuTLS and trying to read a response, the client delivery process
4566 crashed while trying to generate an error log message.
4568 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4569 string, for example:
4571 warn message = H1: something\nH2: something
4573 the text was added as a single header line from Exim's point of view
4574 though it ended up OK in the delivered message. However, searching for the
4575 second and subsequent header lines using $h_h2: did not work. This has been
4576 fixed. Similarly, if a system filter added multiple headers in this way,
4577 the routers could not see them.
4579 20. Expanded the error message when iplsearch is called with an invalid key to
4580 suggest using net-iplsearch in a host list.
4582 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4583 ACLs are no longer actually imposed (and a message to that effect is
4586 22. If a "gecos" field in a passwd entry contained escaped characters, in
4587 particular, if it contained a \" sequence, Exim got it wrong when building
4588 a From: or a Sender: header from that name. A second bug also caused
4589 incorrect handling when an unquoted " was present following a character
4590 that needed quoting.
4592 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4593 was not being matched caselessly.
4595 24. Arranged for all hyphens in the exim.8 source to be escaped with
4598 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4599 properly. Recipient callout cache records were still being keyed to include
4600 the sender, even when use_sender was set false. This led to far more
4601 callouts that were necessary. The sender is no longer included in the key
4602 when use_sender is false.
4604 26. Added "control = submission" modifier to ACLs.
4606 27. Added the ${base62d: operator to decode base 62 numbers.
4608 28. dnsdb lookups can now access SRV records.
4610 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4611 the configuration file.
4613 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4614 (-v) mode. This makes the output for a verbose queue run more intelligible.
4616 31. Added a use_postmaster feature to recipient callouts.
4618 32. Added the $body_zerocount variable, containing the number of binary zero
4619 bytes in the message body.
4621 33. The time of last modification of the "new" subdirectory is now used as the
4622 "mailbox time last read" when there is a quota error for a maildir
4625 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4627 35. Added +ignore_unknown as a special item in host lists.
4629 36. Code for decoding IPv6 addresses in host lists is now included, even if
4630 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4631 address was recognized as an IP address, but was then not correctly decoded
4632 into binary, causing unexpected and incorrect effects when compared with
4639 1. Very minor rewording of debugging text in manualroute to say "list of
4640 hosts" instead of "hostlist".
4642 2. If verify=header_syntax was set, and a header line with an unqualified
4643 address (no domain) and a large number of spaces between the end of the
4644 name and the colon was received, the reception process suffered a buffer
4645 overflow, and (when I tested it) crashed. This was caused by some obsolete
4646 code that should have been removed. The fix is to remove it!
4648 3. When running in the test harness, delay a bit after writing a bounce
4649 message to get a bit more predictability in the log output.
4651 4. Added a call to search_tidyup() just before forking a reception process. In
4652 theory, someone could use a lookup in the expansion of smtp_accept_max_
4653 per_host which, without the tidyup, could leave open a database connection.
4655 5. Added the variables $recipient_data and $sender_data which get set from a
4656 lookup success in an ACL "recipients" or "senders" condition, or a router
4657 "senders" option, similar to $domain_data and $local_part_data.
4659 6. Moved the writing of debug_print from before to after the "senders" test
4662 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4663 problems for message scanning, either using a data ACL, or using
4664 local_scan() because the Received: header was not generated till after they
4665 were called (in order to set the time as the time of reception completion).
4666 I have revised the way this works. The header is now generated after the
4667 body is received, but before the ACL or local_scan() are called. After they
4668 are run, the timestamp in the header is updated.
4674 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4675 before starting a queue runner without re-exec. This happened only when
4676 deliver_drop_privilege was set or when the Exim user was set to root. The
4677 effect of the bug was that timeouts during subsequent deliveries caused
4678 crashes instead of being properly handled. The handler is now left at its
4679 default (and expected) setting.
4681 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4682 message, again when deliver_drop_privilege is set or Exim is run as root.
4683 The bug described in (1) was not present in this case, but the tidying up
4684 of the other signals was missing. I have made the two cases consistent.
4686 3. The ignore_target_hosts setting on a manualroute router was being ignored
4687 for hosts that were looked up using the /MX notation.
4689 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4692 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4693 operated on the sender address. After changing the $sender_address to <>
4694 for the sender address verify, Exim was re-instated it as the original
4695 (before rewriting) address, but remembering that it had rewritten it, so it
4696 wasn't rewriting it again. This bug also had the effect of breaking the
4697 sender address verification caching when the sender address was rewritten.
4699 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4700 This has been changed so that if the ip literal address matches
4701 ignore_target_hosts, the router declines.
4703 7. Added expansion conditions match_domain, match_address, and match_local_
4704 part (NOT match_host).
4706 8. The placeholder for the Received: header didn't have a length field set.
4708 9. Added code to Exim itself and to exim_lock to test for a specific race
4709 condition that could lead to file corruption when using MBX delivery. The
4710 issue is with the lockfile that is created in /tmp. If this file is removed
4711 after a process has opened it but before that process has acquired a lock,
4712 there is the potential for a second process to recreate the file and also
4713 acquire a lock. This could lead to two Exim processes writing to the file
4714 at the same time. The added code performs the same test as UW imapd; it
4715 checks after acquiring the lock that its file descriptor still refers to
4716 the same named file.
4718 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4719 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4722 11. Added the smtp_active_hostname option. If used, this will typically be made
4723 to depend on the incoming interface address. Because $interface_address is
4724 not set up until the daemon has forked a reception process, error responses
4725 that can happen earlier (such as "too many connections") no longer contain
4728 12. If an expansion in a condition on a "warn" statement fails because a lookup
4729 defers, the "warn" statement is abandoned, and the next ACL statement is
4730 processed. Previously this caused the whole ACL to be aborted.
4732 13. Added the iplsearch lookup type.
4734 14. Added ident_timeout as a log selector.
4736 15. Added tls_certificate_verified as a log selector.
4738 16. Added a global option tls_require_ciphers (compare the smtp transport
4739 option of the same name). This controls incoming TLS connections.
4741 17. I finally figured out how to make tls_require_ciphers do a similar thing
4742 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4743 before starting the TLS session.
4745 18. Tabs are now shown as \t in -bP output.
4747 19. If the log selector return_path_on_delivery was set, Exim crashed when
4748 bouncing a message because it had too many Received: header lines.
4750 20. If two routers both had headers_remove settings, and the first one included
4751 a superfluous trailing colon, the final name in the first list and the
4752 first name in the second list were incorrectly joined into one item (with a
4753 colon in the middle).
4759 1. Added -C and -D options to the exinext utility, mainly to make it easier
4760 to include in the automated testing, but these could be helpful when
4761 multiple configurations are in use.
4763 2. The exinext utility was not formatting the output nicely when there was
4764 an alternate port involved in the retry record key, nor when there was a
4765 message id as well (for retries that were specific to a specific message
4766 and a specific host). It was also confused by IPv6 addresses, because of
4767 the additional colons they contain. I have fixed the IPv4 problem, and
4768 patched it up to do a reasonable job for IPv6.
4770 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4771 delivery, the log line now contains "pipelined" if PIPELINING was used.
4773 4. An SMTP transport process used to panic and die if the bind() call to set
4774 an explicit outgoing interface failed. This has been changed; it is now
4775 treated in the same way as a connect() failure.
4777 5. A reference to $sender_host_name in the part of a conditional expansion
4778 that was being skipped was still causing a DNS lookup. This no longer
4781 6. The def: expansion condition was not recognizing references to header lines
4782 that used bh_ and bheader_.
4784 7. Added the _cache feature to named lists.
4786 8. The code for checking quota_filecount in the appendfile transport was
4787 allowing one more file than it should have been.
4789 9. For compatibility with Sendmail, the command line option
4797 and sets the incoming protocol and host name (for trusted callers). The
4798 host name and its colon can be omitted when only the protocol is to be set.
4799 Note the Exim already has two private options, -pd and -ps, that refer to
4800 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4801 "s", but I don't think that's a major issue.
4803 10. A number of refactoring changes to the code, none of which should affect
4806 (a) The number of logging options was getting close to filling up the
4807 32-bit word that was used as a bit map. I have split them into two classes:
4808 those that are passed in the argument to log_write(), and those that are
4809 only ever tested independently outside of that function. These are now in
4810 separate 32-bit words, so there is plenty of room for expansion again.
4811 There is no change in the user interface or the logging behaviour.
4813 (b) When building, for example, log lines, the code previously used a
4814 macro that called string_cat() twice, in order to add two strings. This is
4815 not really sufficiently general. Furthermore, there was one instance where
4816 it was actually wrong because one of the argument was used twice, and in
4817 one call a function was used. (As it happened, calling the function twice
4818 did not affect the overall behaviour.) The macro has been replaced by a
4819 function that can join an arbitrary number of extra strings onto a growing
4822 (c) The code for expansion conditions now uses a table and a binary chop
4823 instead of a serial search (which was left over from when there were very
4824 few conditions). Also, it now recognizes conditions like "pam" even when
4825 the relevant support is not compiled in: a suitably worded error message is
4826 given if an attempt is made to use such a condition.
4828 11. Added ${time_interval:xxxxx}.
4830 12. A bug was causing one of the ddress fields not to be passed back correctly
4831 from remote delivery subprocesses. The field in question was not being
4832 subsequently used, so this caused to problems in practice.
4834 13. Added new log selectors queue_time and deliver_time.
4836 14. Might have fixed a bug in maildirsizefile handling that threw up
4837 "unexpected character" debug warnings, and recalculated the data
4838 unnecessarily. In any case, I expanded the warning message to give more
4841 15. Added the message "Restricted characters in address" to the statements in
4842 the default ACL that block characters like @ and % in local parts.
4844 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4845 Three changes have been made:
4847 (a) There was a serious bug; a negative response to MAIL caused the whole
4848 recipient domain to be cached as invalid, thereby blocking all messages
4849 to all local parts at the same domain, from all senders. This bug has
4850 been fixed. The domain is no longer cached after a negative response to
4851 MAIL if the sender used is not empty.
4853 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4856 (c) A new callout option, "use_sender" has been added for people who want
4857 the modified behaviour.
4863 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4866 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4867 indeed breaks things for older releases.
4869 3. Added additional logging to the case where there is a problem reading data
4870 from a filter that is running in a subprocess using a pipe, in order to
4871 try to track down a specific problem.
4873 4. Testing facility fudge: when running in the test harness and attempting
4874 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4875 getting "No route to host". Convert this to a timeout.
4877 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4880 6. Some OS don't have socklen_t but use size_t instead. This affects the
4881 fifth argument of getsockopt() amongst other things. This is now
4882 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4883 can be set for individual OS. I have set it for SunOS5, OSF1, and
4884 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4885 some earlier ones do not.
4887 7. Change 4.30/15 was not doing the test caselessly.
4889 8. The standard form for an IPv6 address literal was being rejected by address
4890 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4891 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4892 this, as well as the form without the "IPv6" on the front (but only when
4893 address literals are enabled, of course).
4895 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4897 10. Exim crashed if a message with an empty sender address specified by -f
4898 encountered a router with an errors_to setting. This could be provoked only
4899 by a command such as
4903 where an empty string was supplied; "<>" did not hit this bug.
4905 11. Installed PCRE release 4.5.
4907 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4908 remained set. It is now erased.
4910 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4911 times from message ids (which are base 36 rather than the normal 62).
4913 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4914 were being counted as actual protocol errors, and logged if the log
4915 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4916 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4917 and DATA following a set of rejected RCPTs do not count as protocol errors.
4918 In other words, Exim assumes they were pipelined, though this may not
4919 actually be the case. Of course, in all cases the client gets an
4920 appropriate error code.
4922 15. If a lookup fails in an ACL condition, a message about the failure may
4923 be available; it is used if testing the ACL cannot continue, because most
4924 such messages specify what the cause of the deferral is. However, some
4925 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4926 that caused an old message to be retained and used if a later statement
4927 caused a defer, replacing the real cause of the deferral.
4929 16. If an IP address had so many PTR records that the DNS lookup buffer
4930 was not large enough to hold them, Exim could crash while trying to process
4931 the truncated data. It now detects and logs this case.
4933 17. Further to 4.21/58, another change has been made: if (and only if) the
4934 first line of a message (the first header line) ends with CRLF, a bare LF
4935 in a subsequent header line has a space inserted after it, so as not to
4936 terminate the header.
4938 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4939 unnecessarily, used atoi() instead of strtol(), and didn't check the
4940 termination when getting file sizes from file names by regex.
4942 19. Completely re-implemented the support for maildirsize files, in the light
4943 of a number of problems with the previous contributed implementation
4944 (4.30/29). In particular:
4946 . If the quota is zero, the maildirsize file is maintained, but no quota is
4949 . If the maildir directory does not exist, it is created before any attempt
4950 to write a maildirsize file.
4952 . The quota value in the file is just a cache; if the quota is changed in
4953 the transport, the new value overrides.
4955 . A regular expression is available for excluding directories from the
4958 20. The autoreply transport checks the characters in options that define the
4959 message's headers; it allows continued headers, but it was checking with
4960 isspace() after an embedded newline instead of explicitly looking for a
4963 21. If all the "regular" hosts to which an address was routed had passed their
4964 expiry times, and had not reached their retry times, the address was
4965 bounced, even if fallback hosts were defined. Now Exim should go on to try
4968 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4969 equivalent code in the SMTP transport. Some hosts send humungous responses
4970 to HELO/EHLO, more than 1024 it seems.
4972 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4973 gives compiler warnings in some environments. I've now done it "properly",
4976 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4977 (because of problems with the built-in one) was declared to return uschar *
4978 instead of char *, causing compiler failure.
4980 25. Fixed a file descriptor leak when processing alias/forward files.
4982 26. Fixed a minor format string issue in dbfn.c.
4984 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4986 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4987 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4990 29. When a sender address is verified, it is cached, to save repeating the test
4991 when there is more than one recipient in a message. However, when the
4992 verification involves a callout, it is possible for different callout
4993 options to be set for different recipients. It is too complicated to keep
4994 track of this in the cache, so now Exim always runs a verification when a
4995 callout is required, relying on the callout cache for the optimization.
4996 The overhead is duplication of the address routing, but this should not be
4999 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5000 to be verified with callout=postmaster, and the main callout worked but the
5001 postmaster check failed, the verification correctly failed. However, if a
5002 subsequent RCPT command asked for sender verification *without* the
5003 postmaster check, incorrect caching caused this verification also to fail,
5006 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5007 it was not caching the DNS options (qualify_single, search_parents) that
5008 were used when the lookup failed. A subsequent lookup with different
5009 options therefore always gave the same answer, though there were cases
5010 where it should not have. (Example: a "domains = !$mx_any" option on a
5011 dnslookup router: the "domains" option is always processed without any
5012 widening, but the router might have qualify_single set.) Now Exim uses the
5013 cached value only when the same options are set.
5015 32. Added John Jetmore's "exipick" utility to the distribution.
5017 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5018 than a timeout (e.g. a certificate is required, and is not provided), an
5019 Exim server now closes the connection immediately. Previously it waited for
5020 the client to close - but if the client is SSL, it seems that they each
5021 wait for each other, leading to a delay before one of them times out.
5023 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5024 maintained 0.8.x compatibility because I don't think many are using it, and
5025 it is clearly obsolete.
5027 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5030 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5031 client certificate was expired. A simple patch fixes this, though I don't
5032 understand the full logic of why the verify callback is called multiple
5035 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5036 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5037 which causes problems with some clients (such as the Certicom SSL Plus
5038 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5039 disables the coutermeasure allowing Eudora to connect."
5041 38. Exim was not checking that a write() to a log file succeeded. This could
5042 lead to Bad Things if a log got too big, in particular if it hit a file
5043 size limit. Exim now panics and dies if it cannot write to a log file, just
5044 as it does if it cannot open a log file.
5046 39. Modified OS/Makefile-Linux so that it now contains
5048 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5050 The two -D definitions ensure that Exim is compiled with large file
5051 support, which makes it possible to handle log files that are bigger than
5054 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5055 instance) a domain was checked against a named list that involved a lookup,
5056 causing $domain_data to be set, then another domain was checked against the
5057 same list, then the first domain was re-checked, the value of $domain_data
5058 after the final check could be wrong. In particular, if the second check
5059 failed, it could be set empty. This bug probably also applied to
5062 41. The strip_trailing_dot option was not being applied to the address given
5063 with the -f command-line option.
5065 42. The code for reading a message's header from the spool was incrementing
5066 $received_count, but never initializing it. This meant that the value was
5067 incorrect (doubled) while delivering a message in the same process in which
5068 it was received. In the most common configuration of Exim, this never
5069 happens - a fresh exec is done - but it can happen when
5070 deliver_drop_privilege is set.
5072 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5073 it now includes up to 150 characters of the unexpected data in the log
5076 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5077 and building data strings. The size of both of these buffers was 10 000
5078 bytes - far larger than anybody would *ever* want, thought I. Needless to
5079 say, somebody hit the limit. I have increased the maximum line length to
5080 20 000 and the maximum data length of concatenated lines to 100 000. I have
5081 also fixed two bugs, because there was no checking on these buffers. Tsk,
5082 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5083 buffer is too small.
5085 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5086 lsearch lookups. Now it does.
5088 46. When parsing a route_list item in a manualroute router, a fixed-length
5089 buffer was used for the list of hosts. I made this 1024 bytes long,
5090 thinking that nobody would ever have a list of hosts that long. Wrong.
5091 Somebody had a whole pile of complicated expansion conditions, and the
5092 string was silently truncated, leading to an expansion error. It turns out
5093 that it is easier to change to an unlimited length (owing to other changes
5094 that have happened since this code was originally written) than to build
5095 structure for giving a limitation error. The length of the item that
5096 expands into the list of hosts is now unlimited.
5098 47. The lsearch lookup could not handle data where the length of text line was
5099 more than 4095 characters. Such lines were truncated, leading to shortened
5100 data being returned. It should now handle lines of any length.
5102 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5103 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5106 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5107 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5108 no longer happens in comment lines. A list of replacements is now placed
5109 at the head of all of the source files, except those whose only change is
5110 to replace PERL_COMMAND in the very first #! line.
5112 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5113 messages on the queue, with a bottom-up merge sort, using code contributed
5114 by Michael Haardt. This should make operations like -bp somewhat faster on
5115 large queues. It won't affect queue runners, except when queue_run_in_order
5118 51. Installed eximstats 1.31 in the distribution.
5120 52. Added support for SRV lookups to the dnslookup router.
5122 53. If an ACL referred to $message_body or $message_body_end, the value was not
5123 reset for any messages that followed in the same SMTP session.
5125 54. The store-handling optimization for building very long strings was not
5126 differentiating between the different store pools. I don't think this
5127 actually made any difference in practice, but I've tidied it.
5129 55. While running the routers to verify a sender address, $sender_address
5130 was still set to the sender address. This is wrong, because when routing to
5131 send a bounce to the sender, it would be empty. Therefore, I have changed
5132 it so that, while verifying a sender address, $sender_address is set to <>.
5133 (There is no change to what happens when verifying a recipient address.)
5135 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5136 target A or AAAA records (if not already returned) without resetting the
5137 qualify_single or search_parents options of the DNS resolver. These are
5138 inappropriate in this case because the targets of MX and SRV records must
5139 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5140 target that, when qualified, matched something in the local domain. These
5141 two options are now turned off when doing these lookups.
5143 57. It seems that at least some releases of Reiserfs (which does not have the
5144 concept of a fixed number of inodes) returns zero and not -1 for the
5145 number of available inodes. This interacted badly with check_spool_inodes,
5146 which assumed that -1 was the "no such thing" setting. What I have done is
5147 to check that the total number of inodes is greater than zero before doing
5148 the test of how many are available.
5150 58. When a "warn" ACL statement has a log_message modifier, the message is
5151 remembered, and not repeated. This is to avoid a lot of repetition when a
5152 message has many recipients that cause the same warning to be written.
5153 Howewer, Exim was preserving the list of already written lines for an
5154 entire SMTP session, which doesn't seem right. The memory is now reset if a
5155 new message is started.
5157 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5158 debugging output unless log_rewrite was also set.
5160 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5161 of (int)(handle) when we know that handle contains (void *)(-1).
5163 61. The Exim daemon panic-logs an error return when it closes the incoming
5164 connection. However "connection reset by peer" seems to be common, and
5165 isn't really an error worthy of noting specially, so that particular error
5168 62. When Exim is trying to find all the local interfaces, it used to panic and
5169 die if the ioctl to get the interface flags failed. However, it seems that
5170 on at least one OS (Solaris 9) it is possible to have an interface that is
5171 included in the list of interfaces, but for which you get a failure error
5172 for this call. This happens when the interface is not "plumbed" into a
5173 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5174 failure of the "get flags" call assumes that the interface is down.
5176 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5177 makes life easier for people who are doing arithmetic on fields extracted
5178 from dates, where you often get leading zeros that should not be
5179 interpreted as octal.
5181 64. Added qualify_domain to the redirect router, to override the global
5184 65. If a pathologically long header line contained very many addresses (the
5185 report of this problem mentioned 10 000) and each of them was rewritten,
5186 Exim could use up a very large amount of memory. (It kept on making new
5187 copies of the header line as it rewrote, and never released the old ones.)
5188 At the expense of a bit more processing, the header rewriting function has
5189 been changed so that it no longer eats memory in this way.
5191 66. The generation of the Received: header has been moved from the time that a
5192 message starts to be received, to the time that it finishes. The timestamp
5193 in the Received: header should now be very close to that of the <= log
5194 line. There are two side-effects of this change:
5196 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5197 logged header lines no longer include the local Received: line, because
5198 it has not yet been created. The same applies to a copy of the message
5199 that is returned to a non-SMTP sender when a message is rejected.
5201 (b) When a filter file is tested using -bf, no additional Received: header
5202 is added to the test message. After some thought, I decided that this
5205 This change does not affect the value of $received_for. It is still set
5206 after address rewriting, but before local_scan() is called.
5208 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5210 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5211 gave an unhelpful panic error message, and a defer error. I have managed to
5212 change this behaviour so that it now rejects any supplied certificate,
5213 which seems right, as the list of acceptable certificates is empty.
5215 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5216 gave an unhelpful defer error. I have not managed to make this reject any
5217 supplied certificates, but the error message it gives is "no certificate
5218 supplied", which is not helpful.
5220 70. exigrep's output now also includes lines that are not associated with any
5221 message, but which match the given pattern. Implemented by a patch from
5222 Martin Sluka, which also tidied up the Perl a bit.
5224 71. Recipient callout verification, like sender verification, was using <> in
5225 the MAIL FROM command. This isn't really the right thing, since the actual
5226 sender may affect whether the remote host accepts the recipient or not. I
5227 have changed it to use the actual sender in the callout; this means that
5228 the cache record is now keyed on a recipient/sender pair, not just the
5229 recipient address. There doesn't seem to be a real danger of callout loops,
5230 since a callout by the remote host to check the sender would use <>.
5231 [SEE ABOVE: changed after hitting problems.]
5233 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5234 temporary errors. However, in the case of such a code being given after
5235 the end of a data transmission (i.e. after ".") Exim was failing to write
5236 a retry record for the message. (Yes, there was some broken host that was
5237 actually sending 8xx at this point.)
5239 73. An unknown lookup type in a host list could cause Exim to panic-die when
5240 the list was checked. (An example that provoked this was putting <; in the
5241 middle of a list instead of at the start.) If this happened during a DATA
5242 ACL check, a -D file could be left lying around. This kind of configuration
5243 error no longer causes Exim to die; instead it causes a defer errror. The
5244 incident is still logged to the main and panic logs.
5246 74. Buglet left over from Exim 3 conversion. The message "too many messages
5247 in one connection" was written to the rejectlog but not the mainlog, except
5248 when address rewriting (yes!) was being logged.
5250 75. Added write_rejectlog option.
5252 76. When a system filter was run not as root (that is, when system_filter_user
5253 was set), the values of the $n variables were not being returned to the
5254 main process; thus, they were not subsequently available in the $sn
5257 77. Added +return_path_on_delivery log selector.
5259 78. A connection timeout was being treated differently from recipients deferred
5260 when testing hosts_max_try with a message that was older than the host's
5261 retry timeout. (The host should not be counted, thus allowing all hosts to
5262 be tried at least once before bouncing.) This may have been the cause of an
5263 occasionally reported bug whereby a message would remain on the queue
5264 longer than the retry timeout, but would be bounced if a delivery was
5265 forced. I say "may" because I never totally pinned down the problem;
5266 setting up timeout/retry tests is difficult. See also the next item.
5268 79. The ultimate address timeout was not being applied to errors that involved
5269 a combination of host plus message (for example, a timeout on a MAIL
5270 command). When an address resolved to a number of possible hosts, and they
5271 were not all tried for each delivery (e.g. because of hosts_max_try), a
5272 message could remain on the queue longer than the retry timeout.
5274 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5277 81. Fixed an obscure SMTP outgoing bug which required at least the following
5278 conditions: (a) there was another message waiting for the same server;
5279 (b) the server returned 5xx to all RCPT commands in the first message so
5280 that the message was not completed; (c) the server dropped the connection
5281 or gave a negative response to the RSET that Exim sends to abort the
5282 transaction. The observed case was a dropped connection after DATA that had
5283 been sent in pipelining mode. That is, the server had advertised PIPELINING
5284 but was not implementing it correctly. The effect of the bug was incorrect
5285 behaviour, such as trying another host, and this could lead to a crash.
5291 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5292 and daemon.c were passed as pointers to ints; they should have been
5293 pointers to socklen_t variables (which are typically unsigned ints).
5295 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5298 3. Fixed a really odd bug that affected only the testing scheme; patching a
5299 certain fixed string in the binary changed the value of another string that
5300 happened to be identical to the end of the original first string.
5302 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5303 name", it returns that address as the IP address. On some operating
5304 systems (e.g. Solaris), it also passes back the IP address string as the
5305 "host name". However, on others (e.g. Linux), it passes back an empty
5306 string. Exim wasn't checking for this, and was changing the host name to an
5307 empty string, assuming it had been canonicized.
5309 5. Although rare, it is permitted to have more than one PTR record for a given
5310 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5311 all the names associated with an address, because they do in Solaris.
5312 However, it seems that they do not in Linux for data that comes from the
5313 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5314 I found this out when I moved to a new Linux workstation and tried to run
5315 the Exim test suite.
5317 To get round this problem I have changed the code so that it now does its
5318 own call to the DNS to look up PTR records when searching for a host name.
5319 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5320 addresses that are only in /etc/hosts are still found.
5322 This behaviour is, however, controlled by an option called host_lookup_
5323 order, which defaults to "bydns:byaddr". If people want to use the other
5324 order, or indeed, just use one or the other means of lookup, they can
5325 specify it in this variable.
5327 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5328 some operating systems, this comes back from gethostbyaddr() as an empty
5329 string, and this is what Exim used to test for. However, it seems that in
5330 other systems, "." is yielded. Exim now tests for this case too.
5332 7. The values of check_spool_space and check_log_space are now held internally
5333 as a number of kilobytes instead of an absolute number of bytes. If a
5334 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5335 kilobyte. This means that much larger values can be stored.
5337 8. Exim monitor: an attempt to get the action menu when not actually pointing
5338 at a message produces an empty menu entitled "No message selected". This
5339 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5340 no entries in it ("Shell widget menu has zero width and/or height"). So I
5341 have added a single, blank menu entry in this case.
5343 9. Added ${quote_local_part.
5345 10. MIME decoding is now applied to the contents of Subject: header lines when
5348 11. Now that a reference to $sender_host_address automatically causes a reverse
5349 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5350 host lookup before query-style lookups in lists that might use this
5351 variable. This has therefore been abolished, and the "net-" prefix is no
5352 longer necessary for query-style lookups.
5354 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5355 have been a typo for LFLAGS, so it has been changed.
5357 13. The install script calls Exim with "-C /dev/null" in order to find the
5358 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5359 to be output. Howeve, since Exim outputs its version number before the
5360 error, it didn't break the script. It just looked ugly. I fixed this by
5361 always allowing "-C /dev/null" if the caller is root.
5363 14. Ignore overlarge ACL variable number when reading spool file - insurance
5364 against a later release with more variables having written the file.
5366 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5367 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5368 this, as well as the form without the "IPv6" on the front.
5370 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5371 OS/Makefile-Darwin file.
5373 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5374 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5376 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5377 variable. However, some OS use ints for inodes. I've added cast to long int
5378 to get rid of the compiler warning.
5380 19. I had forgotten to lock out "/../" in configuration file names when
5381 ALT_CONFIG_PREFIX was set.
5383 20. Routers used for verification do not need to specify transports. However,
5384 if such a router generated a host list, and callout was configured, Exim
5385 crashed, because it could not find a port number from the (non-existent)
5386 transport. It now assumes port 25 in this circumstance.
5388 21. Added the -t option to exigrep.
5390 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5391 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5392 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5393 them is set, LOOKUP_LSEARCH is forced.
5395 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5397 24. Added sender and host information to the "rejected by local_scan()" log
5398 line; previously there was no indication of these.
5400 25. Added .include_if_exists.
5402 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5403 Linux. It turns out that not all file systems support this. Apparently some
5404 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5405 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5406 not supported on the file descriptor, is now ignored when Exim is trying to
5407 sync a directory. This applies only to Linux.
5409 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5411 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5412 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5415 29. Added support for maildirsize files from supplied patch (modified a bit).
5417 30. The use of :fail: followed by an empty string could lead Exim to respond to
5418 sender verification failures with (e.g.):
5420 550 Verification failed for <xxx>
5421 550 Sender verify failed
5423 where the first response line was missing the '-' that indicates it is not
5424 the final line of the response.
5426 31. The loop for finding the name of the user that called Exim had a hardwired
5427 limit of 10; it now uses the value of finduser_retries, which is used for
5428 all other user lookups.
5430 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5433 33. Exim was neglecting to zero errno before one call of strtol() when
5434 expanding a string and expecting an integer value. On some systems this
5435 resulted in spurious "integer overflow" errors. Also, it was casting the
5436 result into an int without checking.
5438 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5439 did not work. The code looks as if it has *never* worked, though it appears
5440 to have been documented since at least releast 1.62. I have made it work.
5442 35. The "timeout_DNS" error in retry rules, also documented since at least
5443 1.62, also never worked. As it isn't clear exactly what this means, and
5444 clearly it isn't a major issue, I have abolished the feature by treating it
5445 as "timeout", and writing a warning to the main and panic logs.
5447 36. The display of retry rules for -brt wasn't always showing the error code
5450 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5451 timeout_connect_A, timeout_connect_MX.
5453 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5454 to the empty sender.
5456 39. The daemon was not analysing the content of -oX till after it had closed
5457 stderr and disconnected from the controlling terminal. This meant that any
5458 syntax errors were only noted on the panic log, and the return code from
5459 the command was 0. By re-arranging the code a little, I've made the
5460 decoding happen first, so such errors now appear on stderr, and the return
5461 code is 1. However, the actual setting up of the sockets still happens in
5462 the disconnected process, so errors there are still only recorded on the
5465 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5466 connections (as happens on some IP stacks) was logged at start up time as
5467 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5468 it from "IPv6 and IPv4", which means that two separate sockets are being
5471 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5472 says whether AF_INET or AF_INET6 was passed as an argument.
5474 42. Exiwhat output was messed up when time zones were included in log
5477 43. Exiwhat now gives more information about the daemon's listening ports,
5478 and whether -tls-on-connect was used.
5480 44. The "port" option of the smtp transport is now expanded.
5482 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5483 silently ignored. Now an error message is written to the main and panic
5486 46. There's a new ACL modifier called "logwrite" which writes to a log file
5487 as soon as it is encountered.
5489 47. Added $local_user_uid and $local_user_gid at routing time.
5491 48. Exim crashed when trying to verify a sender address that was being
5494 49. Exim was recognizing only a space character after ".include". It now also
5495 recognizes a tab character.
5497 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5498 extracting the relevant information from the specification. The man page no
5499 longer contains scrambled data for the -d option, and I've added a section
5500 at the front about calling Exim under different names.
5502 51. Added "extra_headers" argument to the "mail" command in filter files.
5504 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5507 53. Installed eximstats 1.29.
5509 54. Added transport_filter_timeout as a generic transport option.
5511 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5512 Cc: header lines. This was required by RFC 822, but it not required by RFC
5515 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5516 incoming messages that did not have them. Now it does so only if the
5517 message originates locally, that is, if there is no associated remote host
5518 address. When Resent- header lines are present, this applies to the Resent-
5519 lines rather than the non-Resent- lines.
5521 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5522 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5524 58. Messages for configuration errors now include the name of the main
5525 configuration file - useful now that there may be more than one file in a
5526 list (.included file names were always shown).
5528 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5529 for those rare installations that do not start the daemon as root or run it
5530 setuid root. I've cut out the call to initgroups() if the daemon is not
5533 60. The Exim user and group can now be bound into the binary as text strings
5534 that are looked up at the start of Exim's processing.
5536 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5538 62. Added $mailstore_basename variable.
5540 63. Installed patch to sieve.c from Michael Haardt.
5542 64. When Exim failed to open the panic log after failing to open the main log,
5543 the original message it was trying to log was written to stderr and debug
5544 output, but if they were not available (the usual case in production), it
5545 was lost. Now it is written to syslog before the two lines that record the
5546 failures to open the logs.
5548 65. Users' Exim filters run in subprocesses under the user's uid. It is
5549 possible for a "deliver" command or an alias in a "personal" command to
5550 provoke an address rewrite. If logging of address rewriting is configured,
5551 this fails because the process is not running as root or exim. There may be
5552 a better way of dealing with this, but for the moment (because 4.30 needs
5553 to be released), I have disabled address rewrite logging when running a
5554 filter in a non-root, non-exim process.
5560 1. The buildconfig auxiliary program wasn't quoting the value set for
5561 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5562 not defined. This bug was masked in 4.22 by the effect that was fixed in
5565 2. Some messages that were rejected after a message id was allocated were
5566 shown as "incomplete" by exigrep. It no longer does this for messages that
5567 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5569 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5570 have allow_domain_literals set, the ID did not get logged in the <= line.
5571 Domain literals are now always recognized in Message-ID: header lines.
5573 4. The first argument for a ${extract expansion item is the key name or field
5574 number. Leading and trailing spaces in this item were not being ignored,
5575 causing some misleading effects.
5577 5. When deliver_drop_privilege was set, single queue runner processes started
5578 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5579 same command in the process it spins off) were not dropping privilege.
5581 6. When the daemon running as "exim" started a queue runner, it always
5582 re-executed Exim in the spun-off process. This is a waste of effort when
5583 deliver_drop_privilege is set. The new process now just calls the
5584 queue-runner function directly.
5590 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5593 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5594 ignored. Though the use of -oP was forcing the writing of a pid file, it
5595 was always written to the default place.
5597 3. If the message "no IP address found for host xxxx" is generated during
5598 incoming verification, it is now followed by identification of the incoming
5599 connection (so you can more easily find what provoked it).
5601 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5603 5. Added some features to "harden" Exim a bit more against certain attacks:
5605 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5606 be put in Local/Makefile. This is like the never_users runtime option,
5607 but it cannot be overridden. The default setting is "root".
5609 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5610 prefix string with which any file named in a -C command line option
5613 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5614 is retained for -C and -D only if the caller of Exim is root. Without
5615 it, the exim user may also use -C and -D and retain privilege.
5617 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5618 command line option is disabled.
5620 6. Macro names set by the -D option must start with an upper case letter, just
5621 like macro names defined in the configuration file.
5623 7. Added "dereference=" facility to LDAP.
5625 8. Two instances of the typo "uknown" in the source files are fixed.
5627 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5628 the Configure-Makefile script screwed up while processing it.
5630 10. Incorporated PCRE 4.4.
5632 11. The SMTP synchronization check was not operating right at the start of an
5633 SMTP session. For example, it could not catch a HELO sent before the client
5634 waited for the greeting. There is now a check for outstanding input at the
5635 point when the greeting is written. Because of the duplex, asynchronous
5636 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5637 way, but not yet received, when the check is performed.
5639 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5640 on TCP/IP sockets, because this apparently causes some broken clients to
5643 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5644 unchanged) from the Cygwin maintainer.
5646 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5647 was supported instead of testing for mbx. Effectively a typo.
5649 15. The spa authenticator server code was not checking that the input it
5650 received was valid base64.
5652 16. The debug output line for the "set" modifier in ACLs was not showing the
5653 name of the variable that was being set.
5655 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5657 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5658 lookup on its own. Something else was needed to trigger the lookup. For
5659 example, a match in host_lookup or the need for a host name in a host list.
5660 Now, if $sender_host_name is referenced and the host name has not yet been
5661 looked up, a lookup is performed. If the lookup fails, the variable remains
5662 empty, and $host_lookup_failed is set to "1".
5664 19. Added "eqi" as a case-independent comparison operator.
5666 20. The saslauthd authentication condition could segfault if neither service
5667 nor realm was specified.
5669 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5670 error message that was logged was misleading, and incoming SMTP
5671 connections were dropped. The message is now more accurate, and temporary
5672 errors are given to SMTP connections.
5674 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5675 (except RSET) with a 421 error, until QUIT is received. However, it was
5676 failing to send a response to QUIT.
5678 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5679 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5680 ACL is now run after the helo_try_verify_hosts code.
5682 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5683 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5684 case-independent, so other case variants are also recognized). Apparently
5685 some systems use these upper case variants.
5687 25. If more than two messages were waiting for the same host, and a transport
5688 filter was specified for the transport, Exim sent two messages over the
5689 same TCP/IP connection, and then failed with "socket operation on non-
5690 socket" when it tried to send the third.
5692 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5694 27. The extern definition of crypt16() in expand.c was not being excluded when
5695 the OS had its own crypt16() function.
5697 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5698 as a preferred synonym for return_size_limit, both as an option and as an
5701 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5703 30. Changed the default configuration ACL to relax the local part checking rule
5704 for addresses that are not in any local domains. For these addresses,
5705 slashes and pipe symbols are allowed within local parts, but the sequence
5706 /../ is explicitly forbidden.
5708 31. SPA server authentication was not clearing the challenge buffer before
5711 32. log_message in a "warn" ACL statement was writing to the reject log as
5712 well as to the main log, which contradicts the documentation and doesn't
5713 seem right (because no rejection is happening). So I have stopped it.
5715 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5716 However, I am unable to do any testing of this.
5718 34. Fixed an infelicity in the appendfile transport. When checking directories
5719 for a mailbox, to see if any needed to be created, it was accidentally
5720 using path names with one or more superfluous leading slashes; tracing
5721 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5723 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5724 contents are added to the log line that is written for every discarded
5725 recipient. (Previously a log_message setting was ignored.)
5727 36. The ${quote: operator now quotes the string if it is empty.
5729 37. The install script runs exim in order to find its version number. If for
5730 some reason other than non-existence or emptiness, which it checks, it
5731 could not run './exim', it was installing it with an empty version number,
5732 i.e. as "exim-". This error state is now caught, and the installation is
5735 38. An argument was missing from the function that creates an error message
5736 when Exim fails to connect to the socket for saslauthd authentication.
5737 This could cause Exim to crash, or give a corrupted message.
5739 39. Added isip, isip4, and isip6 to ${if conditions.
5741 40. The ACL variables $acl_xx are now saved with the message, and can be
5742 accessed later in routers, transports, and filters.
5744 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5745 strings in the file are not string-expanded.
5747 42. If a MAIL command specified a SIZE value that was too large to fit into an
5748 int variable, the check against message_size_limit failed. Such values are
5749 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5750 day this will have to be increased, but I don't think I want to be around
5751 when emails are that large.
5758 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5759 iconv() is not standard in FreeBSD.
5761 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5762 IPv6 enabled. The observed symptom was a segmentation fault on return from
5763 the function os_common_find_running_interfaces() in src/os.c.
5765 3. In the check_special_case() function in daemon.c I had used "errno" as an
5766 argument name, which causes warnings on some systems. This was basically a
5767 typo, since it was named "eno" in the comments!
5769 4. The code that waits for the clock to tick (at a resolution of some fraction
5770 of a second) so as to ensure message-id uniqueness was always waiting for
5771 at least one whole tick, when it could have waited for less. [This is
5772 almost certainly not relevant at current processor speeds, where it is
5773 unlikely to ever wait at all. But we try to future-proof.]
5775 5. The function that sleeps for a time interval that includes fractions of a
5776 second contained a race. It did not block SIGALRM between setting the
5777 timer, and suspending (a couple of lines later). If the interval was short
5778 and the sigsuspend() was delayed until after it had expired, the suspension
5779 never ended. On busy systems this could lead to processes getting stuck for
5782 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5783 process, before it forks any delivery processes. The open lookup caching
5784 mechanism meant that the open file or database connection was passed into
5785 the delivery process. The problem was that delivery processes always tidy
5786 up cached lookup data. This could cause a problem for the next delivery
5787 process started by the queue runner, because the external queue runner
5788 process does not know about the closure. So the next delivery process
5789 still has data in the lookup cache. In the case of a file lookup, there was
5790 no problem because closing a file descriptor in a subprocess doesn't affect
5791 the parent. However, if the lookup was caching a connection to a database,
5792 the connection was closed, and the second delivery process was likely to
5793 see errors such as "PGSQL: query failed: server closed the connection
5794 unexpectedly". The problem has been fixed by closing all cached lookups
5795 in a queue runner before running a delivery process.
5797 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5798 seem to have the "const" qualifier which it has on other OS. I've
5801 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5802 *of the same type* (client or server) with the same public name that an
5803 error should be diagnosed.
5805 9. When Exim looked up a host name for an IP address, but failed to find the
5806 original IP address when looking up the host name (a safety check), it
5807 output the message "<ip address> does not match any IP for NULL", which was
5808 confusing, to say the least. The bug was that the host name should have
5809 appeared instead of "NULL".
5811 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5812 user that is built into the binary, and the -C or -D options is used, root
5813 privilege is dropped before the configuration file is read. In addition,
5814 logging is switched to stderr instead of the normal log files. If the
5815 configuration then re-defines the Exim user, the unprivileged environment
5816 is probably not what is expected, so Exim logs a panic warning message (but
5819 However, if deliver_drop_privilege is set, the unprivileged state may well
5820 be exactly what is intended, so the warning has been cut out in that case,
5821 and Exim is allowed to try to write to its normal log files.
5827 1. smtp_return_error_details was not giving details for temporary sender
5828 or receiver verification errors.
5830 2. Diagnose a configuration error if two authenticators have the same public
5833 3. Exim used not to create the message log file for a message until the first
5834 delivery attempt. This could be confusing when incoming messages were held
5835 for policy or load reasons. The message log file is now created at the time
5836 the message is received, and an initial "Received" line is written to it.
5838 4. The automatically generated man page for command line options had a minor
5839 bug that caused no ill effects; however, a more serious problem was that
5840 the procedure for building the man page automatically didn't always
5841 operate. Consequently, release 4.20 contains an out-of-date version. This
5842 shouldn't happen again.
5844 5. When building Exim with embedded Perl support, the script that builds the
5845 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5846 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5848 6. The freeze_tell option was not being used for messages that were frozen on
5849 arrival, either by an ACL or by local_scan().
5851 7. Added the smtp_incomplete_transaction log selector.
5853 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5854 was accepting AUTH without a new EHLO.
5856 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5857 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5860 10. Logging of TCP/IP connections (when configured) now happens in the main
5861 daemon process instead of the child process, so that the TCP/IP connection
5862 count is more accurate (but it can never be perfect).
5864 11. The use of "drop" in a nested ACL was not being handled correctly in the
5865 outer ACL. Now, if condition failure induced by the nested "drop" causes
5866 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5867 or "require"), the connection is dropped.
5869 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5870 that yield "discard" can now be used with an "accept" or a "discard" verb,
5871 but an error is generated for any others (because I can't see a useful way
5872 to define what should happen).
5874 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5875 are now processed in the same way as lines in the Exim configuration file.
5876 In particular, continuation lines are supported.
5878 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5880 15. Added -ti meaning -t -i.
5882 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5883 domains, and warn by logging if others are found.
5885 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5886 returned from GIFCONF when Exim is trying to find the list of interfaces on
5887 a host. The code in os.c has been modified to copy each ifreq to an aligned
5888 structure in all cases.
5890 Also, in some cases, the returned ifreq's were being copied to a 'struct
5891 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5892 means the last couple of bytes of an IPv6 address could be chopped if the
5893 ifreq contained only a normal sockaddr (14 bytes storage).
5895 18. Named domain lists were not supported in the hosts_treat_as_local option.
5896 An entry such as +xxxx was not recognized, and was treated as a literal
5899 19. Ensure that header lines added by a DATA ACL are included in the reject log
5900 if the ACL subsequently rejects the message.
5902 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5903 MD5 (which is deprecated).
5905 21. When testing a filter file using -bf, Exim was writing a message when it
5906 took the sender from a "From " line in the message, but it was not doing so
5907 when it took $return_path from a Return-Path: header line. It now does.
5909 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5910 with a valid header line field name (a series of printing characters
5911 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5913 23. Changed "disc" in the source to "disk" to conform to the documentation and
5914 the book and for uniformity.
5916 24. Ignore Sendmail's -Ooption=value command line item.
5918 25. When execve() failed while trying to run a command in a pipe transport,
5919 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5920 could be confused with a return value of 69 from the command itself. This
5921 has been changed to 127, the value the shell returns if it is asked to run
5922 a non-existent command. The wording for the related log line suggests a
5923 non-existent command as the problem.
5925 26. If received_header_text expands to an empty string, do not add a Received:
5926 header line to the message. (Well, it adds a token one on the spool, but
5927 marks it "old" so that it doesn't get used or transmitted.)
5929 27. Installed eximstats 1.28 (addition of -nt option).
5931 28. There was no check for failure on the call to getsockname() in the daemon
5932 code. This can fail if there is a shortage of resources on the system, with
5933 ENOMEM, for example. A temporary error is now given on failure.
5935 29. Contrary to the C standard, it seems that in some environments, the
5936 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5937 program. Exim now does this explicitly; it affects the formatting of
5938 timestamps using strftime().
5940 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5941 complaints. I've now initialized all the variables, to avoid this.
5943 32. Header lines added by a system filter were not being "seen" during
5944 transport-time rewrites.
5946 33. The info_callback() function passed to OpenSSL is set up with type void
5947 (*)(SSL *, int, int), as described somewhere. However, when calling the
5948 function (actually a macro) that sets it up, the type void(*)() is
5949 expected. I've put in a cast to prevent warnings from picky compilers.
5951 34. If a DNS black list lookup found a CNAME record, but there were no A
5952 records associated with the domain it pointed at, Exim crashed.
5954 35. If a DNS black list lookup returned more than one A record, Exim ignored
5955 all but the first. It now scans all returned addresses if a particular IP
5956 value is being sought. In this situation, the contents of the
5957 $dnslist_value variable are a list of all the addresses, separated by a
5960 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5961 to accept a host name and all its aliases if the forward lookup for any of
5962 them yielded the IP address of the incoming connection. Now it accepts only
5963 those names whose forward lookup yields the correct IP address. Any other
5964 names are discarded. This closes a loophole whereby a rogue DNS
5965 administrator could create reverse DNS records to break through a
5966 wildcarded host restriction in an ACL.
5968 37. If a user filter or a system filter that ran in a subprocess used any of
5969 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5970 the wrong values were passed to the pipe command ($thisaddress had the
5971 value of $0, $0 had the value of $1, etc). This bug was introduced by
5972 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5974 38. Improved the line breaking for long SMTP error messages from ACLs.
5975 Previously, if there was no break point between 40 and 75 characters, Exim
5976 left the rest of the message alone. Two changes have been made: (a) I've
5977 reduced the minimum length to 35 characters; (b) if it can't find a break
5978 point between 35 and 75 characters, it looks ahead and uses the first one
5979 that it finds. This may give the occasional overlong line, but at least the
5980 remaining text gets split now.
5982 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5983 file descriptors might be low, and that setting 1000 would always raise it.
5984 It turns out that in some environments, the limit is already over 1000 and
5985 that lowering it causes trouble. So now Exim takes care not to decrease it.
5987 40. When delivering a message, the value of $return_path is set to $sender_
5988 address at the start of routing (routers may change the value). By an
5989 oversight, this default was not being set up when an address was tested by
5990 -bt or -bv, which affected the outcome if any router or filter referred to
5993 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5994 log only when "message" or "log_message" are set. However, if one of the
5995 conditions was an address verification, or a call to a nested ACL, the
5996 messages generated by the underlying test were being passed through. This
5997 no longer happens. The underlying message is available in $acl_verify_
5998 message for both "message" and "log_message" expansions, so it can be
5999 passed through if needed.
6001 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6002 new expansion $bh_ to give the encoded byte string without charset
6003 translation. Translation happens only if iconv() is available; HAVE_ICONV
6004 indicates this at build time. HEADERS_CHARSET gives the charset to
6005 translate to; headers_charset can change it in the configuration, and
6006 "headers charset" can change it in an individual filter file.
6008 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6009 that creates RFC 2047 encoded "words" labels them as that charset instead
6010 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6011 expansion operator; (ii) when Exim creates a From: line for a local
6012 message; (iii) when a header line is rewritten to include a "phrase" part.
6014 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6015 buggy, causing it to skip the first lines of messages whose message ID
6016 ended in 'D'. This would not have bitten before Exim release 4.14, because
6017 message IDs were unlikely to end in 'D' before then. The effect was to have
6018 incorrect size information for certain domains.
6020 45. #include "config.h" was missing at the start of the crypt16.c module. This
6021 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6024 46. If there was a timeout during a "random" callout check, Exim treated it as
6025 a failure of the random address, and carried on sending RSET and the real
6026 address. If the delay was just some slowness somewhere, the response to the
6027 original RCPT would be taken as a response to RSET and so on, causing
6028 mayhem of various kinds.
6030 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6031 when I implemented it. It didn't allow for the fact that some option values
6032 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6033 the right test for positive values.
6035 48. Domain names in DNS records are case-independent. Exim always looks them up
6036 in lower case. Some resolvers return domain names in exactly the case they
6037 appear in the zone file, that is, they may contain uppercase letters. Not
6038 all resolvers do this - some return always lower case. Exim was treating a
6039 change of case by a resolver as a change of domain, similar to a widening
6040 of a domain abbreviation. This triggered its re-routing code and so it was
6041 trying to route what was effectively the same domain again. This normally
6042 caused routing to fail (because the router wouldn't handle the domain
6043 twice). Now Exim checks for this case specially, and just changes the
6044 casing of the domain that it ultimately uses when it transmits the message
6047 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6050 50. If a filter generated a file delivery with a non-absolute name (possible if
6051 no home directory exists for the router), the forbid_file option was not
6054 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6055 the existing equality matching.
6057 52. Exim was using ints instead of ino_t variables in some places where it was
6058 dealing with inode numbers.
6060 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6061 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6062 called TMPDIR, and if it finds it is different, it changes its value.
6064 54. The smtp_printf() function is now made available to local_scan() so
6065 additional output lines can be written before returning. There is also an
6066 smtp_fflush() function to enable the detection of a dropped connection.
6067 The variables smtp_input and smtp_batched_input are exported to
6070 55. Changed the default runtime configuration: the message "Unknown user"
6071 has been removed from the ACL, and instead placed on the localuser router,
6072 using the cannot_route_message feature. This means that any verification
6073 failures that generate their own messages won't get overridden. Similarly,
6074 the "Unrouteable address" message that was in the ACL for unverifiable
6075 relay addresses has also been removed.
6077 56. Added hosts_avoid_esmtp to the smtp transport.
6079 57. The exicyclog script was not checking for the esoteric option
6080 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6081 will work only if exicyclog is run under the appropriate euid.
6083 58. Following a discussion on the list, the rules by which Exim recognises line
6084 endings on incoming messages have been changed. The -dropcr and drop_cr
6085 options are now no-ops, retained only for backwards compatibility. The
6086 following line terminators are recognized: LF CRLF CR. However, special
6087 processing applies to CR:
6089 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6090 nor a local message in the state where . is a terminator.
6092 (ii) If a bare CR is encountered in a header line, an extra space is added
6093 after the line terminator so as not to end the header. The reasoning
6094 behind this is that bare CRs in header lines are most likely either
6095 to be mistakes, or people trying to play silly games.
6097 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6098 was being incorrectly given as 18 bytes larger than it should have been.
6099 This is a VOB (very old bug).
6101 60. This may never have affected anything current, but just in case it has:
6102 When the local host is found other than at the start of a list of hosts,
6103 the local host, those with the same MX, and any that follow, are discarded.
6104 When the list in question was part of a longer list of hosts, the following
6105 hosts (not currently being processed) were also being discarded. This no
6106 longer happens. I'm not sure if this situation could ever has previously
6109 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6112 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6113 header. This is something that is recommended in a new Internet Draft, and
6114 is something that is documented as being done by Sendmail. There are two
6115 possible values. For messages generated by the autoreply transport, Exim
6118 Auto-Submitted: auto-replied
6120 whereas for all other generated messages (e.g. bounces) it adds
6122 Auto-Submitted: auto-generated
6124 63. The "personal" condition in filters now includes a test for the
6125 Auto-Submitted: header. If it contains the string "auto-" the message it
6126 not considered personal.
6128 64. Added rcpt_include_affixes as a generic transport option.
6130 65. Added queue_only_override (default true).
6132 66. Added the syslog_duplication option.
6134 67. If what should have been the first header line of a message consisted of
6135 a space followed by a colon, Exim was mis-interpreting it as a header line.
6136 It isn't of course - it is syntactically invalid and should therefore be
6137 treated as the start of the message body. The misbehaviour could have
6138 caused a number of strange effects, including loss of data in subsequent
6139 header lines, and spool format errors.
6141 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6142 client host had authenticated. This control can now be exercised by an ACL
6143 for more flexibility.
6145 69. By default, callouts do not happen when testing with -bh. There is now a
6146 variant, -bhc, which does actually run the callout code, including
6147 consulting and updating the callout cache.
6149 70. Added support for saslauthd authentication, courtesy of Alexander
6152 71. If statvfs() failed on the spool or log directories while checking their
6153 size for availability, Exim confusingly gave the error "space shortage".
6154 Furthermore, in debugging mode it crashed with a floating point exception.
6155 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6156 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6157 serious problem, Exim now writes to the main and panic logs when this
6158 happens, with details of the failure. It then refuses to accept the
6159 incoming message, giving the message "spool directory problem" or "log
6160 directory problem" with a 421 code for SMTP messages.
6162 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6163 0, 1, and 2 exist, because some OS complain for execs without them (see
6164 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6165 descriptors. However, the code omitted to check that the open succeeded,
6166 causing mysterious errors if for some reason the permissions on /dev/null
6167 got screwed. Now Exim writes a message to the main and panic logs, and
6168 bombs out if it can't open /dev/null.
6170 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6171 interact so that it is all more flexible. It is supposed to remain
6172 backwards compatible. Also added extra_local_interfaces.
6174 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6175 to bomb out with an assertion failure - to the client this appears as a
6176 connection drop. This problem occurs in the part of the code that was taken
6177 from the Samba project. Fortunately, the assertion is in a very simple
6178 function, so I have fixed this by reproducing the function inline in the
6179 one place where it is called, and arranging for authentication to fail
6180 instead of killing the process with assert().
6182 75. The SPA client code was not working when the server requested OEM rather
6183 than Unicode encoding.
6185 76. Added code to make require_files with a specific uid setting more usable in
6186 the case where statting the file as root fails - usually a non-root-mounted
6187 NFS file system. When this happens and the failure is EACCES, Exim now
6188 forks a subprocess and does the per-uid checking as the relevant uid.
6190 77. Added process_log_path.
6192 78. If log_file_path was not explicitly set, a setting of check_log_space or
6193 check_log_inodes was ignored.
6195 79. If a space check for the spool or log partitions fails, the incident is now
6196 logged. Of course, in the latter case the data may get lost...
6198 80. Added the %p formatting code to string_format() so that it can be used to
6199 print addresses in debug_print(). Adjusted all the address printing in the
6200 debugging in store.c to use %p rather than %d.
6202 81. There was a concern that a line of code in smtp_in.c could overflow a
6203 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6204 initially expressed, the concern was not well-founded, because trailing
6205 spaces are removed early. However, if the trailing spaces were followed by
6206 a NULL, they did not get removed, so the overflow was possible. Two fixes
6209 (a) I re-wrote the offending code in a cleaner fashion.
6210 (b) If an incoming SMTP command contains a NULL character, it is rejected
6213 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6214 runs initgroups(), so that if the Exim user is in any additional groups,
6215 they will be used during message reception.
6221 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob