1 Change log file for Exim from version 4.21
2 -------------------------------------------
5 HS/04 Add support for keep_environment and add_environment options.
8 TL/01 When running the test suite, the README says that variables such as
9 no_msglog_check are global and can be placed anywhere in a specific
10 test's script, however it was observed that placement needed to be near
11 the beginning for it to behave that way. Changed the runtest perl
12 script to read through the entire script once to detect and set these
13 variables, reset to the beginning of the script, and then run through
14 the script parsing/test process like normal.
16 TL/02 The BSD's have an arc4random API. One of the functions to induce
17 adding randomness was arc4random_stir(), but it has been removed in
18 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
19 function when detected.
21 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
22 cause callback expansion.
24 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
25 syntax errors in an expansion can be treated as a string instead of
26 logging or causing an error, due to the internal use of bool_lax
27 instead of bool when processing it.
29 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
30 server certificates when making smtp deliveries.
32 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
34 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
36 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
38 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
39 Merged patch from Sebastian Wiedenroth.
41 JH/05 Fix results-pipe from transport process. Several recipients, combined
42 with certificate use, exposed issues where response data items split
43 over buffer boundaries were not parsed properly. This eventually
44 resulted in duplicates being sent. This issue only became common enough
45 to notice due to the introduction of conection certificate information,
46 the item size being so much larger. Found and fixed by Wolfgang Breyha.
48 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
49 size buffer was used, resulting in syntax errors when an expansion
52 JH/07 Add support for directories of certificates when compiled with a GnuTLS
53 version 3.3.6 or later.
55 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
56 is EXPERIMENTAL_EVENT, the main-configuration and transport options
57 both become "event_action", the variables become $event_name, $event_data
58 and $event_defer_errno. There is a new variable $verify_mode, usable in
59 routers, transports and related events. The tls:cert event is now also
60 raised for inbound connections, if the main configuration event_action
63 TL/06 In test suite, disable OCSP for old versions of openssl which contained
64 early OCSP support, but no stapling (appears to be less than 1.0.0).
66 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
67 server certificate names available under the smtp transport option
68 "tls_verify_cert_hostname" now do not permit multi-component wildcard
71 JH/10 Time-related extraction expansions from certificates now use the main
72 option "timezone" setting for output formatting, and are consistent
73 between OpenSSL and GnuTLS compilations. Bug 1541.
75 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
76 encoded parameter in the incoming message. Bug 1558.
78 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
79 include certificate info, eximon was claiming there were spoolfile
82 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
84 JH/14 Log delivery-related information more consistently, using the sequence
85 "H=<name> [<ip>]" wherever possible.
87 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
88 are problematic for Debian distribution, omit them from the release
91 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
93 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
96 JH/17 Fix a null-indirection in certextract expansions when a nondefault
97 output list separator was used.
102 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
103 checkers that were complaining about end of non-void function with no
106 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
107 This was a regression intruduced in 4.83 by another bugfix.
109 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
111 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
112 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
118 TF/01 Correctly close the server side of TLS when forking for delivery.
120 When a message was received over SMTP with TLS, Exim failed to clear up
121 the incoming connection properly after forking off the child process to
122 deliver the message. In some situations the subsequent outgoing
123 delivery connection happened to have the same fd number as the incoming
124 connection previously had. Exim would try to use TLS and fail, logging
125 a "Bad file descriptor" error.
127 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
128 utilities have not been installed.
130 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
131 temporary space as the ACL may create new global variables.
133 TL/01 LDAP support uses per connection or global context settings, depending
134 upon the detected version of the libraries at build time.
136 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
137 to extract and use the src ip:port in logging and expansions as if it
138 were a direct connection from the outside internet. PPv2 support was
139 updated based on HAProxy spec change in May 2014.
141 JH/02 Add ${listextract {number}{list}{success}{fail}}.
143 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
144 Properly escape header and check for NULL return.
146 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
149 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
151 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
152 characters in header names, implemented as a verify condition.
153 Contributed by Michael Fischer v. Mollard.
155 TL/05 Rename SPF condition results err_perm and err_temp to standardized
156 results permerror and temperror. Previous values are deprecated but
157 still accepted. In a future release, err_perm and err_temp will be
158 completely removed, which will be a backward incompatibility if the
159 ACL tests for either of these two old results. Patch contributed by
160 user bes-internal on the mailing list.
162 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
164 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
165 selectors, in both main and reject logs.
167 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
170 JH/07 Add malware type "sock" for talking to simple daemon.
172 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
174 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
175 routers/transports under cutthrough routing.
177 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
178 numbers. Touch up "bool" conditional to keep the same definition.
180 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
182 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
184 JH/12 Expand items in router/transport headers_add or headers_remove lists
185 individually rather than the list as a whole. Bug 1452.
187 Required for reasonable handling of multiple headers_ options when
188 they may be empty; requires that headers_remove items with embedded
189 colons must have them doubled (or the list-separator changed).
191 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
192 view the policy declared in the DMARC record. Currently, $dmarc_status
193 is a combined value of both the record presence and the result of the
196 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
198 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
199 dnslookup router and the smtp transport (applying to the forward
202 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
203 of ldap servers used for a specific lookup. Patch provided by Heiko
206 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
207 New variable $lookup_dnssec_authenticated for observability.
209 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
210 Patch submitted by Lars Timman.
212 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
214 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
215 Requires trusted mode and valid format message id, aborts otherwise.
216 Patch contributed by Heiko Schlichting.
218 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
219 certextract with support for various fields. Bug 1358.
221 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
222 is requested by default, modifiable by smtp transport option
225 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
226 operate on certificate variables to give certificate fingerprints
227 Also new ${sha256:cert_variable}.
229 JH/23 The PRDR feature is moved from being Experimental into the mainline.
231 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
232 Christian Aistleitner.
234 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
236 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
237 file. Patch from Wolfgang Breyha.
239 JH/25 Expand the coverage of the delivery $host and $host_address to
240 client authenticators run in verify callout. Bug 1476.
242 JH/26 Port service names are now accepted for tls_on_connect_ports, to
243 align with daemon_smtp_ports. Bug 72.
245 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
246 support and error reporting did not work properly.
248 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
249 and is readable. Patch from Andrew Colin Kissa.
251 TL/14 Enhance documentation of ${run expansion and how it parses the
252 commandline after expansion, particularly in the case when an
253 unquoted variable expansion results in an empty value.
255 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
257 PP/02 Fix internal collision of T_APL on systems which support RFC3123
258 by renaming away from it. Addresses GH issue 15, reported by
261 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
263 TL/15 SECURITY: prevent double expansion in math comparison functions
264 (can expand unsanitized data). Not remotely exploitable.
271 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
273 PP/02 Make -n do something, by making it not do something.
274 When combined with -bP, the name of an option is not output.
276 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
279 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
280 $sender_host_name and config options to manage this, and basic check
283 PP/05 DSCP support for outbound connections and control modifier for inbound.
285 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
286 (Only plugin which currently uses this is kerberos4, which nobody should
287 be using, but we should make it available and other future plugins might
288 conceivably use it, even though it would break NAT; stuff *should* be
289 using channel bindings instead).
291 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
292 name; added for Sendmail compatibility; requires admin caller.
293 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
294 just ignore it); requires trusted caller.
295 Also parse but ignore: -Ac -Am -X<logfile>
298 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
300 TL/02 Add +smtp_confirmation as a default logging option.
302 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
303 Patch by Magnus Holmgren from 2007-02-20.
305 TL/04 Bugzilla 1281 - Spec typo.
306 Bugzilla 1283 - Spec typo.
307 Bugzilla 1290 - Spec grammar fixes.
309 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
311 TL/06 Add Experimental DMARC support using libopendmarc libraries.
313 TL/07 Fix an out of order global option causing a segfault. Reported to dev
314 mailing list by by Dmitry Isaikin.
316 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
318 JH/02 Support "G" suffix to numbers in ${if comparisons.
320 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
322 NM/01 Bugzilla 1197 - Spec typo
323 Bugzilla 1196 - Spec examples corrections
325 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
327 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
328 gnutls_enable_pkcs11, but renamed to more accurately indicate its
331 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
332 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
334 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
335 "acl {{name}{arg}...}", and optional args on acl condition
338 JH/05 Permit multiple router/transport headers_add/remove lines.
340 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
342 JH/07 Avoid using a waiting database for a single-message-only transport.
343 Performance patch from Paul Fisher. Bugzilla 1262.
345 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
348 JH/09 Add $headers_added variable, with content from use of ACL modifier
349 add_header (but not yet added to the message). Bugzilla 199.
351 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
352 Pulled from Bugzilla 817 by Wolfgang Breyha.
354 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
356 (nb: this is the same fix as in Exim 4.80.1)
358 JH/11 Add A= logging on delivery lines, and a client_set_id option on
361 JH/12 Add optional authenticated_sender logging to A= and a log_selector
364 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
366 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
367 advertise SMTP AUTH mechanism to us, instead of a generic
368 protocol violation error. Also, make Exim more robust to bad
369 data from the Dovecot auth socket.
371 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
373 When a queue runner is handling a message, Exim first routes the
374 recipient addresses, during which it prunes them based on the retry
375 hints database. After that it attempts to deliver the message to
376 any remaining recipients. It then updates the hints database using
379 So if a recipient address works intermittently, it can get repeatedly
380 deferred at routing time. The retry hints record remains fresh so the
381 address never reaches the final cutoff time.
383 This is a fairly common occurrence when a user is bumping up against
384 their storage quota. Exim had some logic in its local delivery code
385 to deal with this. However it did not apply to per-recipient defers
386 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
388 This change adds a proper retry rule check during routing so that the
389 final cutoff time is checked against the message's age. We only do
390 this check if there is an address retry record and there is not a
391 domain retry record; this implies that previous attempts to handle
392 the address had the retry_use_local_parts option turned on. We use
393 this as an approximation for the destination being like a local
394 delivery, as in LMTP.
396 I suspect this new check makes the old local delivery cutoff check
397 redundant, but I have not verified this so I left the code in place.
399 TF/02 Correct gecos expansion when From: is a prefix of the username.
401 Test 0254 submits a message to Exim with the header
405 When I ran the test suite under the user fanf2, Exim expanded
406 the header to contain my full name, whereas it should have added
407 a Resent-Sender: header. It erroneously treats any prefix of the
408 username as equal to the username.
410 This change corrects that bug.
412 GF/01 DCC debug and logging tidyup
413 Error conditions log to paniclog rather than rejectlog.
414 Debug lines prefixed by "DCC: " to remove any ambiguity.
416 TF/03 Avoid unnecessary rebuilds of lookup-related code.
418 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
419 Bug spotted by Jeremy Harris; was flawed since initial commit.
420 Would have resulted in OCSP responses post-SNI triggering an Exim
421 NULL dereference and crash.
423 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
425 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
426 Bug detection, analysis and fix by Samuel Thibault.
427 Bugzilla 1331, Debian bug #698092.
429 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
431 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
432 Server implementation by Todd Lyons, client by JH.
433 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
434 config variable "prdr_enable" controls whether the server
435 advertises the facility. If the client requests PRDR a new
436 acl_data_smtp_prdr ACL is called once for each recipient, after
437 the body content is received and before the acl_smtp_data ACL.
438 The client is controlled by bolth of: a hosts_try_prdr option
439 on the smtp transport, and the server advertisement.
440 Default client logging of deliveries and rejections involving
441 PRDR are flagged with the string "PRDR".
443 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
444 fclose(). Diagnosis by Todd Lyons.
446 PP/17 Update configure.default to handle IPv6 localhost better.
447 Patch by Alain Williams (plus minor tweaks).
450 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
451 This is now consistent with GnuTLS, and is now documented: the
452 previous undocumented portable approach to treating the option as
453 unset was to force an expansion failure. That still works, and
454 an empty string is now equivalent.
456 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
457 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
458 not performing validation itself.
460 PP/20 Added force_command boolean option to pipe transport.
461 Patch from Nick Koston, of cPanel Inc.
463 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
466 TF/04 Added udpsend ACL modifer and hexquote expansion operator
468 PP/21 Fix eximon continuous updating with timestamped log-files.
469 Broken in a format-string cleanup in 4.80, missed when I repaired the
470 other false fix of the same issue.
471 Report and fix from Heiko Schlichting.
474 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
475 Report from Prashanth Katuri.
477 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
478 It's SecureTransport, so affects any MacOS clients which use the
479 system-integrated TLS libraries, including email clients.
481 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
482 using a MIME ACL for non-SMTP local injection.
483 Report and assistance in diagnosis by Warren Baker.
485 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
487 JH/16 Fix comparisons for 64b. Bugzilla 1385.
489 TL/09 Add expansion variable $authenticated_fail_id to keep track of
490 last id that failed so it may be referenced in subsequent ACL's.
492 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
495 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
496 ldap library initialization, allowing self-signed CA's to be
497 used. Also properly sets require_cert option later in code by
498 using NULL (global ldap config) instead of ldap handle (per
499 session). Bug diagnosis and testing by alxgomz.
501 TL/12 Enhanced documentation in the ratelimit.pl script provided in
502 the src/util/ subdirectory.
504 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
505 renamed to Transport Post Delivery Action by Jeremy Harris, as
508 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
509 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
510 redis_servers = needs to be configured which will be used by the redis
511 lookup. Patch from Warren Baker, of The Packet Hub.
513 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
515 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
516 hostname or reverse DNS when processing a host list. Used suggestions
517 from multiple comments on this bug.
519 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
521 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
522 Missed a few lines, added it to make the runtest require no keyboard
525 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
526 contains upper case chars. Make router use caseful_local_part.
528 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
529 support when GnuTLS has been built with p11-kit.
535 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
537 This, or similar/improved, will also be change PP/11 of 4.82.
543 PP/01 Handle short writes when writing local log-files.
544 In practice, only affects FreeBSD (8 onwards).
545 Bugzilla 1053, with thanks to Dmitry Isaikin.
547 NM/01 Bugzilla 949 - Documentation tweak
549 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
552 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
554 PP/02 Implemented gsasl authenticator.
556 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
558 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
559 `pkg-config foo` for cflags/libs.
561 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
562 with rest of GSASL and with heimdal_gssapi.
564 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
565 `pkg-config foo` for cflags/libs for the TLS implementation.
567 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
568 properties get this fed in as external SSF. A number of robustness
569 and debugging improvements to the cyrus_sasl authenticator.
571 PP/08 cyrus_sasl server now expands the server_realm option.
573 PP/09 Bugzilla 1214 - Log authentication information in reject log.
574 Patch by Jeremy Harris.
576 PP/10 Added dbmjz lookup type.
578 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
580 PP/12 MAIL args handles TAB as well as SP, for better interop with
581 non-compliant senders.
582 Analysis and variant patch by Todd Lyons.
584 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
585 Bug report from Lars Müller <lars@samba.org> (via SUSE),
586 Patch from Dirk Mueller <dmueller@suse.com>
588 PP/13 tls_peerdn now print-escaped for spool files.
589 Observed some $tls_peerdn in wild which contained \n, which resulted
590 in spool file corruption.
592 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
593 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
594 or write after TLS renegotiation, which otherwise led to messages
597 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
598 as a tracking header (ie: a signed header comes before the signature).
599 Patch from Wolfgang Breyha.
601 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
602 comma-sep list; embedded commas doubled.
604 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
606 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
608 Report and patch from Dmitry Banschikov.
610 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
611 Removed SSL_clear() after SSL_new() which led to protocol negotiation
612 failures. We appear to now support TLS1.1+ with Exim.
614 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
615 lets Exim select keys and certificates based upon TLS SNI from client.
616 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
617 before an outbound SMTP session. New log_selector, +tls_sni.
619 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
620 NULL dereference. Report and patch from Alun Jones.
622 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
623 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
624 Not seeing resolver debug output on NetBSD, but suspect this is a
625 resolver implementation change.
627 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
628 Left warnings. Added "eximon gdb" invocation mode.
630 PP/21 Defaulting "accept_8bitmime" to true, not false.
632 PP/22 Added -bw for inetd wait mode support.
634 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
635 locate the relevant includes and libraries. Made this the default.
637 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
638 Bugzilla 1246, report and most of solution from Tomasz Kusy.
640 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
641 This may cause build issues on older platforms.
643 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
644 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
645 gnutls_require_mac & gnutls_require_protocols (no longer supported).
646 Added SNI support via GnuTLS too.
647 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
649 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
651 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
652 Applied second patch from Janne, implementing suggestion to default
653 multiple-strings-in-record handling to match SPF spec.
655 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
657 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
658 read-only, out of scope).
659 Patch from Wolfgang Breyha, report from Stuart Northfield.
661 PP/29 Fix three issues highlighted by clang analyser static analysis.
662 Only crash-plausible issue would require the Cambridge-specific
663 iplookup router and a misconfiguration.
664 Report from Marcin Mirosław.
666 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
668 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
669 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
670 As part of this, removing so much warning spew let me fix some minor
671 real issues in debug logging.
673 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
674 assignment on my part. Fixed.
676 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
677 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
678 Janne Snabb (who went above and beyond: thank you).
680 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
681 string otherwise requires a connection and a bunch more work and it's
682 relatively easy to get wrong. Should also expose TLS library linkage
685 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
686 64-bit ${eval} (JH/03).
688 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
689 GNU libc to support some of the 64-bit stuff, should not lead to
690 conflicts. Defined before os.h is pulled in, so if a given platform
691 needs to override this, it can.
693 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
694 protection layer was required, which is not implemented.
695 Bugzilla 1254, patch from Wolfgang Breyha.
697 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
698 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
699 tls_dhparam take prime identifiers. Also unbreak combination of
700 OpenSSL+DH_params+TLSSNI.
702 PP/39 Disable SSLv2 by default in OpenSSL support.
708 PP/01 Solaris build fix for Oracle's LDAP libraries.
709 Bugzilla 1109, patch from Stephen Usher.
711 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
713 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
716 TF/02 Fix a couple more cases where we did not log the error message
717 when unlink() failed. See also change 4.74-TF/03.
719 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
720 lock up or crash if it happened to be inside a call to libc when it
721 got a SIGUSR1 from exiwhat.
723 The SIGUSR1 handler appends the current process status to the process
724 log which is later printed by exiwhat. It used to use the general
725 purpose logging code to do this, but several functions it calls are
726 not safe for signals.
728 The new output code in the SIGUSR1 handler is specific to the process
729 log, and simple enough that it's easy to inspect for signal safety.
730 Removing some special cases also simplifies the general logging code.
731 Removing the spurious timestamps from the process log simplifies
734 TF/04 Improved ratelimit ACL condition.
736 The /noupdate option has been deprecated in favour of /readonly which
737 has clearer semantics. The /leaky, /strict, and /readonly update modes
738 are mutually exclusive. The update mode is no longer included in the
739 database key; it just determines when the database is updated. (This
740 means that when you upgrde Exim will forget old rate measurements.)
742 Exim now checks that the per_* options are used with an update mode that
743 makes sense for the current ACL. For example, when Exim is processing a
744 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
745 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
746 must specify per_mail/readonly. If you omit the update mode it defaults to
747 /leaky where that makes sense (as before) or /readonly where required.
749 The /noupdate option is now undocumented but still supported for
750 backwards compatibility. It is equivalent to /readonly except that in
751 ACLs where /readonly is required you may specify /leaky/noupdate or
752 /strict/noupdate which are treated the same as /readonly.
754 A useful new feature is the /count= option. This is a generalization
755 of the per_byte option, so that you can measure the throughput of other
756 aggregate values. For example, the per_byte option is now equivalent
757 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
759 The per_rcpt option has been generalized using the /count= mechanism
760 (though it's more complicated than the per_byte equivalence). When it is
761 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
762 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
763 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
764 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
765 Note that using per_rcpt with a non-readonly update mode in more than
766 one ACL will cause the recipients to be double-counted. (The per_mail
767 and per_byte options don't have this problem.)
769 The handling of very low rates has changed slightly. If the computed rate
770 is less than the event's count (usually one) then this event is the first
771 after a long gap. In this case the rate is set to the same as this event's
772 count, so that the first message of a spam run is counted properly.
774 The major new feature is a mechanism for counting the rate of unique
775 events. The new per_addr option counts the number of different
776 recipients that someone has sent messages to in the last time period. It
777 behaves like per_rcpt if all the recipient addresses are different, but
778 duplicate recipient addresses do not increase the measured rate. Like
779 the /count= option this is a general mechanism, so the per_addr option
780 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
781 example, measure the rate that a client uses different sender addresses
782 with the options per_mail/unique=$sender_address. There are further
783 details in the main documentation.
785 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
787 TF/06 Removed a few PCRE remnants.
789 TF/07 Automatically extract Exim's version number from tags in the git
790 repository when doing development or release builds.
792 PP/02 Raise smtp_cmd_buffer_size to 16kB.
793 Bugzilla 879. Patch from Paul Fisher.
795 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
796 Heavily based on revision 40f9a89a from Simon Arlott's tree.
799 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
801 PP/05 Variable $av_failed, true if the AV scanner deferred.
802 Bugzilla 1078. Patch from John Horne.
804 PP/06 Stop make process more reliably on build failure.
805 Bugzilla 1087. Patch from Heiko Schlittermann.
807 PP/07 Make maildir_use_size_file an _expandable_ boolean.
808 Bugzilla 1089. Patch from Heiko Schlittermann.
810 PP/08 Handle ${run} returning more data than OS pipe buffer size.
811 Bugzilla 1131. Patch from Holger Weiß.
813 PP/09 Handle IPv6 addresses with SPF.
814 Bugzilla 860. Patch from Wolfgang Breyha.
816 PP/10 GnuTLS: support TLS 1.2 & 1.1.
818 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
821 PP/11 match_* no longer expand right-hand-side by default.
822 New compile-time build option, EXPAND_LISTMATCH_RHS.
823 New expansion conditions, "inlist", "inlisti".
825 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
827 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
829 PP/14 fix log_write() format string regression from TF/03.
830 Bugzilla 1152. Patch from Dmitry Isaikin.
836 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
838 PP/02 Harmonised TLS library version reporting; only show if debugging.
839 Layout now matches that introduced for other libraries in 4.74 PP/03.
841 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
843 PP/04 New "dns_use_edns0" global option.
845 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
848 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
849 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
851 TK/01 Updated PolarSSL code to 0.14.2.
852 Bugzilla 1097. Patch from Andreas Metzler.
854 PP/07 Catch divide-by-zero in ${eval:...}.
857 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
860 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
861 format-string attack -- SECURITY: remote arbitrary code execution.
863 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
864 time unintentionally subject to list matching rules, letting the header
865 cause arbitrary Exim lookups (of items which can occur in lists, *not*
866 arbitrary string expansion). This allowed for information disclosure.
868 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
869 INT_MIN/-1 -- value coerced to INT_MAX.
875 NM/01 Workround for PCRE version dependency in version reporting
878 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
879 This fixes portability to compilers other than gcc, notably
880 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
882 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
883 makefiles for portability to HP-UX and POSIX correctness.
885 PP/01 Permit LOOKUP_foo enabling on the make command-line.
886 Also via indented variable definition in the Makefile.
887 (Debugging by Oliver Heesakkers).
889 PP/02 Restore caching of spamd results with expanded spamd_address.
890 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
892 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
893 Improves build reliability. Fix from: Frank Elsner
895 NM/02 Fix wide character breakage in the rfc2047 coding
896 Fixes bug 1064. Patch from Andrey N. Oktyabrski
898 NM/03 Allow underscore in dnslist lookups
899 Fixes bug 1026. Patch from Graeme Fowler
901 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
902 Code patches from Adam Ciarcinski of NetBSD.
904 NM/04 Fixed exiqgrep to cope with mailq missing size issue
907 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
908 is logged, to avoid truncation. Patch from John Horne.
910 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
911 Patch from Jakob Hirsch.
913 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
914 SQL string expansion failure details.
915 Patch from Andrey Oktyabrski.
917 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
918 Patch from Simon Arlott.
920 PP/09 New lookups functionality failed to compile on old gcc which rejects
921 extern declarations in function scope.
922 Patch from Oliver Fleischmann
924 PP/10 Use sig_atomic_t for flags set from signal handlers.
925 Check getgroups() return and improve debugging.
926 Fixed developed for diagnosis in bug 927 (which turned out to be
929 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
930 Patch from Mark Zealey.
932 PP/12 Bugzilla 1056: Improved spamd server selection.
933 Patch from Mark Zealey.
935 PP/13 Bugzilla 1086: Deal with maildir quota file races.
936 Based on patch from Heiko Schlittermann.
938 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
939 Patch from Uwe Doering, sign-off by Michael Haardt.
941 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
942 variable declaration deep within a block. Bug and patch from
945 PP/15 lookups-Makefile IRIX compatibilty coercion.
947 PP/16 Make DISABLE_DKIM build knob functional.
949 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
950 Patch by Simon Arlott
952 TF/03 Fix valgrind.h portability to C89 compilers that do not support
953 variable argument macros. Our copy now differs from upstream.
959 TF/01 Failure to get a lock on a hints database can have serious
960 consequences so log it to the panic log.
962 TF/02 Log LMTP confirmation messages in the same way as SMTP,
963 controlled using the smtp_confirmation log selector.
965 TF/03 Include the error message when we fail to unlink a spool file.
967 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
968 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
969 for maintaining out-of-tree patches for some time.
971 PP/01 Bugzilla 139: Documentation and portability issues.
972 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
973 Handle per-OS dynamic-module compilation flags.
975 PP/02 Let /dev/null have normal permissions.
976 The 4.73 fixes were a little too stringent and complained about the
977 permissions on /dev/null. Exempt it from some checks.
978 Reported by Andreas M. Kirchwitz.
980 PP/03 Report version information for many libraries, including
981 Exim version information for dynamically loaded libraries. Created
982 version.h, now support a version extension string for distributors
983 who patch heavily. Dynamic module ABI change.
985 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
986 privilege escalation vulnerability whereby the Exim run-time user
987 can cause root to append content of the attacker's choosing to
990 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
993 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
994 If dropping privileges for untrusted macros, we disabled normal logging
995 on the basis that it would fail; for the Exim run-time user, this is not
996 the case, and it resulted in successful deliveries going unlogged.
997 Fixed. Reported by Andreas Metzler.
1003 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1004 only prepend for the Resent-* case. Fixes regression introduced in
1005 Exim 4.70 by NM/22 for Bugzilla 607.
1007 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1008 increasing numbers of administrators be bitten by this.
1010 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1012 PP/03 Bugzilla 994: added openssl_options main configuration option.
1014 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1016 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1018 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1020 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1021 without a peer certificate, leading to a segfault because of an
1022 assumption that peers always have certificates. Be a little more
1023 paranoid. Problem reported by Martin Tscholak.
1025 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1026 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1027 NB: ClamAV planning to remove STREAM in "middle of 2010".
1028 CL also introduces -bmalware, various -d+acl logging additions and
1029 more caution in buffer sizes.
1031 PP/09 Implemented reverse_ip expansion operator.
1033 PP/10 Bugzilla 937: provide a "debug" ACL control.
1035 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1037 PP/12 Bugzilla 973: Implement --version.
1039 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1041 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1043 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1045 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1046 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1047 ignore trailing whitespace.
1049 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1051 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1054 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1055 Notification from Dr Andrew Aitchison.
1057 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1058 ExtendedDetectionInfo response format.
1059 Notification from John Horne.
1061 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1064 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1065 XSL and documented dependency on system catalogs, with examples of how
1068 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1071 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1072 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1073 configuration file which is writeable by the Exim user or group.
1075 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1076 of configuration files to cover files specified with the -C option if
1077 they are going to be used with root privileges, not just the default
1080 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1081 option (effectively making it always true).
1083 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1084 files to be used while preserving root privileges.
1086 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1087 that rogue child processes cannot use them.
1089 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1090 run-time user, instead of root.
1092 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1093 Exim run-time user without dropping privileges.
1095 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1096 result string, instead of calling string_vformat() twice with the same
1099 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1100 for other users. Others should always drop root privileges if they use
1101 -C on the command line, even for a whitelisted configure file.
1103 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1105 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1111 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1112 $data_path, and $header_path variables; fixed documentation bugs and
1115 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1116 exipick to access non-standard spools, including the "frozen" queue
1119 NM/01 Bugzilla 965: Support mysql stored procedures.
1120 Patch from Alain Williams
1122 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1124 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1125 Patch from Andreas Metzler
1127 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1128 Patch from Kirill Miazine
1130 NM/05 Bugzilla 671: Added umask to procmail example.
1132 JJ/03 installed exipick 20100323.0, fixing doc bug
1134 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1135 directory. Notification and patch from Dan Rosenberg.
1137 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1139 TK/02 Improve log output when DKIM signing operation fails.
1141 MH/01 Treat the transport option dkim_domain as a colon separated
1142 list, not as a single string, and sign the message with each element,
1143 omitting multiple occurences of the same signer.
1145 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1146 Bugzilla 985, 986. Patch by Simon Arlott
1148 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1149 Patch by Simon Arlott
1151 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1152 MBX locking. Notification from Dan Rosenberg.
1158 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1160 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1162 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1164 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1166 NM/04 Bugzilla 915: Flag broken perl installation during build.
1172 TK/01 Added patch by Johannes Berg that expands the main option
1173 "spamd_address" if it starts with a dollar sign.
1175 TK/02 Write list of recipients to X-Envelope-Sender header when building
1176 the mbox-format spool file for content scanning (suggested by Jakob
1179 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1180 (http://www.dcc-servers.net/) support via dccifd. Activated by
1181 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1183 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1184 by Mark Daniel Reidel <mr@df.eu>.
1186 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1187 When building exim an external PCRE library is now needed -
1188 PCRE is a system library on the majority of modern systems.
1189 See entry on PCRE_LIBS in EDITME file.
1191 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1192 conversation. Added nologin parameter to request.
1193 Patch contributed by Kirill Miazine.
1195 TF/01 Do not log submission mode rewrites if they do not change the address.
1197 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1199 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1200 log files in place. Contributed by Roberto Lima.
1202 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1204 TF/03 Bugzilla 615: When checking the local_parts router precondition
1205 after a local_part_suffix or local_part_prefix option, Exim now
1206 does not use the address's named list lookup cache, since this
1207 contains cached lookups for the whole local part.
1209 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1210 Robert Millan. Documentation is in experimental-spec.txt.
1212 TF/04 Bugzilla 668: Fix parallel build (make -j).
1214 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1216 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1217 Patch provided by Jan Srzednicki.
1219 TF/05 Leading white space used to be stripped from $spam_report which
1220 wrecked the formatting. Now it is preserved.
1222 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1223 that they are available at delivery time.
1225 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1227 TF/08 TLS error reporting now respects the incoming_interface and
1228 incoming_port log selectors.
1230 TF/09 Produce a more useful error message if an SMTP transport's hosts
1231 setting expands to an empty string.
1233 NM/06 Bugzilla 744: EXPN did not work under TLS.
1234 Patch provided by Phil Pennock.
1236 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1237 Patch provided by Richard Godbee.
1239 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1240 acl_smtp_notquit, added index entry.
1242 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1243 Patch provided by Eugene Bujak.
1245 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1246 accept(). Patch provided by Maxim Dounin.
1248 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1249 Patch provided by Phil Pennock.
1251 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1253 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1254 Patch provided by Brad "anomie" Jorsch.
1256 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1257 Patch provided by Dean Brooks.
1259 TK/05 Add native DKIM support (does not depend on external libraries).
1261 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1262 Patch provided by Graeme Fowler.
1264 NM/16 Bugzilla 851: Documentation example syntax fix.
1266 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1268 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1271 NM/19 Bugzilla 745: TLS version reporting.
1272 Patch provided by Phil Pennock.
1274 NM/20 Bugzilla 167: bool: condition support.
1275 Patch provided by Phil Pennock.
1277 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1278 clients. Patch provided by Phil Pennock.
1280 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1281 Patch provided by Brad "anomie" Jorsch.
1283 NM/23 Bugzilla 687: Fix misparses in eximstats.
1284 Patch provided by Heiko Schlittermann.
1286 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1287 Patch provided by Heiko Schlittermann.
1289 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1290 plus update to original patch.
1292 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1294 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1295 Patch provided by David Brownlee.
1297 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1299 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1301 NM/30 Bugzilla 888: TLS documentation bugfixes.
1303 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1305 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1306 Unlike the original bugzilla I have changed all shell scripts in src tree.
1308 NM/33 Bugzilla 898: Transport filter timeout fix.
1309 Patch by Todd Rinaldo.
1311 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1312 Patch by Serge Demonchaux.
1314 NM/35 Bugzilla 39: Base64 decode bug fixes.
1315 Patch by Jakob Hirsch.
1317 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1319 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1321 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1323 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1329 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1330 ALT-N's libdkim that I have put here:
1331 http://duncanthrax.net/exim-experimental/
1333 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1334 were called 'true' and it seems that C99 defines that as a reserved
1335 keyword to be used with 'bool' variable types. That means you could
1336 not include C99-style headers which use bools without triggering
1337 build errors in sieve.c.
1339 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1340 as mailq or other aliases. Changed the --help handling significantly
1341 to do whats expected. exim_usage() emits usage/help information.
1343 SC/01 Added the -bylocaldomain option to eximstats.
1345 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1347 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1349 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1355 PH/01 Another patch from the Sieve maintainer.
1357 PH/02 When an IPv6 address is converted to a string for single-key lookup
1358 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1359 dots are used instead of colons so that keys in lsearch files need not
1360 contain colons. This was done some time before quoting was made available
1361 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1362 (notated using the quote facility) so as to distinguish them from IPv4
1363 keys. This meant that lookups for IP addresses in host lists did not work
1364 for iplsearch lookups.
1366 This has been fixed by arranging for IPv6 addresses to be expressed with
1367 colons if the lookup type is iplsearch. This is not incompatible, because
1368 previously such lookups could never work.
1370 The situation is now rather anomolous, since one *can* have colons in
1371 ordinary lsearch keys. However, making the change in all cases is
1372 incompatible and would probably break a number of configurations.
1374 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1377 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1378 conversion specification without a maximum field width, thereby enabling
1379 a rogue spamd server to cause a buffer overflow. While nobody in their
1380 right mind would setup Exim to query an untrusted spamd server, an
1381 attacker that gains access to a server running spamd could potentially
1382 exploit this vulnerability to run arbitrary code as the Exim user.
1384 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1385 $primary_hostname instead of what libspf2 thinks the hosts name is.
1387 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1388 a directory entry by the name of the lookup key. Previously, if a
1389 symlink pointed to a non-existing file or a file in a directory that
1390 Exim lacked permissions to read, a lookup for a key matching that
1391 symlink would fail. Now it is enough that a matching directory entry
1392 exists, symlink or not. (Bugzilla 503.)
1394 PH/03 The body_linecount and body_zerocount variables are now exported in the
1397 PH/04 Added the $dnslist_matched variable.
1399 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1400 This means they are set thereafter only if the connection becomes
1403 PH/06 Added the client_condition to authenticators so that some can be skipped
1404 by clients under certain conditions.
1406 PH/07 The error message for a badly-placed control=no_multiline_responses left
1407 "_responses" off the end of the name.
1409 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1411 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1412 (without spaces) instead of just copying the configuration text.
1414 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1416 PH/11 Added $max_received_linelength.
1418 PH/12 Added +ignore_defer and +include_defer to host lists.
1420 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1421 way in which PCRE > 7.0 is built.
1423 PH/14 Implemented queue_only_load_latch.
1425 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1426 MAIL command. The effect was to mangle the value on 64-bit systems.
1428 PH/16 Another patch from the Sieve maintainer.
1430 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1432 PH/18 If a system quota error occurred while trying to create the file for
1433 a maildir delivery, the message "Mailbox is full" was not appended to the
1434 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1435 only to a quota excession during the actual writing of the file.
1437 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1438 characters?) which causes problems in log lines. The DN values are now
1439 passed through string_printing() before being added to log lines.
1441 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1442 and InterBase are left for another time.)
1444 PH/21 Added message_body_newlines option.
1446 PH/22 Guard against possible overflow in moan_check_errorcopy().
1448 PH/23 POSIX allows open() to be a macro; guard against that.
1450 PH/24 If the recipient of an error message contained an @ in the local part
1451 (suitably quoted, of course), incorrect values were put in $domain and
1452 $local_part during the evaluation of errors_copy.
1458 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1459 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1462 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1463 issue a MAIL command.
1465 PH/02 In an ACL statement such as
1467 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1469 if a client was not listed at all, or was listed with a value other than
1470 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1471 the condition was not true (as it should be), so access was not denied.
1472 The bug was that the ! inversion was incorrectly passed on to the second
1473 item. This has been fixed.
1475 PH/03 Added additional dnslists conditions == and =& which are different from
1476 = and & when the dns lookup returns more than one IP address.
1478 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1479 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1481 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1482 FSYNC, which compiles an option called disable_fsync that allows for
1483 bypassing fsync(). The documentation is heavily laced with warnings.
1485 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1487 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1488 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1489 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1490 including adding "make clean"; (3) Added -fPIC when compiling the test
1491 dynamically loaded module, to get rid of a warning.
1493 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1494 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1495 = 0s. The bug is otherwise harmless.
1497 PH/07 There was a bug in the dovecot authenticator such that the value of
1498 $auth1 could be overwritten, and so not correctly preserved, after a
1499 successful authentication. This usually meant that the value preserved by
1500 the server_setid option was incorrect.
1502 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1504 PH/09 Installed PCRE release 7.0.
1506 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1507 run for batched SMTP input. It is now run at the start of every message
1508 in the batch. While fixing this I discovered that the process information
1509 (output by running exiwhat) was not always getting set for -bs and -bS
1510 input. This is fixed, and it now also says "batched" for BSMTP.
1512 PH/11 Added control=no_pipelining.
1514 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1515 patch, slightly modified), and move the expansion of helo_data till after
1516 the connection is made in the smtp transport (so it can use these
1519 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1521 PH/14 Added log_selector = +pid.
1523 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1525 PH/16 Add ${if forany and ${if forall.
1527 PH/17 Added dsn_from option to vary the From: line in DSNs.
1529 PH/18 Flush SMTP output before performing a callout, unless control =
1530 no_callout_flush is set.
1532 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1533 was true (the default) a successful delivery failed to delete the retry
1534 item, thus causing premature timeout of the address. The bug is now
1537 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1539 PH/21 Long custom messages for fakedefer and fakereject are now split up
1540 into multiline reponses in the same way that messages for "deny" and
1541 other ACL rejections are.
1543 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1544 with slight modification.
1546 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1547 draft, changing the syntax and factoring some duplicate code".
1549 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1550 for deliveries of the second and subsequent messages over the same SMTP
1553 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1554 ${reduce, with only minor "tidies".
1556 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1558 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1559 expansion side effects.
1561 PH/27 When a message times out after an over-quota error from an Exim-imposed
1562 quota, the bounce message says "mailbox is full". This message was not
1563 being given when it was a system quota that was exceeded. It now should
1566 MH/03 Made $recipients available in local_scan(). local_scan() already has
1567 better access to the recipient list through recipients_list[], but
1568 $recipients can be useful in postmaster-provided expansion strings.
1570 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1571 in the case of a MAIL command with additional options following the
1572 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1573 were accidentally chopped off.
1575 PH/29 SMTP synchronization checks are implemented when a command is read -
1576 there is a check that no more input is waiting when there shouldn't be
1577 any. However, for some commands, a delay in an ACL can mean that it is
1578 some time before the response is written. In this time, more input might
1579 arrive, invalidly. So now there are extra checks after an ACL has run for
1580 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1581 pipelining has not been advertised.
1583 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1585 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1586 returned in the SMTP response when smtp_return_error_details was false.
1587 This has been fixed.
1589 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1590 instead of the C I/O that was originally supplied, because problems were
1591 reported on Solaris.
1593 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1594 Exim which did not show up earlier: it was assuming that a call to
1595 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1596 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1597 was a macro that became an assignment, so it seemed to work. This has
1598 changed to a proper function call with a void return, hence the compile
1599 error. Exim's code has been fixed.
1601 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1604 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1606 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1608 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1609 because it turns out to be too big or there is a timeout) there is no
1610 "Completed" line in the log. When some messages of this type were
1611 selected by exigrep, they were listed as "not completed". Others were
1612 picked up by some special patterns. I have improved the selection
1613 criteria to be more general.
1615 PH/38 The host_find_failed option in the manualroute router can now be set
1616 to "ignore", to completely ignore a host whose IP address cannot be
1617 found. If all hosts are ignored, the behaviour is controlled by the new
1618 host_all_ignored option.
1620 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1621 homing or because of multiple MX records with /mx) generated more than
1622 one IP address, and the following item turned out to be the local host,
1623 all the secondary addresses of the first item were incorrectly removed
1624 from the list, along with the local host and any following hosts (which
1625 is what is supposed to happen).
1627 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1628 whoever called Exim into the -H file. In the case of the daemon it was
1629 behaving confusingly. When first started, it used values for whoever
1630 started the daemon, but after a SIGHUP it used the Exim user (because it
1631 calls itself on a restart). I have changed the code so that it now always
1634 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1635 message are rejected with the same error (e.g. no authentication or bad
1636 sender address), and a DATA command is nevertheless sent (as can happen
1637 with PIPELINING or a stupid MUA), the error message that was given to the
1638 RCPT commands is included in the rejection of the DATA command. This is
1639 intended to be helpful for MUAs that show only the final error to their
1642 PH/42 Another patch from the Sieve maintainer.
1644 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1645 Eximstats - Fixed some broken HTML links and added missing column headers
1647 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1648 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1650 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1652 PH/43 Yet another patch from the Sieve maintainer.
1654 PH/44 I found a way to check for a TCP/IP connection going away before sending
1655 the response to the final '.' that terminates a message, but only in the
1656 case where the client has not sent further data following the '.'
1657 (unfortunately, this is allowed). However, in many cases there won't be
1658 any further data because there won't be any more messages to send. A call
1659 to select() can be used: if it shows that the input is "ready", there is
1660 either input waiting, or the socket has been closed. An attempt to read
1661 the next input character can distinguish the two cases. Previously, Exim
1662 would have sent an OK response which the client would never have see.
1663 This could lead to message repetition. This fix should cure that, at
1664 least in a lot of common cases.
1666 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1667 advertised in response to EHLO.
1673 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1674 fixed by 4.65/MH/01 (is this a record?) are fixed:
1676 (i) An empty string was always treated as zero by the numeric comparison
1677 operators. This behaviour has been restored.
1679 (ii) It is documented that the numeric comparison operators always treat
1680 their arguments as decimal numbers. This was broken in that numbers
1681 starting with 0 were being interpreted as octal.
1683 While fixing these problems I realized that there was another issue that
1684 hadn't been noticed. Values of message_size_limit (both the global option
1685 and the transport option) were treated as octal if they started with 0.
1686 The documentation was vague. These values are now always treated as
1687 decimal, and I will make that clear in the documentation.
1693 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1694 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1697 MH/01 Don't check that the operands of numeric comparison operators are
1698 integers when their expansion is in "skipping" mode (fixes bug
1699 introduced by 4.64-PH/07).
1701 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1702 child addresses, Exim now panics and dies. Previously, because the count
1703 is held in a short int, deliveries were likely to be lost. As such a
1704 large number of recipients for a single message is ridiculous
1705 (performance will be very, very poor), I have chosen to impose a limit
1706 rather than extend the field.
1712 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1713 leftover -K file (the existence of which was triggered by #402).
1714 While we were at it, introduced process PID as part of the -K
1715 filename. This should rule out race conditions when creating
1718 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1719 processing considerably. Previous code took too long for large mails,
1720 triggering a timeout which in turn triggers #401.
1722 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1723 in the DK code in transports.c. sendfile() is not really portable,
1724 hence the _LINUX specificness.
1726 TF/01 In the add_headers option to the mail command in an Exim filter,
1727 there was a bug that Exim would claim a syntax error in any
1728 header after the first one which had an odd number of characters
1731 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1732 callout verification, Exim cached a "reject" for the entire domain. This
1733 is correct for most verifications, but it is not correct for a recipient
1734 verification with use_sender or use_postmaster set, because in that case
1735 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1736 case of MAIL FROM:<> rejection from other early rejections (e.g.
1737 rejection of HELO). When verifying a recipient using a non-null MAIL
1738 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1739 Whatever the result of the callout, the value of the domain cache is
1740 left unchanged (for any other kind of callout, getting as far as trying
1741 RCPT means that the domain itself is ok).
1743 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1746 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1747 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1748 session, the daemon ignores ECONNECT errors and logs others; it now
1749 ignores EPIPE as well.
1751 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1752 (quoted-printable decoding).
1754 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1755 later the small subsequent patch to fix an introduced bug.
1757 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1759 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1761 PH/08 An error is now given if message_size_limit is specified negative.
1763 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1764 to be given (somewhat) arbitrary names.
1766 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1769 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1770 miscellaneous code fixes
1772 PH/10 Added the log_reject_target ACL modifier to specify where to log
1775 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1776 hostname. This is wrong, because it relates to the incoming message (and
1777 probably the interface on which it is arriving) and not to the outgoing
1778 callout (which could be using a different interface). This has been
1779 changed to use the value of the helo_data option from the smtp transport
1780 instead - this is what is used when a message is actually being sent. If
1781 there is no remote transport (possible with a router that sets up host
1782 addresses), $smtp_active_hostname is used.
1784 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1785 tweaks were necessary in order to get it to work (see also 21 below):
1786 (a) The code assumed that strncpy() returns a negative number on buffer
1787 overflow, which isn't the case. Replaced with Exim's string_format()
1789 (b) There were several signed/unsigned issues. I just did the minimum
1790 hacking in of casts. There is scope for a larger refactoring.
1791 (c) The code used strcasecmp() which is not a standard C function.
1792 Replaced with Exim's strcmpic() function.
1793 (d) The code set only $1; it now sets $auth1 as well.
1794 (e) A simple test gave the error "authentication client didn't specify
1795 service in request". It would seem that Dovecot has changed its
1796 interface. Fortunately there's a specification; I followed it and
1797 changed what the client sends and it appears to be working now.
1799 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1802 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1803 address A is aliased to B and C, where B exists and C does not. Without
1804 -v the output is "A verified" because verification stops after a
1805 successful redirection if more than one address is generated. However,
1806 with -v the child addresses are also verified. Exim was outputting "A
1807 failed to verify" and then showing the successful verification for C,
1808 with its parentage. It now outputs "B failed to verify", showing B's
1809 parentage before showing the successful verification of C.
1811 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1812 look up a TXT record in a specific list after matching in a combined
1815 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1816 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1817 they consult the DNS. I had assumed they would set it the way they
1818 wanted; and indeed my experiments on Linux seem to show that in some
1819 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1820 To be on the safe side, however, I have now made the interface to
1821 host_find_byname() similar to host_find_bydns(), with an argument
1822 containing the DNS resolver options. The host_find_byname() function now
1823 sets these options at its start, just as host_find_bydns() does. The smtp
1824 transport options dns_qualify_single and dns_search_parents are passed to
1825 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1826 of host_find_byname() use the default settings of RES_DEFNAMES
1827 (qualify_single) but not RES_DNSRCH (search_parents).
1829 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1830 spool_read_header() do less string testing, by means of a preliminary
1831 switch on the second character of optional "-foo" lines. (This is
1832 overdue, caused by the large number of possibilities that now exist.
1833 Originally there were few.) While I was there, I also converted the
1834 str(n)cmp tests so they don't re-test the leading "-" and the first
1835 character, in the hope this might squeeze out yet more improvement.
1837 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1838 flag allowing group syntax was set by the header_syntax check but not
1839 turned off, possible causing trouble later; (2) The flag was not being
1840 set at all for the header_verify test, causing "group"-style headers to
1841 be rejected. I have now set it in this case, and also caused header_
1842 verify to ignore an empty address taken from a group. While doing this, I
1843 came across some other cases where the code for allowing group syntax
1844 while scanning a header line wasn't quite right (mostly, not resetting
1845 the flag correctly in the right place). These bugs could have caused
1846 trouble for malformed header lines. I hope it is now all correct.
1848 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1849 with the "reply" argument non-NULL. The code, however (which originally
1850 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1851 but it didn't always do it. This confused somebody who was copying the
1852 code for some other use. I have removed all the tests.
1854 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1855 feature that was used to support insecure browsers during the U.S. crypto
1856 embargo. It requires special client support, and Exim is probably the
1857 only MTA that supported it -- and would never use it because real RSA is
1858 always available. This code has been removed, because it had the bad
1859 effect of slowing Exim down by computing (never used) parameters for the
1860 RSA_EXPORT functionality.
1862 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1863 authenticator to fail if there's a tab character in the incoming data
1864 (there should never be unless someone is messing about, as it's supposed
1865 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1866 if the connection is using TLS or if the remote IP is the same as the
1867 local IP, and the "valid-client-cert option" if a client certificate has
1870 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1871 authenticators. This can be used for authorization after authentication
1872 succeeds. (In the case of plaintext, it servers for both authentication
1875 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1876 if any retry times were supplied.
1878 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1879 connection, where there is no client IP address to check. In this
1880 situation, the verify now always succeeds.
1882 PH/25 Applied John Jetmore's -Mset patch.
1884 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1886 PH/27 In a string expansion for a processed (not raw) header when multiple
1887 headers of the same name were present, leading whitespace was being
1888 removed from all of them, but trailing whitespace was being removed only
1889 from the last one. Now trailing whitespace is removed from each header
1890 before concatenation. Completely empty headers in a concatenation (as
1891 before) are ignored.
1893 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1894 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1896 PH/29 [Removed. This was a change that I later backed out, and forgot to
1897 correct the ChangeLog entry (that I had efficiently created) before
1898 committing the later change.]
1900 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1901 address errors (4xx response to RCPT) over the same connection as other
1902 messages routed to the same hosts. Such deliveries are always "forced",
1903 so retry times are not inspected. This resulted in far too many retries
1904 for the affected addresses. The effect occurred only when there were more
1905 hosts than the hosts_max_try setting in the smtp transport when it had
1906 the 4xx errors. Those hosts that it had tried were not added to the list
1907 of hosts for which the message was waiting, so if all were tried, there
1908 was no problem. Two fixes have been applied:
1910 (i) If there are any address or message errors in an SMTP delivery, none
1911 of the hosts (tried or untried) are now added to the list of hosts
1912 for which the message is waiting, so the message should not be a
1913 candidate for sending over the same connection that was used for a
1914 successful delivery of some other message. This seems entirely
1915 reasonable: after all the message is NOT "waiting for some host".
1916 This is so "obvious" that I'm not sure why it wasn't done
1917 previously. Hope I haven't missed anything, but it can't do any
1918 harm, as the worst effect is to miss an optimization.
1920 (ii) If, despite (i), such a delivery is accidentally attempted, the
1921 routing retry time is respected, so at least it doesn't keep
1922 hammering the server.
1924 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1925 in ${readsocket because some servers need this prod.
1927 PH/32 Added some extra debug output when updating a wait-xxx database.
1929 PH/33 The hint "could be header name not terminated by colon", which has been
1930 given for certain expansion errors for a long time, was not being given
1931 for the ${if def:h_colon_omitted{... case.
1933 PH/34 The spec says: "With one important exception, whenever a domain list is
1934 being scanned, $domain contains the subject domain." There was at least
1935 one case where this was not true.
1937 PH/35 The error "getsockname() failed: connection reset by peer" was being
1938 written to the panic log as well as the main log, but it isn't really
1939 panic-worthy as it just means the connection died rather early on. I have
1940 removed the panic log writing for the ECONNRESET error when getsockname()
1943 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1944 runs only) independently of the message's sender address. This meant
1945 that, if the 4xx error was in fact related to the sender, a different
1946 message to the same recipient with a different sender could confuse
1947 things. In particualar, this can happen when sending to a greylisting
1948 server, but other circumstances could also provoke similar problems.
1949 I have changed the default so that the retry time for these errors is now
1950 based a combination of the sender and recipient addresses. This change
1951 can be overridden by setting address_retry_include_sender=false in the
1954 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1955 remote server are returned as part of bounce messages. This was not
1956 happening for LMTP over a pipe (the lmtp transport), but now it is the
1957 same for both kinds of LMTP.
1959 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1960 in header lines that were in fact CNAMEs. This is no longer the case.
1962 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1963 and queue runs started by the daemon processed all messages. This has
1964 been fixed so that -R and -S can now usefully be given with -q<time>.
1966 PH/40 Import PCRE release 6.7 (fixes some bugs).
1968 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1970 PH/42 Give an error if -q is specified more than once.
1972 PH/43 Renamed the variables $interface_address and $interface_port as
1973 $received_ip_address and $received_port, to make it clear that these
1974 values apply to message reception, and not to the outgoing interface when
1975 a message is delivered. (The old names remain recognized, of course.)
1977 PH/44 There was no timeout on the connect() call when using a Unix domain
1978 socket in the ${readsocket expansion. There now is.
1980 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1981 be meaningful with "accept".
1983 SC/01 Eximstats V1.43
1984 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1986 SC/02 Eximstats V1.44
1987 Use a glob alias rather than an array ref in the generated
1988 parser. This improves both readability and performance.
1990 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1991 Collect SpamAssassin and rejection statistics.
1992 Don't display local sender or destination tables unless
1993 there is data to show.
1994 Added average volumes into the top table text output.
1996 SC/04 Eximstats V1.46
1997 Collect data on the number of addresses (recipients)
1998 as well as the number of messages.
2000 SC/05 Eximstats V1.47
2001 Added 'Message too big' to the list of mail rejection
2002 reasons (thanks to Marco Gaiarin).
2004 SC/06 Eximstats V1.48
2005 Mainlog lines which have GMT offsets and are too short to
2006 have a flag are now skipped.
2008 SC/07 Eximstats V1.49 (Alain Williams)
2009 Added the -emptyok flag.
2011 SC/08 Eximstats V1.50
2012 Fixes for obtaining the IP address from reject messages.
2014 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2015 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2016 whitesspace changes from 4.64-PH/27
2018 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2021 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2022 are found, allow negative numbers in numeric criteria)
2024 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2026 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2027 to match changes made in 4.64-PH/43
2029 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2031 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2032 "domain" is actually an IP address, and force a failure. This locks out
2033 those revolvers/nameservers that support "A-for-A" lookups, in
2034 contravention of the specifications.
2036 PH/48 When a host name was looked up from an IP address, and the subsequent
2037 forward lookup of the name timed out, the host name was left in
2038 $sender_host_name, contrary to the specification.
2040 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2041 restricted to single-key lookups, Exim was not diagnosing an error if
2042 * or *@ was used with a query-style lookup.
2044 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2046 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2047 long ago, but noone interested enough thought of it. Let's just say that
2048 the "1.1" means that there are some new functions that weren't there at
2049 some point in the past.
2051 PH/51 Error processing for expansion failure of helo_data from an smtp
2052 transport during callout processing was broken.
2054 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2055 tested/used via the -bh/-bhc/-bs options.
2057 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2058 bug, fixed in subsequent PCRE releases).
2060 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2061 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2063 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2069 SC/01 Use a glob alias rather than an array ref in eximstats generated
2070 parser. This improves both readability and performance.
2072 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2073 Don't display local sender or destination tables in eximstats unless
2074 there is data to show.
2075 Added average volumes into the eximstats top table text output.
2077 SC/03 Collect data on the number of addresses (recipients) as well
2078 as the number of messages in eximstats.
2080 TF/01 Correct an error in the documentation for the redirect router. Exim
2081 does (usually) call initgroups() when daemonizing.
2083 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2084 with consistent privilege compared to when running as a daemon.
2086 TF/03 Note in the spec that $authenticated_id is not set for local
2087 submissions from trusted users.
2089 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2090 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2092 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2093 by adding some example configuration directives to the default
2094 configuration file. A little bit of work is required to uncomment the
2095 directives and define how usernames and passwords are checked, but
2096 there is now a framework to start from.
2098 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2099 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2100 without this. I don't know how relevant this is to other LDAP libraries.
2102 PH/02 Add the verb name to the "unknown ACL verb" error.
2104 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2106 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2108 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2109 directory not expanded when it should be if an expanded home directory
2110 was set for the address (which is overridden by the transport).
2112 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2115 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2116 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2117 because it is too late at that time, and has no effect.
2119 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2120 security issue with \' (bugzilla #107). I could not use the
2121 PQescapeStringConn() function, because it needs a PGconn value as one of
2124 PH/08 When testing addresses using -bt, indicate those final addresses that
2125 are duplicates that would not cause an additional delivery. At least one
2126 person was confused, thinking that -bt output corresponded to deliveries.
2127 (Suppressing duplicates isn't a good idea as you lose the information
2128 about possibly different redirections that led to the duplicates.)
2130 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2131 systems where poll() doesn't work, in particular OS X.
2133 PH/10 Added more information to debugging output for retry time not reached.
2135 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2136 operations in malware.c.
2138 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2141 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2142 syslog_duplication set false, log lines that would normally be written
2143 both the the main log and to the reject log were not written to syslog at
2146 PH/14 In the default configuration, change the use of "message" in ACL warn
2147 statements to "add_header".
2149 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2150 not followed by a command (e.g. "seen endif").
2152 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2153 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2156 PH/17 Added extra conditions to the default value of delay_warning_condition
2160 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2161 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2162 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2165 The Auto-Submitted: and various List- headers are standardised, whereas I
2166 don't think Precedence: ever was.
2168 PH/18 Refactored debugging code in route_finduser() to show more information,
2169 in particular, the error code if getpwnam() issues one.
2171 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2172 This is apparently needed in addition to the PH/07 change above to avoid
2173 any possible encoding problems.
2175 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2176 but not after initializing Perl.
2178 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2179 output them only if debugging. By default they are written stderr,
2180 apparently, which is not desirable.
2182 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2185 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2188 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2190 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2191 authenticated or an ident call has been made. Suppress the default
2192 values for $authenticated_id and $authenticated_sender (but permit -oMai
2193 and -oMas) when testing with -bh.
2195 PH/24 Re-jigged the order of the tests in the default configuration so that the
2196 tests for valid domains and recipients precede the DNS black list and CSA
2197 tests, on the grounds that those ones are more expensive.
2199 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2200 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2201 This bug exists in every version of Exim that I still have, right back to
2204 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2205 However, an attempt to turn on case-sensitivity in a regex key by
2206 including (?-i) didn't work because the subject string was already
2207 lowercased, and the effects were non-intuitive. It turns out that a
2208 one-line patch can be used to allow (?-i) to work as expected.
2214 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2215 other effects) broke the use of negated acl sub-conditions.
2217 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2220 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2221 "Deny" causes Exim to reject the incoming connection with a 554 error.
2222 Unfortunately, if there is a major crisis, such as a disk failure,
2223 tcp-wrappers gives "deny", whereas what one would like would be some
2224 kind of temporary error. A kludge has been added to help with this.
2225 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2226 554 error is used if errno is still zero or contains ENOENT (which occurs
2227 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2230 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2232 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2233 errors. Otherwise a message that provokes a temporary error (when other
2234 messages do not) can cause a whole host to time out.
2236 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2237 addresses were routed directly to files or pipes from a redirect router.
2238 File deliveries just didn't batch; pipe deliveries might have suffered
2241 PH/06 A failure to get a lock for a hints database would erroneously always say
2242 "Failed to get write lock", even when it was really a read lock.
2244 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2245 of 0600. This has been changed to use the value of the lockfile_mode
2246 option (which defaults to 0600).
2248 PH/08 Applied small patch from the Sieve maintainer.
2250 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2251 folder from quota calculations, a direct delivery into this folder messed
2252 up the contents of the maildirsize file. This was because the regex was
2253 used only to exclude .Trash (or whatever) when the size of the mailbox
2254 was calculated. There was no check that a delivery was happening into an
2255 excluded directory. This bug has been fixed by ignoring all quota
2256 processing for deliveries into excluded directories.
2258 PH/10 Added the maildirfolder_create_regex option to appendfile.
2264 PH/01 The code for finding all the local interface addresses on a FreeBSD
2265 system running IPv6 was broken. This may well have applied to all BSD
2266 systems, as well as to others that have similar system calls. The broken
2267 code found IPv4 interfaces correctly, but gave incorrect values for the
2268 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2269 that it would not match correctly against @[] and not recognize the IPv6
2272 PH/02 The ipliteral router was not recognizing addresses of the form user@
2273 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2275 PH/03 Added disable_ipv6.
2277 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2278 decoded form, because it is most often used to construct To: headers
2279 lines in autoreplies, and the decoded form may well be syntactically
2280 invalid. However, $reply_address has leading white space removed, and all
2281 newlines turned into spaces so that the autoreply transport does not
2284 PH/05 If group was specified without a user on a router, and no group or user
2285 was specified on a transport, the group from the router was ignored.
2287 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2288 for visible compile-time settings that can be used to change these
2289 numbers, for those that want even more. Backwards compatibility with old
2290 spool files has been maintained. However, going back to a previous Exim
2291 release will lost any variables that are in spool files.
2293 PH/07 Two small changes when running in the test harness: increase delay when
2294 passing a TCP/IP connection to a new process, in case the original
2295 process has to generate a bounce, and remove special handling of
2296 127.0.0.2 (sic), which is no longer necessary.
2298 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2299 be the same on different OS.
2301 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2304 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2305 whether --show-vars was specified or not
2307 JJ/02 exipick: Added support for new ACL variable spool format introduced
2310 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2311 syntactically invalid From: or Reply-to: line, and a filter used this to
2312 generate an autoreply, and therefore failed to obtain an address for the
2313 autoreply, Exim could try to deliver to a non-existent relative file
2314 name, causing unrelated and misleading errors. What now happens is that
2315 it logs this as a hard delivery error, but does not attempt to create a
2318 PH/11 The exinext utility has a -C option for testing purposes, but although
2319 the given file was scanned by exinext itself; it wasn't being passed on
2320 when Exim was called.
2322 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2323 an end-of-file indication when reading a command response.
2325 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2326 compiled. In many other places in Exim, IPv6 addresses are always
2327 recognized, so I have changed this. It also means that IPv4 domain
2328 literals of the form [IPV4:n.n.n.n] are now always recognized.
2330 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2331 used if the router is not running as root, for example, when verifying at
2332 ACL time, or when using -bh. The debugging output from this situation was
2333 non-existent - all you got was a failure to exec. I have made two
2336 (a) Failures to set uid/gid, the current directory, or a process leader
2337 in a subprocess such as that created by queryprogram now generate
2338 suitable debugging ouput when -d is set.
2340 (b) The queryprogram router detects when it is not running as root,
2341 outputs suitable debugging information if -d is set, and then runs
2342 the subprocess without attempting to change uid/gid.
2344 PH/15 Minor change to Makefile for building test_host (undocumented testing
2347 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2348 additional section of a DNS packet that returns MX or SRV records.
2349 Instead, it always explicitly searches for A/AAAA records. This avoids
2350 major problems that occur when a DNS server includes only records of one
2351 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2352 fixed another bug: if SRV records were looked up and the corresponding
2353 address records were *not* found in the additional section, the port
2354 values from the SRV records were lost.
2356 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2357 using the correct key (the original address) when searching the retry
2358 rules in order to find which one to use for generating the retry hint.
2360 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2361 adding the default one. Similarly, if it contains a Reply-To: header, the
2362 errors_reply_to option, if set, is not used.
2364 PH/19 When calculating a retry time, Exim used to measure the "time since
2365 failure" by looking at the "first failed" field in the retry record. Now
2366 it does not use this if it is later than than the arrival time of the
2367 message. Instead it uses the arrival time. This makes for better
2368 behaviour in cases where some deliveries succeed, thus re-setting the
2369 "first failed" field. An example is a quota failure for a huge message
2370 when small messages continue to be delivered. Without this change, the
2371 "time since failure" will always be short, possible causing more frequent
2372 delivery attempts for the huge message than are intended.
2373 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2375 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2376 $1, $2, $3) because the numerical variables can be reset during some
2377 expansion items (e.g. "match"), thereby losing the authentication data.
2379 PH/21 Make -bV show the size of off_t variables so that the test suite can
2380 decide whether to run tests for quotas > 2G.
2382 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2383 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2384 filecount value is greater than 2G or if a quota value is greater than 2G
2385 on a system where the size of off_t is not greater than 4, a panic error
2388 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2389 never match. The debug and -bh output now contains an explicit error
2390 message indicating a malformed IPv4 address or mask.
2392 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2393 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2394 PH/23 above applies.
2396 PH/25 Do not write to syslog when running in the test harness. The only
2397 occasion when this arises is a failure to open the main or panic logs
2398 (for which there is an explicit test).
2400 PH/26 Added the /no_tell option to "control=freeze".
2402 PH/27 If a host name lookup failed very early in a connection, for example, if
2403 the IP address matched host_lookup and the reverse lookup yielded a name
2404 that did not have a forward lookup, an error message of the form "no IP
2405 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2406 could be logged. Now it outputs the IP address instead of "NULL".
2408 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2409 allows the sender and the authenticated sender to be set when
2410 submitting a message from within Exim. Since child_open_exim() is
2411 documented for local_scan(), the new function should be too.
2413 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2414 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2415 results in an empty string is now treated as unset.
2417 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2419 PH/31 Added sender_verify_fail logging option.
2421 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2422 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2423 tidied the source and removed it altogether.
2425 PH/33 When a queue run was abandoned because the load average was too high, a
2426 log line was always written; now it is written only if the queue_run log
2427 selector is set. In addition, the log line for abandonment now contains
2428 information about the queue run such as the pid. This is always present
2429 in "start" and "stop" lines but was omitted from the "abandon" line.
2431 PH/34 Omit spaces between a header name and the colon in the error message that
2432 is given when verify = headers_syntax fails (if there are lots of them,
2433 the message gets confusing).
2435 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2436 names, as there are now some PTR records that contain slashes. This check
2437 is only to protect against broken name servers that fall over on strange
2438 characters, so the fact that it applies to all lookups doesn't matter.
2440 PH/36 Now that the new test suite is complete, we can remove some of the
2441 special code in Exim that was needed for the old test suite. For example,
2442 sorting DNS records because real resolvers return them in an arbitrary
2443 order. The new test suite's fake resolver always returns records in the
2446 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2447 bounces) except when queue_only is set, to avoid logging races between
2448 the different processes.
2450 PH/38 Panic-die if .include specifies a non-absolute path.
2452 PH/39 A tweak to the "H" retry rule from its user.
2454 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2455 a label. They prevented compilation on older perls.
2457 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2458 a warning to be raised on newish perls.
2460 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2461 on queue. Changes to match documented behaviour of showing count of
2462 messages matching specified criteria.
2464 PH/40 Changed the default ident timeout from 30s to 5s.
2466 PH/41 Added support for the use of login_cap features, on those BSD systems
2467 that have them, for controlling the resources used by pipe deliveries.
2469 PH/42 The content-scanning code uses fopen() to create files in which to put
2470 message data. Previously it was not paying any attention to the mode of
2471 the files. Exim runs with umask(0) because the rest of the code creates
2472 files with open(), and sets the required mode explicitly. Thus, these
2473 files were ending up world-writeable. This was not a big issue, because,
2474 being within the spool directory, they were not world-accessible. I have
2475 created a function called modefopen, which takes an additional mode
2476 argument. It sets umask(777), creates the file, chmods it to the required
2477 mode, then resets the umask. All the relevant calls to fopen() in the
2478 content scanning code have been changed to use this function.
2480 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2481 to 24 hours. This avoids potential overflow problems when processing G
2482 and H retry rules. I suspect nobody ever tinkers with this value.
2484 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2486 PH/45 When the plaintext authenticator is running as a client, the server's
2487 challenges are checked to ensure they are valid base64 strings. By
2488 default, the authentication attempt is cancelled if an invalid string is
2489 received. Setting client_ignore_invalid_base64 true ignores these errors.
2490 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2491 they are received. Thus, the responses can be made to depend on the
2492 challenges. If an invalid string is ignored, an empty string is placed in
2495 PH/46 Messages that are created by the autoreply transport now contains a
2496 References: header, in accordance with RFCs 2822 and 3834.
2498 PH/47 Added authenticated_sender_force to the smtp transport.
2500 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2502 PH/49 Installed latest patch from the Sieve maintainer.
2504 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2505 was also set, the appendfile transport was unnecessarily scanning a
2506 directory of message files (e.g. for maildir delivery) to find the count
2507 of files (along with the size), even though it did not need this
2508 information. It now does the scan only if it needs to find either the
2509 size of the count of files.
2511 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2513 PH/52 Two bugs concerned with error handling when the smtp transport is
2516 (i) Exim was not creating retry information for temporary errors given
2517 for individual recipients after the DATA command when the smtp transport
2518 was used in LMTP mode. This meant that they could be retried too
2519 frequently, and not timed out correctly.
2521 (ii) Exim was setting the flag that allows error details to be returned
2522 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2523 recipients that were returned after the DATA command.
2525 PH/53 This is related to PH/52, but is more general: for any failing address,
2526 when detailed error information was permitted to be returned to the
2527 sender, but the error was temporary, then after the final timeout, only
2528 "retry timeout exceeded" was returned. Now it returns the full error as
2529 well as "retry timeout exceeded".
2531 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2532 do this, and (what is worse) MTAs that accept it.
2534 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2535 will now be deprecated.
2537 PH/56 New os.c-cygwin from the Cygwin maintainer.
2539 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2540 formats (previously only available in exim formats via -bpr, -bpru,
2541 and -bpra. Now also available in native and exiqgrep formats)
2543 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2544 with very large, slow to parse queues
2546 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2548 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2550 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2551 responses to authentication challenges, though it was showing the
2552 challenges; (ii) I've removed the CR characters from the debug output for
2555 PH/58 Allow for the insertion of a newline as well as a space when a string
2556 is turned into more than one encoded-word during RFC 2047 encoding. The
2557 Sieve code now uses this.
2559 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2560 data_4xx, lost_connection, tls_required.
2562 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2563 message was being sent as an SMTP response.
2565 PH/61 Add -l and -k options to exicyclog.
2567 PH/62 When verifying, if an address was redirected to one new address, so that
2568 verification continued, and the new address failed or deferred after
2569 having set something in $address_data, the value of $address_data was not
2570 passed back to the ACL. This was different to the case when no
2571 redirection occurred. The value is now passed back in both cases.
2573 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2574 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2575 use login_cap.h, so on its own it isn't the distinguishing feature. The
2576 new name refers directly to the setclassresources() function.
2578 PH/65 Added configuration files for NetBSD3.
2580 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2582 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2583 is preferred over IPv4.
2585 PH/68 The bounce_return_message and bounce_return_body options were not being
2586 honoured for bounces generated during the reception of non-SMTP messages.
2587 In particular, this applied to messages rejected by the ACL. This bug has
2588 been fixed. However, if bounce_return_message is true and bounce_return_
2589 body is false, the headers that are returned for a non-SMTP message
2590 include only those that have been read before the error was detected.
2591 (In the case of an ACL rejection, they have all been read.)
2593 PH/69 The HTML version of the specification is now built in a directory called
2594 spec_html instead of spec.html, because the latter looks like a path with
2595 a MIME-type, and this confuses some software.
2597 PH/70 Catch two compiler warnings in sieve.c.
2599 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2600 function verify_get_ident() calls ip_connect() to connect a socket, but
2601 if the "connect()" function timed out, ip_connect() used to close the
2602 socket. However, verify_get_ident() also closes the socket later, and in
2603 between Exim writes to the log, which may get opened at this point. When
2604 the socket was closed in ip_connect(), the log could get the same file
2605 descriptor number as the socket. This naturally causes chaos. The fix is
2606 not to close the socket in ip_connect(); the socket should be closed by
2607 the function that creates it. There was only one place in the code where
2608 this was missing, in the iplookup router, which I don't think anybody now
2609 uses, but I've fixed it anyway.
2611 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2612 well as to direct DNS lookups. Otherwise the handling of names in host
2613 lists is inconsistent and therefore confusing.
2619 PH/01 Two changes to the default runtime configuration:
2621 (1) Move the checks for relay_from_hosts and authenticated clients from
2622 after to before the (commented out) DNS black list checks.
2624 (2) Add control=submission to the relay_from_hosts and authenticated
2625 clients checks, on the grounds that messages accepted by these
2626 statements are most likely to be submissions.
2628 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2630 (1) Generate an error if the third argument for the ${prvs expansion is
2633 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2636 (3) Reset the variables that are obtained from the first argument of
2637 ${prvscheck and used in the second argument before leaving the code,
2638 because their memory is reclaimed, so using them afterwards may do
2641 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2642 one (it's much easier than Tom thought :-).
2644 (5) Because of (4), we can now allow for the use of $prvscheck_result
2645 inside the third argument.
2647 PH/03 For some reason, the default setting of PATH when running a command from
2648 a pipe transport was just "/usr/bin". I have changed it to
2651 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2652 anything to be listed in the output from -bV.
2654 PH/05 When a filter generated an autoreply, the entire To: header line was
2655 quoted in the delivery log line, like this:
2657 => >A.N.Other <ano@some.domain> <original@ddress> ...
2659 This has been changed so that it extracts the operative address. There
2660 may be more than one such address. If so, they are comma-separated, like
2663 => >ano@some.domain,ona@other.domain <original@ddress> ...
2665 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2666 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2667 not being looked up in the rDNS to get a host name, Exim was showing the
2668 IP address twice in Received: lines, even though the IP addresses were
2669 identical. For example:
2671 Received: from [1.2.3.4] (helo=[1.2.3.4])
2673 However, if the real host name was known, it was omitting the HELO data
2674 if it matched the actual IP address. This has been tidied up so that it
2675 doesn't show the same IP address twice.
2677 PH/07 When both +timestamp and +memory debugging was on, the value given by
2678 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2679 called by the debug printing, thereby overwriting the timestamp buffer.
2680 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2682 PH/08 When the original message was included in an autoreply transport, it
2683 always said "this is a copy of the message, including all the headers",
2684 even if body_only or headers_only was set. It now gives an appropriate
2687 PH/09 Applied a patch from the Sieve maintainer which:
2689 o fixes some comments
2690 o adds the (disabled) notify extension core
2691 o adds some debug output for the result of if/elsif tests
2692 o points to the current vacation draft in the documentation
2693 and documents the missing references header update
2697 o fixes a bug in processing the envelope test (when testing
2698 multiple envelope elements, the last element determinted the
2701 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2702 Electronic Mail") by including:
2704 Auto-submitted: auto-generated
2706 in the messages that it generates (bounce messages and others, such as
2707 warnings). In the case of bounce messages for non-SMTP mesages, there was
2708 also a typo: it was using "Auto_submitted" (underscore instead of
2709 hyphen). Since every message generated by Exim is necessarily in response
2710 to another message, thes have all been changed to:
2712 Auto-Submitted: auto-replied
2714 in accordance with these statements in the RFC:
2716 The auto-replied keyword:
2718 - SHOULD be used on messages sent in direct response to another
2719 message by an automatic process,
2721 - MUST NOT be used on manually-generated messages,
2723 - MAY be used on Delivery Status Notifications (DSNs) and Message
2724 Disposition Notifications (MDNs),
2726 - MUST NOT be used on messages generated by automatic or periodic
2727 processes, except for messages which are automatic responses to
2730 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2731 to the default Received: header definition.
2733 PH/12 Added log selector acl_warn_skipped (default on).
2735 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2736 variables because (a) they are in the wrong storage pool and (b) even if
2737 they were copied, it wouldn't work properly because of the caching.
2739 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2740 checking when decoding. Apparently there are clients that generate
2741 overlong encoded strings. Why am I not surprised?
2743 PH/15 If the first argument of "${if match_address" was not empty, but did not
2744 contain an "@" character, Exim crashed. Now it writes a panic log message
2745 and treats the condition as false.
2747 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2749 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2750 extension "envelope-auth". The code is finished and in agreement with
2751 other implementations, but there is no documentation so far and in fact,
2752 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2753 not changing the active code.
2755 Print executed "if" and "elsif" statements when debugging is used. This
2756 helps a great deal to understand what a filter does.
2758 Document more things not specified clearly in RFC3028. I had all this
2759 sorted out, when out of a sudden new issues came to my mind. Oops."
2761 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2764 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2765 canonical form (as documented). However, after a host name lookup from
2766 the IP address, check_host() was doing a simple string comparison with
2767 addresses acquired from the DNS when checking that the found name did
2768 have the original IP as one of its addresses. Since any found IPv6
2769 addresses are likely to be in abbreviated form, the comparison could
2770 fail. Luckily, there already exists a function for doing the comparison
2771 by converting both addresses to binary, so now that is used instead of
2772 the text comparison.
2774 PH/20 There was another similar case to PH/19, when a complete host name was
2775 given in a host list; looking up its IP address could give an abbreviated
2776 form, whereas the current host's name might or might not be abbreviated.
2777 The same fix has been applied.
2783 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2784 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2787 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2788 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2790 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2792 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2793 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2794 Also, for "Auto-Submitted", treat anything other than "no" as
2795 non-personal, in accordance with RFC 3834. (Previously it treated
2796 anything starting "auto-" as non-personal.)
2798 TF/01 The control=submission/name=... option had a problem with syntax
2799 errors if the name included a slash character. The /name= option
2800 now slurps the rest of the string, so it can include any characters
2801 but it must come last in the list of options (after /sender_retain
2804 PH/05 Some modifications to the interface to the fake nameserver for the new
2812 TK/01 Added the "success_on_redirect" address verification option. See
2813 NewStuff for rationale and an example.
2815 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2817 PH/02 Patch to exigrep to allow it to work on syslog lines.
2819 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2820 fread() to skip over the body file's header line, because in Cygwin the
2821 header line is locked and is inaccessible.
2823 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2824 co-exist for some time) to make it clear that it is the Exim ID that is
2825 referenced, not the Message-ID: header line.
2827 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2828 string_format() function, because snprintf() does not exist on all
2831 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2832 ${stat: expansion item.
2834 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2835 protocol synchronization error", to keep the pedants happy.
2837 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2838 well as for IRIX systems, when gcc is being used. See the host.c source
2841 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2843 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2846 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2847 transport and to the smtp transport in LMTP mode.
2849 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2851 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2852 run a filter in a subprocess. This could lead to confusion in subsequent
2853 lookups in the parent process. There should also be a search_tidyup() at
2854 the end of the subprocess.
2856 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2857 only if the host matched helo_try_verify_hosts, which caused the
2858 verification to occur when the EHLO/HELO command was issued. The ACL just
2859 tested the remembered result. Now, if a previous verification attempt has
2860 not happened, "verify = helo" does it there and then.
2862 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2864 TK/03 Fix log output including CR from clamd.
2866 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2867 exist provoked a memory error which could cause a segfault.
2869 PH/15 Installed PCRE 6.2
2871 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2873 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2874 of the problem. Specifically, suggested +O2 rather than +O1 for the
2877 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2879 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2880 record, the port was not correctly logged when the outgoing_port log
2881 selector was set (it logged the transort's default port).
2883 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2884 fallback_hosts, and "hosts" in the smtp transport.
2886 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2887 host errors such as "Connection refused".
2889 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2890 authentication with radiusclient 0.4.9:
2892 - Error returned from rc_read_config was caught wrongly
2893 - Username/password not passed on to radius server due to wrong length.
2895 The presumption is that some radiusclient API changes for 4.51/PH/17
2896 were not taken care of correctly. The code is still untested by me (my
2897 Linux distribution still has 0.3.2 of radiusclient), but it was
2898 contributed by a Radius user.
2900 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2901 expanding the "port" option of the smtp transport.
2903 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2904 while reading a MIME header. Thanks to Tom Hughes for a patch.
2906 PH/24 Include config.h inside local_scan.h so that configuration settings are
2909 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2910 that in an ACL for RCPT (for example), you can examine exactly what was
2913 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2914 commands, but it was not correctly comparing the address with the actual
2915 client host address. Thus, it would show the EHLO address in Received:
2916 header lines when this was not necessary.
2918 PH/27 Added the % operator to ${eval:}.
2920 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2921 it should be ignoring failures (because with -C, for example, it has lost
2922 privilege). It wasn't ignoring creation failures other than "already
2925 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2928 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2929 because an input line was too long, either on its own, or by virtue of
2930 too many continuations, the temporary file was not being removed, and the
2931 return code was incorrect.
2933 PH/31 Missing "BOOL" in function definition in filtertest.c.
2935 PH/32 Applied Sieve patches from the maintainer.
2937 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2939 PH/33 Added "verify = not_blind".
2941 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2942 Local/Makefile (with some defaults set). These are used in built scripts
2943 such as exicyclog, but they have never been used in the exim_install
2944 script (though there are many overriding facilities there). I have
2945 arranged that the exim_install script now takes note of these two
2948 PH/35 Installed configuration files for Dragonfly.
2950 PH/36 When a locally submitted message by a trusted user did not contain a
2951 From: header, and the sender address was obtained from -f or from an SMTP
2952 MAIL command, and the trusted user did not use -F to supply a sender
2953 name, $originator_name was incorrectly used when constructing a From:
2954 header. Furthermore, $originator_name was used for submission mode
2955 messages from external hosts without From: headers in a similar way,
2956 which is clearly wrong.
2958 PH/37 Added control=suppress_local_fixups.
2960 PH/38 When log_selector = +received_sender was set, and the addition of the
2961 sender made the log line's construction buffer exactly full, or one byte
2962 less than full, an overflow happened when the terminating "\n" was
2965 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2966 when the result of a list match is failure because a DNS lookup failed.
2968 PH/40 RM_COMMAND is now used in the building process.
2970 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2971 the "build-* directories that it finds.
2973 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2974 address in a domain literal was a prefix of an interface address.
2976 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2977 when verifying a sender address, unless rewrite_headers is false.
2979 PH/44 Wrote a long comment about why errors_to addresses are verified as
2980 recipients, not senders.
2982 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2983 the ratelimit ACL was added.
2985 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2987 PH/46 Added extra information about PostgreSQL errors to the error string.
2989 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2990 avoiding the need to install special zones in a real server. This is
2991 backwards compatible; if it can't find the fake resolver, it drops back.
2992 Thus, both old and new test suites can be run.
2994 TF/02 Added util/ratelimit.pl
2996 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2997 clock is set back in time.
2999 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3000 Candler <B.Candler@pobox.com>.
3002 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3003 OK for addresses that are the result of redirections.
3005 PH/48 A number of further additions for the benefit of the new test suite,
3006 including a fake gethostbyname() that interfaces to the fake DNS resolver
3009 TF/06 The fix for widen_domains has also been applied to qualify_single and
3010 search_parents which are the other dnslookup options that can cause
3013 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3016 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3018 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3019 service" if it encountered an MX record with an empty target hostname.
3020 The message is now "an MX or SRV record indicated no SMTP service".
3022 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3023 if the DNS of the sending site is misconfigured. This is quite a
3024 common situation. This change restores the behaviour of treating a
3025 helo verification defer as a failure.
3027 PH/51 If self=fail was set on a router, the bounce message did not include the
3028 actual error message.
3034 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3036 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3037 command itself ended in error, the underlying message about the transport
3038 filter timeout was being overwritten with the pipe command error. Now the
3039 underlying error message should be appended to the second error message.
3041 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3043 PH/02 Reduce the amount of output that "make" produces by default. Full output
3044 can still be requested.
3046 PH/03 The warning log line about a condition test deferring for a "warn" verb
3047 was being output only once per connection, rather than after each
3048 occurrence (because it was using the same function as for successful
3049 "warn" verbs). This seems wrong, so I have changed it.
3051 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3052 it should not have, which might have caused a crash in the right
3053 circumstances, but probably never did.
3055 PH/04 Installed a modified version of Tony Finch's patch to make submission
3056 mode fix the return path as well as the Sender: header line, and to
3057 add a /name= option so that you can make the user's friendly name appear
3060 TF/03 Added the control = fakedefer ACL modifier.
3062 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3063 Mark Lowes for thorough testing.
3065 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3067 TK/03 Merged latest SRS patch from Miles Wilton.
3069 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3070 with the definition in sysexits.h (which is #included earlier).
3071 Fortunately, Exim does not actually use EX_OK. The code used to try to
3072 preserve the sysexits.h value, by assumimg that macro definitions were
3073 scanned for macro replacements. I have been disabused of this notion,
3074 so now the code just undefines EX_OK before #including unistd.h.
3076 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3077 in the smtp transport. When a block could not be written in a single
3078 write() function, the timeout was being re-applied to each part-write.
3079 This seems wrong - if the receiver was accepting one byte at a time it
3080 would take for ever. The timeout is now adjusted when this happens. It
3081 doesn't have to be particularly precise.
3083 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3084 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3086 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3087 without a domain if the check to <postmaster@domain> fails.
3089 SC/01 Eximstats: added -xls and the ability to specify output files
3090 (patch written by Frank Heydlauf).
3092 SC/02 Eximstats: use FileHandles for outputing results.
3094 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3096 SC/04 Eximstats: fixed display of large numbers with -nvr option
3098 SC/05 Eximstats: fixed merging of reports with empty tables.
3100 SC/06 Eximstats: added the -include_original_destination flag
3102 SC/07 Eximstats: removed tabs and trailing whitespace.
3104 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3106 TK/06 MBOX spool code: Add real "From " MBOX separator line
3107 so the .eml file is really in mbox format (even though
3108 most programs do not really care). Patch from Alex Miller.
3110 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3111 The latter is generated from $received_to and is only set if the
3112 message has one envelope recipient. SA can use these headers,
3113 obviously out-of-the-box. Patch from Alex Miller.
3115 PH/08 The ${def test on a variable was returning false if the variable's
3116 value was "0", contrary to what the specification has always said!
3117 The result should be true unless the variable is empty.
3119 PH/09 The syntax error of a character other than { following "${if
3120 def:variable_name" (after optional whitespace) was not being diagnosed.
3121 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3122 accidental colon was present, for example, could give incorrect results.
3124 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3125 result is used (not including appendfile, where other changes are about
3128 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3129 This involved changing a lot of size variables from int to off_t. It
3130 should work with maildirs and everything.
3132 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3133 spamd dying while we are connected to it.
3135 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3138 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3139 function for BDB 4.3.
3141 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3143 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3144 the sources for PCRE have been re-arranged and more files are now
3147 PH/15 The code I had for printing potentially long long variables in PH/11
3148 above was not the best (it lost precision). The length of off_t variables
3149 is now inspected at build time, and an appropriate printing format (%ld
3150 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3151 to be "long long int" or "long int". This is needed for the internal
3152 formatting function string_vformat().
3154 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3155 the configuration file to be ":syslog", then the script "guesses" where
3156 the logs files are, rather than using the compiled in default. In our
3157 case the guess is not the same as the compiled default, so the script
3158 suddenly stopped working when I started to use syslog. The patch checks
3159 to see if log_file_path is "". If so, it attempts to read it from exim
3160 with no configuration file to get the compiled in version, before it
3161 falls back to the previous guessing code."
3163 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3164 implementing BATV in an Exim configuration. See NewStuff for the gory
3167 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3168 Makefile that are specific to HP-UX.
3170 PH/18 If the "use_postmaster" option was set for a recipient callout together
3171 with the "random" option, the postmaster address was used as the MAIL
3172 FROM address for the random test, but not for the subsequent recipient
3173 test. It is now used for both.
3175 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3176 patch removes a few documentation additions to RFC 3028, because the
3177 latest draft now contains them. It adds the new en;ascii-case comparator
3178 and a new error check for 8bit text in MIME parts. Comparator and
3179 require names are now matched exactly. I enabled the subaddress
3180 extension, but it is not well tested yet (read: it works for me)."
3182 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3183 rework some of the code of TK/09 above to avoid the hardwired use of
3184 "%lld" and "long long". Replaced the call to snprintf() with a call to
3187 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3188 records point to non-existent hosts", "retry timeout exceeded", and
3189 "retry time not reached for any host after a long failure period".
3191 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3192 experimental DomainKeys support:
3194 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3195 (2) On an error such as an illegally used "control", the wrong name for
3196 the control was given.
3198 These problems did NOT occur unless DomainKeys support was compiled.
3200 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3202 PH/24 Added ${if match_ip condition.
3204 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3205 databases so that it will be absolutely obvious if a crash occurs in the
3206 DB library. This is a regular occurrence (often caused by mis-matched
3209 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3210 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3211 picked up on a user's system that detects such things. There doesn't seem
3212 to be a gcc warning option for this - only an attribute that has to be
3213 put on the function's prototype. It seems that in Fedora Core 4 they have
3214 set this on a number of new functions. No doubt there will be more in due
3217 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3218 specify a transport. However, if an address that was verified by such a
3219 router was the subject of a callout, Exim crashed because it tried to
3220 read the rcpt_include_affixes from the non-existent transport. Now it
3221 just assumes that the setting of that option is false. This bug was
3222 introduced by 4.51/PH/31.
3224 PH/28 Changed -d+all to exclude +memory, because that information is very
3225 rarely of interest, but it makes the output a lot bigger. People tend to
3226 do -d+all out of habit.
3228 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3229 code in os-type was giving problems when libc.so lives in lib64, like on
3232 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3233 aren't the modern standard, and it seems that some systems' include files
3234 don't always have them. Exim was already checking for some of the newer
3235 ones like T_AAAA, and defining it itself. I've added checks for all the
3236 record types that Exim uses.
3238 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3239 not automatically generating a new one, as it is supposed to. This
3240 prevented TLS from working. If the file did exist, but contained invalid
3241 data, a new version was generated, as expected. It was only the case of a
3242 non-existent file that was broken.
3244 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3245 with a change in libdomainkeys > 0.64.
3247 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3248 from DNS. If the selector record carries the flag, it now has
3249 precedence over the domain-wide flag.
3251 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3253 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3254 the use of a port name that isn't defined in /etc/services) occurred, the
3255 message was deferred as in a normal delivery, and thus remained on the
3256 spool, instead of being failed because of the mua_wrapper setting. This
3257 is now fixed, and I tidied up some of the mua_wrapper messages at the
3260 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3261 the messages in a hash of arrays rather than using individual hashes.
3262 This is a bit cleaner and results in dramatic memory savings, albeit
3263 at a slight CPU cost.
3265 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3266 as requested by Marc Sherman.
3268 SC/10 Eximstats: added histograms for user specified patterns as requested
3271 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3273 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3274 fopen() in the content-scanning modules that did not already have it.
3280 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3281 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3283 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3285 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3287 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3288 to test Sieve filters that use "vacation".
3290 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3291 that changes the way the GnuTLS parameters are stored in the cache file.
3292 The new format can be generated externally. For backward compatibility,
3293 if the data in the cache doesn't make sense, Exim assumes it has read an
3294 old-format file, and it generates new data and writes a new file. This
3295 means that you can't go back to an older release without removing the
3298 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3299 work if there are any delivery delays because "one_time" forces the
3300 parent to be marked "delivered", so its unseen clone is never tried
3301 again. For this reason, Exim now forbids the simultaneous setting of
3304 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3305 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3306 entry if you want to know the details. The fix, however, neglected to
3307 consider the case when local delivery batching is involved. The test for
3308 "previously delivered" was not happening when checking to see if an
3309 address could be batched with a previous (undelivered) one; under
3310 certain circumstances this could lead to multiple deliveries to the same
3313 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3314 in its include files, and this causes problems building Exim.
3316 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3317 header_syntax) but Exim was just ignoring anything given after a slash.
3318 In particular, this caused confusion with an attempt to use "verify =
3319 reverse_host_lookup/defer_ok". An error is now given when options are
3320 supplied for verify items that do not have them. (Maybe reverse_host_
3321 lookup should have a defer_ok option, but that's a different point.)
3323 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3324 defined by RFC 821) to 2048, because there were problems with some AUTH
3325 commands, and RFC 1869 says the size should be increased for extended
3326 SMTP commands that take arguments.
3328 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3331 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3332 "unknown" error; now it says that the functionality isn't in the binary.
3334 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3335 an address' error message when a string expansion fails (syntax or
3336 whatever). Otherwise the password may appear in the log. Following change
3337 PH/42 below, there is no longer a chance of it appearing in a bounce
3340 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3342 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3343 of its addresses was ever tried. (Bugzilla bug #2.)
3345 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3346 the result incorrectly in the debug output. (It correctly added a newline
3347 to what was transported.)
3349 TF/01 Added $received_time.
3351 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3352 commented out examples of how to interface to a virus scanner and to
3353 SpamAssassin. Also added commented examples of av_scanner and
3354 spamd_address settings.
3356 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3357 and controls are allowed in which ACLs. There were a couple of minor
3358 errors. Some of the entries in the conditions table (which is a table of
3359 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3360 negation of where the condition IS allowed.
3362 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3364 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3365 header file does not have a version number, so I've had to invent a new
3366 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3367 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3368 radiusclient), but it was contributed by a Radius user.
3370 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3371 files or directories, for OpenSSL.
3373 PH/19 When an Exim process that is running as root has to create an Exim log
3374 file, it does so in a subprocess that runs as exim:exim so as to get the
3375 ownership right at creation (otherwise, other Exim processes might see
3376 the file with the wrong ownership). There was no test for failure of this
3377 fork() call, which would lead to the process getting stuck as it waited
3378 for a non-existent subprocess. Forks do occasionally fail when resources
3379 run out. I reviewed all the other calls to fork(); they all seem to check
3382 PH/20 When checking for unexpected SMTP input at connect time (before writing
3383 the banner), Exim was not dealing correctly with a non-positive return
3384 from the read() function. If the client had disconnected by this time,
3385 the result was a log entry for a synchronization error with an empty
3386 string after "input=" when read() returned zero. If read() returned -1
3387 (an event I could not check), uninitialized data bytes were printed.
3388 There were reports of junk text (parts of files, etc) appearing after
3391 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3393 PH/22 Added support for macro redefinition, and (re)definition in between
3394 driver and ACL definitions.
3396 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3397 forgetting to use the resulting value; it was using the unexpanded value.
3399 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3400 hadn't been configured. The fix is from Juergen Kreileder, who
3401 understands it better than I do:
3403 "Here's what I see happening with three configured cyrus_sasl
3404 authenticators configured (plain, login, cram-md5):
3406 On startup auth_cyrus_sasl_init() gets called for each of these.
3407 This means three calls to sasl_listmech() without a specified mech_list.
3408 => SASL tests which mechs of all available mechs actually work
3409 => three warnings about OTP not working
3410 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3412 With the patch, sasl_listmech() also gets called three times. But now
3413 SASL's mech_list option is set to the server_mech specified in the the
3414 authenticator. Or in other words, the answer from sasl_listmech()
3415 gets limited to just the mech you're testing for (which is different
3417 => the return list contains just 'plain' or 'login', 'cram-md5' or
3418 nothing depending on the value of ob->server_mech.
3420 I've just tested the patch: Authentication still works fine,
3421 unavailable mechs specified in the exim configuration are still
3422 caught, and the auth.log warnings about OTP are gone."
3424 PH/25 When debugging is enabled, the contents of the command line are added
3425 to the debugging output, even when log_selector=+arguments is not
3428 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3429 answer is "GNU", and only if the return is "GNU/something" is the answer
3432 PH/27 $acl_verify_message is now set immediately after the failure of a
3433 verification in an ACL, and so is available in subsequent modifiers. In
3434 particular, the message can be preserved by coding like this:
3436 warn !verify = sender
3437 set acl_m0 = $acl_verify_message
3439 Previously, $acl_verify_message was set only while expanding "message"
3440 and "log_message" when a very denied access.
3442 PH/28 Modified OS/os.c-Linux with
3444 -#ifndef OS_LOAD_AVERAGE
3445 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3447 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3448 nomenclature these days.)
3450 PH/29 Installed patch from the Sieve maintainer that adds the options
3451 sieve_useraddress and sieve_subaddress to the redirect router.
3453 PH/30 In these circumstances:
3454 . Two addresses routed to the same list of hosts;
3455 . First host does not offer TLS;
3456 . First host accepts first address;
3457 . First host gives temporary error to second address;
3458 . Second host offers TLS and a TLS session is established;
3459 . Second host accepts second address.
3460 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3461 and peerdn, if requested) that were in fact used only for the second
3464 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3465 attention to any local part prefix or suffix that was matched by the
3466 router that accepted the address. It now behaves in the same way as it
3467 does for delivery: the affixes are removed from the local part unless
3468 rcpt_include_affixes is set on the transport.
3470 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3471 timeout during the DATA phase of an incoming message.
3473 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3474 applied a patch sanctioned by the Sieve maintainer.
3476 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3477 the uid or gid is negative. A case of a negative gid caused this to be
3478 noticed. The fix allows for either to be negative.
3480 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3481 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3484 PH/36 The change PH/12 above was broken. Fixed it.
3486 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3487 the grounds that routing the same address twice would always produce the
3488 same answer. This might have been true once, but it is certainly no
3489 longer true now. Routing a child address may depend on the previous
3490 routing that produced that child. Some complicated redirection strategies
3491 went wrong when messages had multiple recipients, and made Exim's
3492 behaviour dependent on the order in which the addresses were given.
3494 I have moved the duplicate checking until after the routing is complete.
3495 Exim scans the addresses that are assigned to local and remote
3496 transports, and removes any duplicates. This means that more work will be
3497 done, as duplicates will always all be routed, but duplicates are
3498 presumably rare, so I don't expect this is of any significance.
3500 For deliveries to pipes, files, and autoreplies, the duplicate checking
3501 still happens during the routing process, since they are not going to be
3504 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3505 It corrects a timeout issue with spamd. This is Ian's comment: "The
3506 background is that sometimes spamd either never reads data from a
3507 connection it has accepted, or it never writes response data. The exiscan
3508 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3509 blindly assumes that writes won't block so it may never time out."
3511 PH/39 Allow G after quota size as well as K and M.
3513 PH/40 The value set for $authenticated_id in an authenticator may not contain
3514 binary zeroes or newlines because the value is written to log lines and
3515 to spool files. There was no check on this. Now the value is run through
3516 the string_printing() function so that such characters are converted to
3517 printable escape sequences.
3519 PH/41 $message_linecount is a new variable that contains the total number of
3520 lines in the message. Compare $body_linecount, which is the count for the
3523 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3524 bounce and delay warning messages, except in certain special cases, which
3527 (a) An SMTP error message from a remote host;
3528 (b) A message specified in a :fail: redirection;
3529 (c) A message specified in a "fail" command in a system filter;
3530 (d) A message specified in a FAIL return from the queryprogram router;
3531 (e) A message specified by the cannot_route_message router option.
3533 In these cases only, Exim does include the error details in bounce and
3534 warning messages. There are also a few cases where bland messages such
3535 as "unrouteable address" or "local delivery error" are given.
3537 PH/43 $value is now also set for the "else" part of a ${run expansion.
3539 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3540 being worked on, but at least Exim now implements the latest version to
3543 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3544 process to complete was treated as a delivery failure, a timeout while
3545 writing the message to the pipe was logged, but erroneously treated as a
3546 successful delivery. Such timeouts include transport filter timeouts. For
3547 consistency with the overall process timeout, these timeouts are now
3548 treated as errors, giving rise to delivery failures by default. However,
3549 there is now a new Boolean option for the pipe transport called
3550 timeout_defer, which, if set TRUE, converts the failures into defers for
3551 both kinds of timeout. A transport filter timeout is now identified in
3554 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3555 systems where "make" and "gmake" are different, calling "gmake" at top
3556 level broke things. I've arranged for the value of $(MAKE) to be passed
3557 from the Makefile to this script so that it can call the same version of
3561 A note about Exim versions 4.44 and 4.50
3562 ----------------------------------------
3564 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3565 changes of various kinds. As a consequence, a big documentation update was
3566 needed. This delayed the release for rather longer than seemed good, especially
3567 in the light of a couple of (minor) security issues. Therefore, the changes
3568 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3569 release. So 4.44 and 4.50 are in effect two different branches that both start
3572 I have left the 4.50 change log unchanged; it contains all the changes since
3573 4.43. The change log for 4.44 is below; many of its items are identical to
3574 those for 4.50. This seems to be the most sensible way to preserve the
3575 historical information.
3581 1. Minor wording change to the doc/README.SIEVE file.
3583 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3584 computation of the current number of files was incorrect.
3586 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3587 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3590 4. Give more explanation in the error message when the command for a transport
3591 filter fails to execute.
3593 5. There are several places where Exim runs a non-Exim command in a
3594 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3595 was being done only for the command run by the queryprogram router. It is
3596 now done for all such subprocesses. The other cases are: ${run, transport
3597 filters, and the commands run by the lmtp and pipe transports.
3599 6. Added CONFIGURE_GROUP build-time option.
3601 7. Some older OS have a limit of 256 on the maximum number of file
3602 descriptors. Exim was using setrlimit() to set 1000 as a large value
3603 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3604 systems. I've change it so that if it can't get 1000, it tries for 256.
3606 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3607 was an oversight, and furthermore, ever since the addition of extra
3608 controls (e.g. 4.43/32), the checks on when to allow different forms of
3609 "control" were broken. There should now be diagnostics for all cases when a
3610 control that does not make sense is encountered.
3612 9. Added the /retain_sender option to "control=submission".
3614 10. $recipients is now available in the predata ACL (oversight).
3616 11. Tidy the search cache before the fork to do a delivery from a message
3617 received from the command line. Otherwise the child will trigger a lookup
3618 failure and thereby defer the delivery if it tries to use (for example) a
3619 cached ldap connection that the parent has called unbind on.
3621 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3622 of $address_data from the recipient verification was clobbered by the
3623 sender verification.
3625 13. The value of address_data from a sender verification is now available in
3626 $sender_address_data in subsequent conditions in the ACL statement.
3628 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3630 15. Added a new option "connect=<time>" to callout options, to set a different
3633 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3634 was its contents. (It was OK if the option was not defined at all.)
3636 17. A "Completed" log line is now written for messages that are removed from
3637 the spool by the -Mrm option.
3639 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3640 information about exactly what failed.
3642 19. Added -dd to debug only the daemon process.
3644 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3645 handles timeouts, both on the server side and network timeouts. Renamed the
3646 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3648 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3649 It is now set to "smtps".
3651 22. $host_address is now set to the target address during the checking of
3652 ignore_target_hosts.
3654 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3655 being passed; this would have caused $sender_host_name to have been used if
3656 matching the list had actually called for a host name (not very likely,
3657 since this list is usually IP addresses). A host name is now passed as
3660 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3661 code that specifies a non-restarting handler (typically sigaction() in
3662 modern systems) in an attempt to fix a rare and obscure crash bug.
3664 25. Narrowed the window for a race in the daemon that could cause it to ignore
3665 SIGCHLD signals. This is not a major problem, because they are used only to
3666 wake it up if nothing else does.
3668 26. A malformed maildirsize file could cause Exim to calculate negative values
3669 for the mailbox size or file count. Odd effects could occur as a result.
3670 The maildirsize information is now recalculated if the size or filecount
3673 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3674 support for a long time. Removed HAVE_SYS_VFS_H.
3676 28. Installed the latest version of exipick from John Jetmore.
3678 29. In an address list, if the pattern was not a regular expression, an empty
3679 subject address (from a bounce message) matched only if the pattern was an
3680 empty string. Non-empty patterns were not even tested. This was the wrong
3681 because it is perfectly reasonable to use an empty address as part of a
3682 database query. An empty address is now tested by patterns that are
3683 lookups. However, all the other forms of pattern expect the subject to
3684 contain a local part and a domain, and therefore, for them, an empty
3685 address still always fails if the pattern is not itself empty.
3687 30. Exim went into a mad DNS loop when attempting to do a callout where the
3688 host was specified on an smtp transport, and looking it up yielded more
3689 than one IP address.
3691 31. Re-factored the code for checking spool and log partition space into a
3692 function that finds that data and another that does the check. The former
3693 is then used to implement four new variables: $spool_space, $log_space,
3694 $spool_inodes, and $log_inodes.
3696 32. The RFC2047 encoding function was originally intended for short strings
3697 such as real names; it was not keeping to the 75-character limit for
3698 encoded words that the RFC imposes. It now respects the limit, and
3699 generates multiple encoded words if necessary. To be on the safe side, I
3700 have increased the buffer size for the ${rfc2047: expansion operator from
3703 33. It is now permitted to omit both strings after an "if" condition; if the
3704 condition is true, the result is "true". As before, when the second string
3705 is omitted, a false condition yields an empty string. This makes it less
3706 cumbersome to write custom ACL and router conditions.
3708 34. Failure to deliver a bounce message always caused it to be frozen, even if
3709 there was an errors_to setting on the router. The errors_to setting is now
3712 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3713 canonical form (fully expanded) before being placed in
3714 $sender_host_address.
3716 36. The table in the code that translates DNS record types into text (T_A to
3717 "A" for instance) was missing entries for NS and CNAME. It is just possible
3718 that this could have caused confusion if both these types were looked up
3719 for the same domain, because the text type is used as part of Exim's
3720 per-process caching. But the chance of anyone hitting this buglet seems
3723 37. The dnsdb lookup has been extended in a number of ways.
3725 (1) There is a new type, "zns", which walks up the domain tree until it
3726 finds some nameserver records. It should be used with care.
3728 (2) There is a new type, "mxh", which is like "mx" except that it returns
3729 just the host names, not the priorities.
3731 (3) It is now possible to give a list of domains (or IP addresses) to be
3732 looked up. The behaviour when one of the lookups defers can be
3733 controlled by a keyword.
3735 (4) It is now possible to specify the separator character for use when
3736 multiple records are returned.
3738 38. The dnslists ACL condition has been extended: it is now possible to supply
3739 a list of IP addresses and/or domains to be looked up in a particular DNS
3742 39. Added log_selector=+queue_time_overall.
3744 40. When running the queue in the test harness, wait just a tad after forking a
3745 delivery process, to get repeatability of debugging output.
3747 41. Include certificate and key file names in error message when GnuTLS fails
3748 to set them up, because the GnuTLS error message doesn't include the name
3749 of the failing file when there is a problem reading it.
3751 42. Allow both -bf and -bF in the same test run.
3753 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3755 44. The "Exiscan patch" is now merged into the mainline Exim source.
3757 45. Sometimes the final signoff response after QUIT could fail to get
3758 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3759 < 0 before doing a fflush(). This bug looks as though it goes back to the
3760 introduction of TLS in release 3.20, but "sometimes" must have been rare
3761 because the tests only now provoked it.
3763 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3764 (this can affect the format of dates).
3766 47. exim_tidydb, when checking for the continued existence of a message for
3767 which it has found a message-specific retry record, was not finding
3768 messages that were in split spool directories. Consequently, it was
3769 deleting retry records that should have stayed in existence.
3771 48. Steve fixed some bugs in eximstats.
3773 49. The SPA authentication driver was not abandoning authentication and moving
3774 on to the next authenticator when an expansion was forced to fail,
3775 contradicting the general specification for all authenticators. Instead it
3776 was generating a temporary error. It now behaves as specified.
3778 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3779 (the order specifies the preference for clients). The order is now AES256,
3780 AES128, 3DES, ARCFOUR128.
3782 51. Small patch to Sieve code - explicitly set From: when generating an
3785 52. Exim crashed if a remote delivery caused a very long error message to be
3786 recorded - for instance if somebody sent an entire SpamAssassin report back
3787 as a large number of 550 error lines. This bug was coincidentally fixed by
3788 increasing the size of one of Exim's internal buffers (big_buffer) that
3789 happened as part of the Exiscan merge. However, to be on the safe side, I
3790 have made the code more robust (and fixed the comments that describe what
3793 53. Now that there can be additional text after "Completed" in log lines (if
3794 the queue_time_overall log selector is set), a one-byte patch to exigrep
3795 was needed to allow it to recognize "Completed" as not the last thing in
3798 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3799 patch that reportedly fixes this has been added. I am not expert enough to
3800 create a test for it. This is what the patch creator wrote:
3802 "I found a little strange behaviour of ldap code when working with
3803 Windows 2003 AD Domain, where users was placed in more than one
3804 Organization Units. When I tried to give exim partial DN, the exit code
3805 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3806 But simultaneously result of request was absolutely normal ldap result,
3807 so I produce this patch..."
3809 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3810 REFERENCE, so I have modified the code to exclude the patch when that macro
3813 55. Some experimental protocols are using DNS PTR records for new purposes. The
3814 keys for these records are domain names, not reversed IP addresses. The
3815 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3816 leaves it alone. Component reversal etc. now happens only for IP addresses.
3819 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3821 57. Double the size of the debug message buffer (to 2048) so that more of very
3822 long debug lines gets shown.
3824 58. The exicyclog utility now does better if the number of log files to keep
3825 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3827 59. Two changes related to the smtp_active_hostname option:
3829 (1) $smtp_active_hostname is now available as a variable.
3830 (2) The default for smtp_banner uses $smtp_active_hostname instead
3831 of $primary_hostname.
3833 60. The host_aton() function is supposed to be passed a string that is known
3834 to be a valid IP address. However, in the case of IPv6 addresses, it was
3835 not checking this. This is a hostage to fortune. Exim now panics and dies
3836 if the condition is not met. A case was found where this could be provoked
3837 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3838 components; fortuitously, this particular loophole had already been fixed
3839 by change 4.50/55 above.
3841 If there are any other similar loopholes, the new check in host_aton()
3842 itself should stop them being exploited. The report I received stated that
3843 data on the command line could provoke the exploit when Exim was running as
3844 exim, but did not say which command line option was involved. All I could
3845 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3846 running as the user.
3849 61. There was a buffer overflow vulnerability in the SPA authentication code
3850 (which came originally from the Samba project). I have added a test to the
3851 spa_base64_to_bits() function which I hope fixes it.
3854 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3855 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3857 63. The daemon start-up calls getloadavg() while still root for those OS that
3858 need the first call to be done as root, but it missed one case: when
3859 deliver_queue_load_max is set with deliver_drop_privilege. This is
3860 necessary for the benefit of the queue runner, because there is no re-exec
3861 when deliver_drop_privilege is set.
3863 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3864 This has been fixed.
3866 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3867 was in use, was not putting the data itself into the right store pool;
3868 consequently, it could be overwritten for a subsequent message in the same
3869 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3872 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3874 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3875 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3876 the function were treating the return as a boolean value, which happened to
3877 work because 0=false and not-0=true, but is not correct code.
3879 68. The host_aton() function was not handling scoped IPv6 addresses (those
3880 with, for example, "%eth0" on the end) correctly.
3882 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3883 negated items (that is, ~something) in unsigned ints. Some compilers
3884 apparently mutter when there is no cast.
3886 70. If an address verification called from an ACL failed, and did not produce a
3887 user-specific message (i.e. there was only a "system" message), nothing was
3888 put in $acl_verify_message. In this situation, it now puts the system
3891 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3892 session; change 4.31/43 added the unwanted input to the log line - except
3893 that it did not do this in the start of session case. It now does.
3895 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3896 This is wrong and can cause the other end to generate a synchronization
3897 error if it is another Exim or anything else that does the synchronization
3898 check. A QUIT command is no longer sent after a timeout.
3900 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3901 during host lookups.
3903 74. The defer_ok option of callout verification was not working if it was used
3904 when verifying addresses in header lines, that is, for this case:
3906 verify = header_sender/callout=defer_ok
3908 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3909 those file descriptors could be used for SMTP connections. If anything
3910 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3911 could be sent to the SMTP client, causing chaos. The daemon now opens
3912 stdin, stdout, and stderr to /dev/null when it puts itself into the
3915 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3916 log by default. The user can override this with suitable Perl magic.
3918 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3919 the log message when discard triggers, was not working for the DATA ACL or
3920 for the non-SMTP ACL.
3922 78. Error message wording change in sieve.c.
3924 79. If smtp_accept_max_per_host was set, the number of connections could be
3925 restricted to fewer than expected, because the daemon was trying to set up
3926 a new connection before checking whether the processes handling previous
3927 connections had finished. The check for completed processes is now done
3928 earlier. On busy systems, this bug wouldn't be noticed because something
3929 else would have woken the daemon, and it would have reaped the completed
3932 80. If a message was submitted locally by a user whose login name contained one
3933 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3934 It caused a spool format error. I have fixed the spool reading code. A
3935 related problem was that the "from" clause in the Received: line became
3936 illegal because of the space(s). It is now covered by ${quote_local_part.
3938 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3941 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3943 83. Patch from Sieve maintainer to fix unterminated string problem in
3944 "vacation" handling.
3946 84. Some minor changes to the Linux configuration files to help with other
3947 OS variants using glibc.
3949 85. One more patch for Sieve to update vacation handling to latest spec.
3952 ----------------------------------------------------
3953 See the note above about the 4.44 and 4.50 releases.
3954 ----------------------------------------------------
3960 1. Change 4.43/35 introduced a bug that caused file counts to be
3961 incorrectly computed when quota_filecount was set in an appendfile
3964 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3965 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3968 3. Give more explanation in the error message when the command for a transport
3969 filter fails to execute.
3971 4. There are several places where Exim runs a non-Exim command in a
3972 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3973 was being done only for the command run by the queryprogram router. It is
3974 now done for all such subprocesses. The other cases are: ${run, transport
3975 filters, and the commands run by the lmtp and pipe transports.
3977 5. Some older OS have a limit of 256 on the maximum number of file
3978 descriptors. Exim was using setrlimit() to set 1000 as a large value
3979 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3980 systems. I've change it so that if it can't get 1000, it tries for 256.
3982 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3983 was an oversight, and furthermore, ever since the addition of extra
3984 controls (e.g. 4.43/32), the checks on when to allow different forms of
3985 "control" were broken. There should now be diagnostics for all cases when a
3986 control that does not make sense is encountered.
3988 7. $recipients is now available in the predata ACL (oversight).
3990 8. Tidy the search cache before the fork to do a delivery from a message
3991 received from the command line. Otherwise the child will trigger a lookup
3992 failure and thereby defer the delivery if it tries to use (for example) a
3993 cached ldap connection that the parent has called unbind on.
3995 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3996 of $address_data from the recipient verification was clobbered by the
3997 sender verification.
3999 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4000 was its contents. (It was OK if the option was not defined at all.)
4002 11. A "Completed" log line is now written for messages that are removed from
4003 the spool by the -Mrm option.
4005 12. $host_address is now set to the target address during the checking of
4006 ignore_target_hosts.
4008 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4009 being passed; this would have caused $sender_host_name to have been used if
4010 matching the list had actually called for a host name (not very likely,
4011 since this list is usually IP addresses). A host name is now passed as
4014 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4015 code that specifies a non-restarting handler (typically sigaction() in
4016 modern systems) in an attempt to fix a rare and obscure crash bug.
4018 15. Narrowed the window for a race in the daemon that could cause it to ignore
4019 SIGCHLD signals. This is not a major problem, because they are used only to
4020 wake it up if nothing else does.
4022 16. A malformed maildirsize file could cause Exim to calculate negative values
4023 for the mailbox size or file count. Odd effects could occur as a result.
4024 The maildirsize information is now recalculated if the size or filecount
4027 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4028 support for a long time. Removed HAVE_SYS_VFS_H.
4030 18. Updated exipick to current release from John Jetmore.
4032 19. Allow an empty sender to be matched against a lookup in an address list.
4033 Previously the only cases considered were a regular expression, or an
4036 20. Exim went into a mad DNS lookup loop when doing a callout where the
4037 host was specified on the transport, if the DNS lookup yielded more than
4040 21. The RFC2047 encoding function was originally intended for short strings
4041 such as real names; it was not keeping to the 75-character limit for
4042 encoded words that the RFC imposes. It now respects the limit, and
4043 generates multiple encoded words if necessary. To be on the safe side, I
4044 have increased the buffer size for the ${rfc2047: expansion operator from
4047 22. Failure to deliver a bounce message always caused it to be frozen, even if
4048 there was an errors_to setting on the router. The errors_to setting is now
4051 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4052 canonical form (fully expanded) before being placed in
4053 $sender_host_address.
4055 24. Updated eximstats to version 1.33
4057 25. Include certificate and key file names in error message when GnuTLS fails
4058 to set them up, because the GnuTLS error message doesn't include the name
4059 of the failing file when there is a problem reading it.
4061 26. Expand error message when OpenSSL has problems setting up cert/key files.
4064 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4065 (this can affect the format of dates).
4067 28. exim_tidydb, when checking for the continued existence of a message for
4068 which it has found a message-specific retry record, was not finding
4069 messages that were in split spool directories. Consequently, it was
4070 deleting retry records that should have stayed in existence.
4072 29. eximstats updated to version 1.35
4073 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4074 1.35 - bugfix such that pie charts by volume are generated correctly
4076 30. The SPA authentication driver was not abandoning authentication and moving
4077 on to the next authenticator when an expansion was forced to fail,
4078 contradicting the general specification for all authenticators. Instead it
4079 was generating a temporary error. It now behaves as specified.
4081 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4082 (the order specifies the preference for clients). The order is now AES256,
4083 AES128, 3DES, ARCFOUR128.
4085 31. Small patch to Sieve code - explicitly set From: when generating an
4088 32. Exim crashed if a remote delivery caused a very long error message to be
4089 recorded - for instance if somebody sent an entire SpamAssassin report back
4090 as a large number of 550 error lines. This bug was coincidentally fixed by
4091 increasing the size of one of Exim's internal buffers (big_buffer) that
4092 happened as part of the Exiscan merge. However, to be on the safe side, I
4093 have made the code more robust (and fixed the comments that describe what
4096 33. Some experimental protocols are using DNS PTR records for new purposes. The
4097 keys for these records are domain names, not reversed IP addresses. The
4098 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4099 leaves it alone. Component reversal etc. now happens only for IP addresses.
4102 34. The host_aton() function is supposed to be passed a string that is known
4103 to be a valid IP address. However, in the case of IPv6 addresses, it was
4104 not checking this. This is a hostage to fortune. Exim now panics and dies
4105 if the condition is not met. A case was found where this could be provoked
4106 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4107 components; fortuitously, this particular loophole had already been fixed
4108 by change 4.50/55 or 4.44/33 above.
4110 If there are any other similar loopholes, the new check in host_aton()
4111 itself should stop them being exploited. The report I received stated that
4112 data on the command line could provoke the exploit when Exim was running as
4113 exim, but did not say which command line option was involved. All I could
4114 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4115 running as the user.
4118 35. There was a buffer overflow vulnerability in the SPA authentication code
4119 (which came originally from the Samba project). I have added a test to the
4120 spa_base64_to_bits() function which I hope fixes it.
4123 36. The daemon start-up calls getloadavg() while still root for those OS that
4124 need the first call to be done as root, but it missed one case: when
4125 deliver_queue_load_max is set with deliver_drop_privilege. This is
4126 necessary for the benefit of the queue runner, because there is no re-exec
4127 when deliver_drop_privilege is set.
4129 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4130 was in use, was not putting the data itself into the right store pool;
4131 consequently, it could be overwritten for a subsequent message in the same
4132 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4135 38. Sometimes the final signoff response after QUIT could fail to get
4136 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4137 < 0 before doing a fflush(). This bug looks as though it goes back to the
4138 introduction of TLS in release 3.20, but "sometimes" must have been rare
4139 because the tests only now provoked it.
4145 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4146 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4147 Now it returns an integer. A number of calls were still expecting a T/F
4148 return. Fortuitously, in all cases, the tests worked in OK situations,
4149 which is the norm. However, things would have gone wrong on any write
4150 failures on the smtp file descriptor. This function is used when sending
4151 messages over SMTP and also when doing verify callouts.
4153 2. When Exim is called to do synchronous delivery of a locally submitted
4154 message (the -odf or -odi options), it no longer closes stderr before doing
4157 3. Implemented the mua_wrapper option.
4159 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4161 5. Implemented the functions header_remove(), header_testname(),
4162 header_add_at_position(), and receive_remove_recipient(), and exported them
4165 6. If an ACL "warn" statement specified the addition of headers, Exim already
4166 inserted X-ACL-Warn: at the start if there was no header name. However, it
4167 was not making this test for the second and subsequent header lines if
4168 there were newlines in the string. This meant that an invalid header could
4169 be inserted if Exim was badly configured.
4171 7. Allow an ACL "warn" statement to add header lines at the start or after all
4172 the Received: headers, as well as at the end.
4174 8. Added the rcpt_4xx retry error code.
4176 9. Added postmaster_mailfrom=xxx to callout verification option.
4178 10. Added mailfrom=xxxx to the callout verification option, for verify=
4181 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4182 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4184 12. Inserted some casts to stop certain compilers warning when using pointer
4185 differences as field lengths or precisions in printf-type calls (mostly
4186 affecting debugging statements).
4188 13. Added optional readline() support for -be (dynamically loaded).
4190 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4191 same clock tick as a message's arrival, so that its received time was the
4192 same as the "first fail" time on the retry record, and that message
4193 remained on the queue past the ultimate address timeout, every queue runner
4194 would try a delivery (because it was past the ultimate address timeout) but
4195 after another failure, the ultimate address timeout, which should have then
4196 bounced the address, did not kick in. This was a "< instead of <=" error;
4197 in most cases the first failure would have been in the next clock tick
4198 after the received time, and all would be well.
4200 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4201 being recognized when the domain list was tested by the match_domain
4202 condition in an expansion string.
4204 16. Added the ${str2b64: operator.
4206 17. Exim was always calling setrlimit() to set a large limit for the number of
4207 processes, without checking whether the existing limit was already
4208 adequate. (It did check for the limit on file descriptors.) Furthermore,
4209 errors from getrlimit() and setrlimit() were being ignored. Now they are
4210 logged to the main and panic logs, but Exim does carry on, to try to do its
4211 job under whatever limits there are.
4213 18. Imported PCRE 5.0.
4215 19. Trivial typo in log message " temporarily refused connection" (the leading
4218 20. If the log selector return_path_on_delivery was set and an address was
4219 redirected to /dev/null, the delivery process crashed because it assumed
4220 that a return path would always be set for a "successful" delivery. In this
4221 case, the whole delivery is bypassed as an optimization, and therefore no
4224 21. Internal re-arrangement: the function for sending a challenge and reading
4225 a response while authentication was assuming a zero-terminated challenge
4226 string. It's now changed to take a pointer and a length, to allow for
4227 binary data in such strings.
4229 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4231 23. Exim was not respecting finduser_retries when seeking the login of the
4232 uid under which it was called; it was always trying 10 times. (The default
4233 setting of finduser_retries is zero.) Also, it was sleeping after the final
4234 failure, which is pointless.
4236 24. Implemented tls_on_connect_ports.
4238 25. Implemented acl_smtp_predata.
4240 26. If the domain in control=submission is set empty, Exim assumes that the
4241 authenticated id is a complete email address when it generates From: or
4242 Sender: header lines.
4244 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4245 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4246 chown and chgrp in /bin and hostname in /usr/bin.
4248 28. Exim was keeping the "process log" file open after each use, just as it
4249 does for the main log. This opens the possibility of it remaining open for
4250 long periods when the USR1 signal hits a daemon. Occasional processlog
4251 errors were reported, that could have been caused by this. Anyway, it seems
4252 much more sensible not to leave this file open at all, so that is what now
4255 29. The long-running daemon process does not normally write to the log once it
4256 has entered its main loop, and it closes the log before doing so. This is
4257 so that log files can straightforwardly be renamed and moved. However,
4258 there are a couple of unusual error situations where the daemon does write
4259 log entries, and I had neglected to close the log afterwards.
4261 30. The text of an SMTP error response that was received during a remote
4262 delivery was being truncated at 512 bytes. This is too short for some of
4263 the long messages that one sometimes sees. I've increased the limit to
4266 31. It is now possible to make retry rules that apply only when a message has a
4267 specific sender, in particular, an empty sender.
4269 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4270 it possible to be selective about when SMTP synchronization is enforced.
4272 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4274 32. Implemented hosts_connection_nolog.
4276 33. Added an ACL for QUIT.
4278 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4281 35. Added mailbox_size and mailbox_filecount to appendfile.
4283 36. Added control = no_multiline_responses to ACLs.
4285 37. There was a bug in the logic of the code that waits for the clock to tick
4286 in the case where the clock went backwards by a substantial amount such
4287 that the microsecond fraction of "now" was more than the microsecond
4288 fraction of "then" (but the whole seconds number was less).
4290 38. Added support for the libradius Radius client library this is found on
4291 FreeBSD (previously only the radiusclient library was supported).
4297 1. When certain lookups returned multiple values in the form name=value, the
4298 quoting of the values was not always being done properly. Specifically:
4299 (a) If the value started with a double quote, but contained no whitespace,
4301 (b) If the value contained whitespace other than a space character (i.e.
4302 tabs or newlines or carriage returns) it was not quoted.
4303 This fix has been applied to the mysql and pgsql lookups by writing a
4304 separate quoting function and calling it from the lookup code. The fix
4305 should probably also be applied to nisplus, ibase and oracle lookups, but
4306 since I cannot test any of those, I have not disturbed their existing code.
4308 2. A hit in the callout cache for a specific address caused a log line with no
4309 reason for rejecting RCPT. Now it says "Previous (cached) callout
4310 verification failure".
4312 3. There was an off-by-one bug in the queryprogram router. An over-long
4313 return line was truncated at 256 instead of 255 characters, thereby
4314 overflowing its buffer with the terminating zero. As well as fixing this, I
4315 have increased the buffer size to 1024 (and made a note to document this).
4317 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4318 when Exim is waiting for an SMTP response from a remote server, Exim
4319 restarts its select() call on the socket, thereby resetting its timeout.
4320 This is not a problem when such interrupts are rare. Somebody set up a cron
4321 job to run exiwhat every 2 minutes, which is less than the normal select()
4322 timeout (5 or 10 minutes). This meant that the select() timeout never
4323 kicked in because it was always reset. I have fixed this by comparing the
4324 time when an interrupt arrives with the time at the start of the first call
4325 to select(). If more time than the timeout has elapsed, the interrupt is
4326 treated as a timeout.
4328 5. Some internal re-factoring in preparation for the addition of Sieve
4329 extensions (by MH). In particular, the "personal" test is moved to a
4330 separate function, and given an option for scanning Cc: and Bcc: (which is
4331 not set for Exim filters).
4333 6. When Exim created an email address using the login of the caller as the
4334 local part (e.g. when creating a From: or Sender: header line), it was not
4335 quoting the local part when it contained special characters such as @.
4337 7. Installed new OpenBSD configuration files.
4339 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4340 try to make them clearer.
4342 9. Callout options, other than the timeout value, were being ignored when
4343 verifying sender addresses in header lines. For example, when using
4345 verify = header_sender/callout=no_cache
4347 the cache was (incorrectly) being used.
4349 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4350 only the Cygwin environment.
4352 11. When return_path_on_delivery was set as a log selector, if different remote
4353 addresses in the same message used different return paths and parallel
4354 remote delivery occurred, the wrong values would sometimes be logged.
4355 (Whenever a remote delivery process finished, the return path value from
4356 the most recently started remote delivery process was logged.)
4358 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4359 lines when AUTH and/or TLS are in use. This is the "received protocol"
4360 field. Exim used to use "asmtp" for authenticated SMTP, without any
4361 indication (in the protocol name) for TLS use. Now it follows the RFC and
4362 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4363 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4364 names appear in log lines as well as in Received: header lines.
4366 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4367 extensions, and comparison tests, and to fix some bugs.
4369 14. Changes to the "personal" filter test:
4371 (1) The test was buggy in that it was just doing the equivalent of
4372 "contains" tests on header lines. For example, if a user's address was
4373 anne@some.where, the "personal" test would incorrectly be true for
4375 To: susanne@some.where
4377 This test is now done by extracting each address from the header in turn,
4378 and checking the entire address. Other tests that are part of "personal"
4379 are now done using regular expressions (for example, to check local parts
4380 of addresses in From: header lines).
4382 (2) The list of non-personal local parts in From: addresses has been
4383 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4384 taken from the Sieve specification recommendations.
4386 (3) If the message contains any header line starting with "List-" it is
4387 treated as non-personal.
4389 (4) The test for "circular" in the Subject: header line has been removed
4390 because it now seems ill-conceived.
4392 15. Minor typos in src/EDITME comments corrected.
4394 16. Installed latest exipick from John Jetmore.
4396 17. If headers_add on a router specified a text string that was too long for
4397 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4398 of string_sprintf() is now avoided.
4400 18. $message_body_size was not set (it was always zero) when running the DATA
4401 ACL and the local_scan() function.
4403 19. For the "mail" command in an Exim filter, no default was being set for
4404 the once_repeat time, causing a random time value to be used if "once" was
4405 specified. (If the value happened to be <= 0, no repeat happened.) The
4406 default is now 0s, meaning "never repeat". The "vacation" command was OK
4407 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4408 (I found it when inspecting the code).
4410 20. There is now an overall timeout for performing a callout verification. It
4411 defaults to 4 times the callout timeout, which applies to individual SMTP
4412 commands during the callout. The overall timeout applies when there is more
4413 than one host that can be tried. The timeout is checked before trying the
4414 next host. This prevents very long delays if there are a large number of
4415 hosts and all are timing out (e.g. when the network connections are timing
4416 out). The value of the overall timeout can be changed by specifying an
4417 additional sub-option for "callout", called "maxwait". For example:
4419 verify = sender/callout=5s,maxwait=20s
4421 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4422 to the end before writing, but this should make it even safer).
4424 22. Exim was forgetting that it had advertised PIPELINING for the second and
4425 subsequent messages on an SMTP connection. It was also not resetting its
4426 memory on STARTTLS and an internal HELO.
4428 23. When Exim logs an SMTP synchronization error within a session, it now
4429 records whether PIPELINING has been advertised or not.
4431 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4432 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4433 rather than long int.
4435 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4437 26. Added the never_mail option to autoreply.
4443 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4444 crash if the getsockname() call failed; this can happen if a connection is
4445 closed very soon after it is established. The problem was simply in the
4446 order in which certain operations were done, causing Exim to try to write
4447 to the SMTP stream before it had set up the file descriptor. The bug has
4448 been fixed by making things happen in the correct order.
4454 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4455 before the connection was closed, thus losing the rejection response.
4457 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4458 some early Solaris releases, but causes trouble in current releases where
4459 socklen_t is defined.
4461 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4464 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4467 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4468 that releasing the top of it at the end releases what was used for sub-
4469 expansions (unless the block got too big). However, discard this block if
4470 the first thing is a variable or header, so that we can use its block when
4471 it is dynamic (useful for very large $message_headers, for example).
4473 6. Lookups now cache *every* query, not just the most recent. A new, separate
4474 store pool is used for this. It can be recovered when all lookup caches are
4475 flushed. Lookups now release memory at the end of their result strings.
4476 This has involved some general refactoring of the lookup sources.
4478 7. Some code has been added to the store_xxx() functions to reduce the amount
4479 of flapping under certain conditions.
4481 8. log_incoming_interface used to affect only the <= reception log lines. Now
4482 it causes the local interface and port to be added to several more SMTP log
4483 lines, for example "SMTP connection from", and rejection lines.
4485 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4487 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4489 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4490 could be overwritten at the end of the current message (or the start of a
4491 new message if it was set in a HELO ACL). The value is now preserved for
4492 the duration of the SMTP connection.
4494 12. If a transport had a headers_rewrite setting, and a matching header line
4495 contained an unqualified address, that address was qualified, even if it
4496 did not match any rewriting rules. The underlying bug was that the values
4497 of the flags that permit the existence of unqualified sender and recipient
4498 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4499 non-local messages, and by -bnq for local messages) were not being
4500 preserved with the message after it was received.
4502 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4503 "next input=" as part of the text comprising the host identity instead of
4504 the correct text. The code was using the same buffer for two different
4505 strings. However, depending on which order the printing function evaluated
4506 its arguments, the bug did not always show up. Under Linux, for example, my
4507 test suite worked just fine.
4509 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4510 doesn't work with some older versions of Perl. It has been changed to "my",
4511 which in any case is probably the better facility to use.
4513 15. A really picky compiler found some instances of statements for creating
4514 error messages that either had too many or two few arguments for the format
4517 16. The size of the buffer for calls to the DNS resolver has been increased
4518 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4519 for addresses that have a lot of PTR records. This alleviates a problem; it
4520 does not fully solve it.
4522 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4523 buffer now truncates the list and logs the incident, which is the same
4524 action as happens when Exim is looking up a host name and its aliases.
4525 Previously in this situation something unpredictable would happen;
4526 sometimes it was "internal error: store_reset failed".
4528 18. If a server dropped the connection unexpectedly when an Exim client was
4529 using GnuTLS and trying to read a response, the client delivery process
4530 crashed while trying to generate an error log message.
4532 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4533 string, for example:
4535 warn message = H1: something\nH2: something
4537 the text was added as a single header line from Exim's point of view
4538 though it ended up OK in the delivered message. However, searching for the
4539 second and subsequent header lines using $h_h2: did not work. This has been
4540 fixed. Similarly, if a system filter added multiple headers in this way,
4541 the routers could not see them.
4543 20. Expanded the error message when iplsearch is called with an invalid key to
4544 suggest using net-iplsearch in a host list.
4546 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4547 ACLs are no longer actually imposed (and a message to that effect is
4550 22. If a "gecos" field in a passwd entry contained escaped characters, in
4551 particular, if it contained a \" sequence, Exim got it wrong when building
4552 a From: or a Sender: header from that name. A second bug also caused
4553 incorrect handling when an unquoted " was present following a character
4554 that needed quoting.
4556 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4557 was not being matched caselessly.
4559 24. Arranged for all hyphens in the exim.8 source to be escaped with
4562 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4563 properly. Recipient callout cache records were still being keyed to include
4564 the sender, even when use_sender was set false. This led to far more
4565 callouts that were necessary. The sender is no longer included in the key
4566 when use_sender is false.
4568 26. Added "control = submission" modifier to ACLs.
4570 27. Added the ${base62d: operator to decode base 62 numbers.
4572 28. dnsdb lookups can now access SRV records.
4574 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4575 the configuration file.
4577 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4578 (-v) mode. This makes the output for a verbose queue run more intelligible.
4580 31. Added a use_postmaster feature to recipient callouts.
4582 32. Added the $body_zerocount variable, containing the number of binary zero
4583 bytes in the message body.
4585 33. The time of last modification of the "new" subdirectory is now used as the
4586 "mailbox time last read" when there is a quota error for a maildir
4589 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4591 35. Added +ignore_unknown as a special item in host lists.
4593 36. Code for decoding IPv6 addresses in host lists is now included, even if
4594 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4595 address was recognized as an IP address, but was then not correctly decoded
4596 into binary, causing unexpected and incorrect effects when compared with
4603 1. Very minor rewording of debugging text in manualroute to say "list of
4604 hosts" instead of "hostlist".
4606 2. If verify=header_syntax was set, and a header line with an unqualified
4607 address (no domain) and a large number of spaces between the end of the
4608 name and the colon was received, the reception process suffered a buffer
4609 overflow, and (when I tested it) crashed. This was caused by some obsolete
4610 code that should have been removed. The fix is to remove it!
4612 3. When running in the test harness, delay a bit after writing a bounce
4613 message to get a bit more predictability in the log output.
4615 4. Added a call to search_tidyup() just before forking a reception process. In
4616 theory, someone could use a lookup in the expansion of smtp_accept_max_
4617 per_host which, without the tidyup, could leave open a database connection.
4619 5. Added the variables $recipient_data and $sender_data which get set from a
4620 lookup success in an ACL "recipients" or "senders" condition, or a router
4621 "senders" option, similar to $domain_data and $local_part_data.
4623 6. Moved the writing of debug_print from before to after the "senders" test
4626 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4627 problems for message scanning, either using a data ACL, or using
4628 local_scan() because the Received: header was not generated till after they
4629 were called (in order to set the time as the time of reception completion).
4630 I have revised the way this works. The header is now generated after the
4631 body is received, but before the ACL or local_scan() are called. After they
4632 are run, the timestamp in the header is updated.
4638 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4639 before starting a queue runner without re-exec. This happened only when
4640 deliver_drop_privilege was set or when the Exim user was set to root. The
4641 effect of the bug was that timeouts during subsequent deliveries caused
4642 crashes instead of being properly handled. The handler is now left at its
4643 default (and expected) setting.
4645 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4646 message, again when deliver_drop_privilege is set or Exim is run as root.
4647 The bug described in (1) was not present in this case, but the tidying up
4648 of the other signals was missing. I have made the two cases consistent.
4650 3. The ignore_target_hosts setting on a manualroute router was being ignored
4651 for hosts that were looked up using the /MX notation.
4653 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4656 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4657 operated on the sender address. After changing the $sender_address to <>
4658 for the sender address verify, Exim was re-instated it as the original
4659 (before rewriting) address, but remembering that it had rewritten it, so it
4660 wasn't rewriting it again. This bug also had the effect of breaking the
4661 sender address verification caching when the sender address was rewritten.
4663 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4664 This has been changed so that if the ip literal address matches
4665 ignore_target_hosts, the router declines.
4667 7. Added expansion conditions match_domain, match_address, and match_local_
4668 part (NOT match_host).
4670 8. The placeholder for the Received: header didn't have a length field set.
4672 9. Added code to Exim itself and to exim_lock to test for a specific race
4673 condition that could lead to file corruption when using MBX delivery. The
4674 issue is with the lockfile that is created in /tmp. If this file is removed
4675 after a process has opened it but before that process has acquired a lock,
4676 there is the potential for a second process to recreate the file and also
4677 acquire a lock. This could lead to two Exim processes writing to the file
4678 at the same time. The added code performs the same test as UW imapd; it
4679 checks after acquiring the lock that its file descriptor still refers to
4680 the same named file.
4682 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4683 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4686 11. Added the smtp_active_hostname option. If used, this will typically be made
4687 to depend on the incoming interface address. Because $interface_address is
4688 not set up until the daemon has forked a reception process, error responses
4689 that can happen earlier (such as "too many connections") no longer contain
4692 12. If an expansion in a condition on a "warn" statement fails because a lookup
4693 defers, the "warn" statement is abandoned, and the next ACL statement is
4694 processed. Previously this caused the whole ACL to be aborted.
4696 13. Added the iplsearch lookup type.
4698 14. Added ident_timeout as a log selector.
4700 15. Added tls_certificate_verified as a log selector.
4702 16. Added a global option tls_require_ciphers (compare the smtp transport
4703 option of the same name). This controls incoming TLS connections.
4705 17. I finally figured out how to make tls_require_ciphers do a similar thing
4706 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4707 before starting the TLS session.
4709 18. Tabs are now shown as \t in -bP output.
4711 19. If the log selector return_path_on_delivery was set, Exim crashed when
4712 bouncing a message because it had too many Received: header lines.
4714 20. If two routers both had headers_remove settings, and the first one included
4715 a superfluous trailing colon, the final name in the first list and the
4716 first name in the second list were incorrectly joined into one item (with a
4717 colon in the middle).
4723 1. Added -C and -D options to the exinext utility, mainly to make it easier
4724 to include in the automated testing, but these could be helpful when
4725 multiple configurations are in use.
4727 2. The exinext utility was not formatting the output nicely when there was
4728 an alternate port involved in the retry record key, nor when there was a
4729 message id as well (for retries that were specific to a specific message
4730 and a specific host). It was also confused by IPv6 addresses, because of
4731 the additional colons they contain. I have fixed the IPv4 problem, and
4732 patched it up to do a reasonable job for IPv6.
4734 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4735 delivery, the log line now contains "pipelined" if PIPELINING was used.
4737 4. An SMTP transport process used to panic and die if the bind() call to set
4738 an explicit outgoing interface failed. This has been changed; it is now
4739 treated in the same way as a connect() failure.
4741 5. A reference to $sender_host_name in the part of a conditional expansion
4742 that was being skipped was still causing a DNS lookup. This no longer
4745 6. The def: expansion condition was not recognizing references to header lines
4746 that used bh_ and bheader_.
4748 7. Added the _cache feature to named lists.
4750 8. The code for checking quota_filecount in the appendfile transport was
4751 allowing one more file than it should have been.
4753 9. For compatibility with Sendmail, the command line option
4761 and sets the incoming protocol and host name (for trusted callers). The
4762 host name and its colon can be omitted when only the protocol is to be set.
4763 Note the Exim already has two private options, -pd and -ps, that refer to
4764 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4765 "s", but I don't think that's a major issue.
4767 10. A number of refactoring changes to the code, none of which should affect
4770 (a) The number of logging options was getting close to filling up the
4771 32-bit word that was used as a bit map. I have split them into two classes:
4772 those that are passed in the argument to log_write(), and those that are
4773 only ever tested independently outside of that function. These are now in
4774 separate 32-bit words, so there is plenty of room for expansion again.
4775 There is no change in the user interface or the logging behaviour.
4777 (b) When building, for example, log lines, the code previously used a
4778 macro that called string_cat() twice, in order to add two strings. This is
4779 not really sufficiently general. Furthermore, there was one instance where
4780 it was actually wrong because one of the argument was used twice, and in
4781 one call a function was used. (As it happened, calling the function twice
4782 did not affect the overall behaviour.) The macro has been replaced by a
4783 function that can join an arbitrary number of extra strings onto a growing
4786 (c) The code for expansion conditions now uses a table and a binary chop
4787 instead of a serial search (which was left over from when there were very
4788 few conditions). Also, it now recognizes conditions like "pam" even when
4789 the relevant support is not compiled in: a suitably worded error message is
4790 given if an attempt is made to use such a condition.
4792 11. Added ${time_interval:xxxxx}.
4794 12. A bug was causing one of the ddress fields not to be passed back correctly
4795 from remote delivery subprocesses. The field in question was not being
4796 subsequently used, so this caused to problems in practice.
4798 13. Added new log selectors queue_time and deliver_time.
4800 14. Might have fixed a bug in maildirsizefile handling that threw up
4801 "unexpected character" debug warnings, and recalculated the data
4802 unnecessarily. In any case, I expanded the warning message to give more
4805 15. Added the message "Restricted characters in address" to the statements in
4806 the default ACL that block characters like @ and % in local parts.
4808 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4809 Three changes have been made:
4811 (a) There was a serious bug; a negative response to MAIL caused the whole
4812 recipient domain to be cached as invalid, thereby blocking all messages
4813 to all local parts at the same domain, from all senders. This bug has
4814 been fixed. The domain is no longer cached after a negative response to
4815 MAIL if the sender used is not empty.
4817 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4820 (c) A new callout option, "use_sender" has been added for people who want
4821 the modified behaviour.
4827 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4830 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4831 indeed breaks things for older releases.
4833 3. Added additional logging to the case where there is a problem reading data
4834 from a filter that is running in a subprocess using a pipe, in order to
4835 try to track down a specific problem.
4837 4. Testing facility fudge: when running in the test harness and attempting
4838 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4839 getting "No route to host". Convert this to a timeout.
4841 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4844 6. Some OS don't have socklen_t but use size_t instead. This affects the
4845 fifth argument of getsockopt() amongst other things. This is now
4846 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4847 can be set for individual OS. I have set it for SunOS5, OSF1, and
4848 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4849 some earlier ones do not.
4851 7. Change 4.30/15 was not doing the test caselessly.
4853 8. The standard form for an IPv6 address literal was being rejected by address
4854 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4855 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4856 this, as well as the form without the "IPv6" on the front (but only when
4857 address literals are enabled, of course).
4859 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4861 10. Exim crashed if a message with an empty sender address specified by -f
4862 encountered a router with an errors_to setting. This could be provoked only
4863 by a command such as
4867 where an empty string was supplied; "<>" did not hit this bug.
4869 11. Installed PCRE release 4.5.
4871 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4872 remained set. It is now erased.
4874 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4875 times from message ids (which are base 36 rather than the normal 62).
4877 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4878 were being counted as actual protocol errors, and logged if the log
4879 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4880 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4881 and DATA following a set of rejected RCPTs do not count as protocol errors.
4882 In other words, Exim assumes they were pipelined, though this may not
4883 actually be the case. Of course, in all cases the client gets an
4884 appropriate error code.
4886 15. If a lookup fails in an ACL condition, a message about the failure may
4887 be available; it is used if testing the ACL cannot continue, because most
4888 such messages specify what the cause of the deferral is. However, some
4889 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4890 that caused an old message to be retained and used if a later statement
4891 caused a defer, replacing the real cause of the deferral.
4893 16. If an IP address had so many PTR records that the DNS lookup buffer
4894 was not large enough to hold them, Exim could crash while trying to process
4895 the truncated data. It now detects and logs this case.
4897 17. Further to 4.21/58, another change has been made: if (and only if) the
4898 first line of a message (the first header line) ends with CRLF, a bare LF
4899 in a subsequent header line has a space inserted after it, so as not to
4900 terminate the header.
4902 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4903 unnecessarily, used atoi() instead of strtol(), and didn't check the
4904 termination when getting file sizes from file names by regex.
4906 19. Completely re-implemented the support for maildirsize files, in the light
4907 of a number of problems with the previous contributed implementation
4908 (4.30/29). In particular:
4910 . If the quota is zero, the maildirsize file is maintained, but no quota is
4913 . If the maildir directory does not exist, it is created before any attempt
4914 to write a maildirsize file.
4916 . The quota value in the file is just a cache; if the quota is changed in
4917 the transport, the new value overrides.
4919 . A regular expression is available for excluding directories from the
4922 20. The autoreply transport checks the characters in options that define the
4923 message's headers; it allows continued headers, but it was checking with
4924 isspace() after an embedded newline instead of explicitly looking for a
4927 21. If all the "regular" hosts to which an address was routed had passed their
4928 expiry times, and had not reached their retry times, the address was
4929 bounced, even if fallback hosts were defined. Now Exim should go on to try
4932 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4933 equivalent code in the SMTP transport. Some hosts send humungous responses
4934 to HELO/EHLO, more than 1024 it seems.
4936 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4937 gives compiler warnings in some environments. I've now done it "properly",
4940 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4941 (because of problems with the built-in one) was declared to return uschar *
4942 instead of char *, causing compiler failure.
4944 25. Fixed a file descriptor leak when processing alias/forward files.
4946 26. Fixed a minor format string issue in dbfn.c.
4948 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4950 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4951 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4954 29. When a sender address is verified, it is cached, to save repeating the test
4955 when there is more than one recipient in a message. However, when the
4956 verification involves a callout, it is possible for different callout
4957 options to be set for different recipients. It is too complicated to keep
4958 track of this in the cache, so now Exim always runs a verification when a
4959 callout is required, relying on the callout cache for the optimization.
4960 The overhead is duplication of the address routing, but this should not be
4963 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4964 to be verified with callout=postmaster, and the main callout worked but the
4965 postmaster check failed, the verification correctly failed. However, if a
4966 subsequent RCPT command asked for sender verification *without* the
4967 postmaster check, incorrect caching caused this verification also to fail,
4970 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4971 it was not caching the DNS options (qualify_single, search_parents) that
4972 were used when the lookup failed. A subsequent lookup with different
4973 options therefore always gave the same answer, though there were cases
4974 where it should not have. (Example: a "domains = !$mx_any" option on a
4975 dnslookup router: the "domains" option is always processed without any
4976 widening, but the router might have qualify_single set.) Now Exim uses the
4977 cached value only when the same options are set.
4979 32. Added John Jetmore's "exipick" utility to the distribution.
4981 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4982 than a timeout (e.g. a certificate is required, and is not provided), an
4983 Exim server now closes the connection immediately. Previously it waited for
4984 the client to close - but if the client is SSL, it seems that they each
4985 wait for each other, leading to a delay before one of them times out.
4987 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4988 maintained 0.8.x compatibility because I don't think many are using it, and
4989 it is clearly obsolete.
4991 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4994 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4995 client certificate was expired. A simple patch fixes this, though I don't
4996 understand the full logic of why the verify callback is called multiple
4999 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5000 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5001 which causes problems with some clients (such as the Certicom SSL Plus
5002 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5003 disables the coutermeasure allowing Eudora to connect."
5005 38. Exim was not checking that a write() to a log file succeeded. This could
5006 lead to Bad Things if a log got too big, in particular if it hit a file
5007 size limit. Exim now panics and dies if it cannot write to a log file, just
5008 as it does if it cannot open a log file.
5010 39. Modified OS/Makefile-Linux so that it now contains
5012 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5014 The two -D definitions ensure that Exim is compiled with large file
5015 support, which makes it possible to handle log files that are bigger than
5018 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5019 instance) a domain was checked against a named list that involved a lookup,
5020 causing $domain_data to be set, then another domain was checked against the
5021 same list, then the first domain was re-checked, the value of $domain_data
5022 after the final check could be wrong. In particular, if the second check
5023 failed, it could be set empty. This bug probably also applied to
5026 41. The strip_trailing_dot option was not being applied to the address given
5027 with the -f command-line option.
5029 42. The code for reading a message's header from the spool was incrementing
5030 $received_count, but never initializing it. This meant that the value was
5031 incorrect (doubled) while delivering a message in the same process in which
5032 it was received. In the most common configuration of Exim, this never
5033 happens - a fresh exec is done - but it can happen when
5034 deliver_drop_privilege is set.
5036 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5037 it now includes up to 150 characters of the unexpected data in the log
5040 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5041 and building data strings. The size of both of these buffers was 10 000
5042 bytes - far larger than anybody would *ever* want, thought I. Needless to
5043 say, somebody hit the limit. I have increased the maximum line length to
5044 20 000 and the maximum data length of concatenated lines to 100 000. I have
5045 also fixed two bugs, because there was no checking on these buffers. Tsk,
5046 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5047 buffer is too small.
5049 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5050 lsearch lookups. Now it does.
5052 46. When parsing a route_list item in a manualroute router, a fixed-length
5053 buffer was used for the list of hosts. I made this 1024 bytes long,
5054 thinking that nobody would ever have a list of hosts that long. Wrong.
5055 Somebody had a whole pile of complicated expansion conditions, and the
5056 string was silently truncated, leading to an expansion error. It turns out
5057 that it is easier to change to an unlimited length (owing to other changes
5058 that have happened since this code was originally written) than to build
5059 structure for giving a limitation error. The length of the item that
5060 expands into the list of hosts is now unlimited.
5062 47. The lsearch lookup could not handle data where the length of text line was
5063 more than 4095 characters. Such lines were truncated, leading to shortened
5064 data being returned. It should now handle lines of any length.
5066 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5067 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5070 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5071 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5072 no longer happens in comment lines. A list of replacements is now placed
5073 at the head of all of the source files, except those whose only change is
5074 to replace PERL_COMMAND in the very first #! line.
5076 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5077 messages on the queue, with a bottom-up merge sort, using code contributed
5078 by Michael Haardt. This should make operations like -bp somewhat faster on
5079 large queues. It won't affect queue runners, except when queue_run_in_order
5082 51. Installed eximstats 1.31 in the distribution.
5084 52. Added support for SRV lookups to the dnslookup router.
5086 53. If an ACL referred to $message_body or $message_body_end, the value was not
5087 reset for any messages that followed in the same SMTP session.
5089 54. The store-handling optimization for building very long strings was not
5090 differentiating between the different store pools. I don't think this
5091 actually made any difference in practice, but I've tidied it.
5093 55. While running the routers to verify a sender address, $sender_address
5094 was still set to the sender address. This is wrong, because when routing to
5095 send a bounce to the sender, it would be empty. Therefore, I have changed
5096 it so that, while verifying a sender address, $sender_address is set to <>.
5097 (There is no change to what happens when verifying a recipient address.)
5099 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5100 target A or AAAA records (if not already returned) without resetting the
5101 qualify_single or search_parents options of the DNS resolver. These are
5102 inappropriate in this case because the targets of MX and SRV records must
5103 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5104 target that, when qualified, matched something in the local domain. These
5105 two options are now turned off when doing these lookups.
5107 57. It seems that at least some releases of Reiserfs (which does not have the
5108 concept of a fixed number of inodes) returns zero and not -1 for the
5109 number of available inodes. This interacted badly with check_spool_inodes,
5110 which assumed that -1 was the "no such thing" setting. What I have done is
5111 to check that the total number of inodes is greater than zero before doing
5112 the test of how many are available.
5114 58. When a "warn" ACL statement has a log_message modifier, the message is
5115 remembered, and not repeated. This is to avoid a lot of repetition when a
5116 message has many recipients that cause the same warning to be written.
5117 Howewer, Exim was preserving the list of already written lines for an
5118 entire SMTP session, which doesn't seem right. The memory is now reset if a
5119 new message is started.
5121 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5122 debugging output unless log_rewrite was also set.
5124 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5125 of (int)(handle) when we know that handle contains (void *)(-1).
5127 61. The Exim daemon panic-logs an error return when it closes the incoming
5128 connection. However "connection reset by peer" seems to be common, and
5129 isn't really an error worthy of noting specially, so that particular error
5132 62. When Exim is trying to find all the local interfaces, it used to panic and
5133 die if the ioctl to get the interface flags failed. However, it seems that
5134 on at least one OS (Solaris 9) it is possible to have an interface that is
5135 included in the list of interfaces, but for which you get a failure error
5136 for this call. This happens when the interface is not "plumbed" into a
5137 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5138 failure of the "get flags" call assumes that the interface is down.
5140 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5141 makes life easier for people who are doing arithmetic on fields extracted
5142 from dates, where you often get leading zeros that should not be
5143 interpreted as octal.
5145 64. Added qualify_domain to the redirect router, to override the global
5148 65. If a pathologically long header line contained very many addresses (the
5149 report of this problem mentioned 10 000) and each of them was rewritten,
5150 Exim could use up a very large amount of memory. (It kept on making new
5151 copies of the header line as it rewrote, and never released the old ones.)
5152 At the expense of a bit more processing, the header rewriting function has
5153 been changed so that it no longer eats memory in this way.
5155 66. The generation of the Received: header has been moved from the time that a
5156 message starts to be received, to the time that it finishes. The timestamp
5157 in the Received: header should now be very close to that of the <= log
5158 line. There are two side-effects of this change:
5160 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5161 logged header lines no longer include the local Received: line, because
5162 it has not yet been created. The same applies to a copy of the message
5163 that is returned to a non-SMTP sender when a message is rejected.
5165 (b) When a filter file is tested using -bf, no additional Received: header
5166 is added to the test message. After some thought, I decided that this
5169 This change does not affect the value of $received_for. It is still set
5170 after address rewriting, but before local_scan() is called.
5172 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5174 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5175 gave an unhelpful panic error message, and a defer error. I have managed to
5176 change this behaviour so that it now rejects any supplied certificate,
5177 which seems right, as the list of acceptable certificates is empty.
5179 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5180 gave an unhelpful defer error. I have not managed to make this reject any
5181 supplied certificates, but the error message it gives is "no certificate
5182 supplied", which is not helpful.
5184 70. exigrep's output now also includes lines that are not associated with any
5185 message, but which match the given pattern. Implemented by a patch from
5186 Martin Sluka, which also tidied up the Perl a bit.
5188 71. Recipient callout verification, like sender verification, was using <> in
5189 the MAIL FROM command. This isn't really the right thing, since the actual
5190 sender may affect whether the remote host accepts the recipient or not. I
5191 have changed it to use the actual sender in the callout; this means that
5192 the cache record is now keyed on a recipient/sender pair, not just the
5193 recipient address. There doesn't seem to be a real danger of callout loops,
5194 since a callout by the remote host to check the sender would use <>.
5195 [SEE ABOVE: changed after hitting problems.]
5197 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5198 temporary errors. However, in the case of such a code being given after
5199 the end of a data transmission (i.e. after ".") Exim was failing to write
5200 a retry record for the message. (Yes, there was some broken host that was
5201 actually sending 8xx at this point.)
5203 73. An unknown lookup type in a host list could cause Exim to panic-die when
5204 the list was checked. (An example that provoked this was putting <; in the
5205 middle of a list instead of at the start.) If this happened during a DATA
5206 ACL check, a -D file could be left lying around. This kind of configuration
5207 error no longer causes Exim to die; instead it causes a defer errror. The
5208 incident is still logged to the main and panic logs.
5210 74. Buglet left over from Exim 3 conversion. The message "too many messages
5211 in one connection" was written to the rejectlog but not the mainlog, except
5212 when address rewriting (yes!) was being logged.
5214 75. Added write_rejectlog option.
5216 76. When a system filter was run not as root (that is, when system_filter_user
5217 was set), the values of the $n variables were not being returned to the
5218 main process; thus, they were not subsequently available in the $sn
5221 77. Added +return_path_on_delivery log selector.
5223 78. A connection timeout was being treated differently from recipients deferred
5224 when testing hosts_max_try with a message that was older than the host's
5225 retry timeout. (The host should not be counted, thus allowing all hosts to
5226 be tried at least once before bouncing.) This may have been the cause of an
5227 occasionally reported bug whereby a message would remain on the queue
5228 longer than the retry timeout, but would be bounced if a delivery was
5229 forced. I say "may" because I never totally pinned down the problem;
5230 setting up timeout/retry tests is difficult. See also the next item.
5232 79. The ultimate address timeout was not being applied to errors that involved
5233 a combination of host plus message (for example, a timeout on a MAIL
5234 command). When an address resolved to a number of possible hosts, and they
5235 were not all tried for each delivery (e.g. because of hosts_max_try), a
5236 message could remain on the queue longer than the retry timeout.
5238 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5241 81. Fixed an obscure SMTP outgoing bug which required at least the following
5242 conditions: (a) there was another message waiting for the same server;
5243 (b) the server returned 5xx to all RCPT commands in the first message so
5244 that the message was not completed; (c) the server dropped the connection
5245 or gave a negative response to the RSET that Exim sends to abort the
5246 transaction. The observed case was a dropped connection after DATA that had
5247 been sent in pipelining mode. That is, the server had advertised PIPELINING
5248 but was not implementing it correctly. The effect of the bug was incorrect
5249 behaviour, such as trying another host, and this could lead to a crash.
5255 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5256 and daemon.c were passed as pointers to ints; they should have been
5257 pointers to socklen_t variables (which are typically unsigned ints).
5259 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5262 3. Fixed a really odd bug that affected only the testing scheme; patching a
5263 certain fixed string in the binary changed the value of another string that
5264 happened to be identical to the end of the original first string.
5266 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5267 name", it returns that address as the IP address. On some operating
5268 systems (e.g. Solaris), it also passes back the IP address string as the
5269 "host name". However, on others (e.g. Linux), it passes back an empty
5270 string. Exim wasn't checking for this, and was changing the host name to an
5271 empty string, assuming it had been canonicized.
5273 5. Although rare, it is permitted to have more than one PTR record for a given
5274 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5275 all the names associated with an address, because they do in Solaris.
5276 However, it seems that they do not in Linux for data that comes from the
5277 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5278 I found this out when I moved to a new Linux workstation and tried to run
5279 the Exim test suite.
5281 To get round this problem I have changed the code so that it now does its
5282 own call to the DNS to look up PTR records when searching for a host name.
5283 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5284 addresses that are only in /etc/hosts are still found.
5286 This behaviour is, however, controlled by an option called host_lookup_
5287 order, which defaults to "bydns:byaddr". If people want to use the other
5288 order, or indeed, just use one or the other means of lookup, they can
5289 specify it in this variable.
5291 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5292 some operating systems, this comes back from gethostbyaddr() as an empty
5293 string, and this is what Exim used to test for. However, it seems that in
5294 other systems, "." is yielded. Exim now tests for this case too.
5296 7. The values of check_spool_space and check_log_space are now held internally
5297 as a number of kilobytes instead of an absolute number of bytes. If a
5298 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5299 kilobyte. This means that much larger values can be stored.
5301 8. Exim monitor: an attempt to get the action menu when not actually pointing
5302 at a message produces an empty menu entitled "No message selected". This
5303 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5304 no entries in it ("Shell widget menu has zero width and/or height"). So I
5305 have added a single, blank menu entry in this case.
5307 9. Added ${quote_local_part.
5309 10. MIME decoding is now applied to the contents of Subject: header lines when
5312 11. Now that a reference to $sender_host_address automatically causes a reverse
5313 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5314 host lookup before query-style lookups in lists that might use this
5315 variable. This has therefore been abolished, and the "net-" prefix is no
5316 longer necessary for query-style lookups.
5318 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5319 have been a typo for LFLAGS, so it has been changed.
5321 13. The install script calls Exim with "-C /dev/null" in order to find the
5322 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5323 to be output. Howeve, since Exim outputs its version number before the
5324 error, it didn't break the script. It just looked ugly. I fixed this by
5325 always allowing "-C /dev/null" if the caller is root.
5327 14. Ignore overlarge ACL variable number when reading spool file - insurance
5328 against a later release with more variables having written the file.
5330 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5331 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5332 this, as well as the form without the "IPv6" on the front.
5334 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5335 OS/Makefile-Darwin file.
5337 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5338 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5340 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5341 variable. However, some OS use ints for inodes. I've added cast to long int
5342 to get rid of the compiler warning.
5344 19. I had forgotten to lock out "/../" in configuration file names when
5345 ALT_CONFIG_PREFIX was set.
5347 20. Routers used for verification do not need to specify transports. However,
5348 if such a router generated a host list, and callout was configured, Exim
5349 crashed, because it could not find a port number from the (non-existent)
5350 transport. It now assumes port 25 in this circumstance.
5352 21. Added the -t option to exigrep.
5354 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5355 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5356 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5357 them is set, LOOKUP_LSEARCH is forced.
5359 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5361 24. Added sender and host information to the "rejected by local_scan()" log
5362 line; previously there was no indication of these.
5364 25. Added .include_if_exists.
5366 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5367 Linux. It turns out that not all file systems support this. Apparently some
5368 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5369 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5370 not supported on the file descriptor, is now ignored when Exim is trying to
5371 sync a directory. This applies only to Linux.
5373 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5375 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5376 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5379 29. Added support for maildirsize files from supplied patch (modified a bit).
5381 30. The use of :fail: followed by an empty string could lead Exim to respond to
5382 sender verification failures with (e.g.):
5384 550 Verification failed for <xxx>
5385 550 Sender verify failed
5387 where the first response line was missing the '-' that indicates it is not
5388 the final line of the response.
5390 31. The loop for finding the name of the user that called Exim had a hardwired
5391 limit of 10; it now uses the value of finduser_retries, which is used for
5392 all other user lookups.
5394 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5397 33. Exim was neglecting to zero errno before one call of strtol() when
5398 expanding a string and expecting an integer value. On some systems this
5399 resulted in spurious "integer overflow" errors. Also, it was casting the
5400 result into an int without checking.
5402 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5403 did not work. The code looks as if it has *never* worked, though it appears
5404 to have been documented since at least releast 1.62. I have made it work.
5406 35. The "timeout_DNS" error in retry rules, also documented since at least
5407 1.62, also never worked. As it isn't clear exactly what this means, and
5408 clearly it isn't a major issue, I have abolished the feature by treating it
5409 as "timeout", and writing a warning to the main and panic logs.
5411 36. The display of retry rules for -brt wasn't always showing the error code
5414 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5415 timeout_connect_A, timeout_connect_MX.
5417 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5418 to the empty sender.
5420 39. The daemon was not analysing the content of -oX till after it had closed
5421 stderr and disconnected from the controlling terminal. This meant that any
5422 syntax errors were only noted on the panic log, and the return code from
5423 the command was 0. By re-arranging the code a little, I've made the
5424 decoding happen first, so such errors now appear on stderr, and the return
5425 code is 1. However, the actual setting up of the sockets still happens in
5426 the disconnected process, so errors there are still only recorded on the
5429 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5430 connections (as happens on some IP stacks) was logged at start up time as
5431 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5432 it from "IPv6 and IPv4", which means that two separate sockets are being
5435 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5436 says whether AF_INET or AF_INET6 was passed as an argument.
5438 42. Exiwhat output was messed up when time zones were included in log
5441 43. Exiwhat now gives more information about the daemon's listening ports,
5442 and whether -tls-on-connect was used.
5444 44. The "port" option of the smtp transport is now expanded.
5446 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5447 silently ignored. Now an error message is written to the main and panic
5450 46. There's a new ACL modifier called "logwrite" which writes to a log file
5451 as soon as it is encountered.
5453 47. Added $local_user_uid and $local_user_gid at routing time.
5455 48. Exim crashed when trying to verify a sender address that was being
5458 49. Exim was recognizing only a space character after ".include". It now also
5459 recognizes a tab character.
5461 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5462 extracting the relevant information from the specification. The man page no
5463 longer contains scrambled data for the -d option, and I've added a section
5464 at the front about calling Exim under different names.
5466 51. Added "extra_headers" argument to the "mail" command in filter files.
5468 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5471 53. Installed eximstats 1.29.
5473 54. Added transport_filter_timeout as a generic transport option.
5475 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5476 Cc: header lines. This was required by RFC 822, but it not required by RFC
5479 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5480 incoming messages that did not have them. Now it does so only if the
5481 message originates locally, that is, if there is no associated remote host
5482 address. When Resent- header lines are present, this applies to the Resent-
5483 lines rather than the non-Resent- lines.
5485 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5486 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5488 58. Messages for configuration errors now include the name of the main
5489 configuration file - useful now that there may be more than one file in a
5490 list (.included file names were always shown).
5492 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5493 for those rare installations that do not start the daemon as root or run it
5494 setuid root. I've cut out the call to initgroups() if the daemon is not
5497 60. The Exim user and group can now be bound into the binary as text strings
5498 that are looked up at the start of Exim's processing.
5500 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5502 62. Added $mailstore_basename variable.
5504 63. Installed patch to sieve.c from Michael Haardt.
5506 64. When Exim failed to open the panic log after failing to open the main log,
5507 the original message it was trying to log was written to stderr and debug
5508 output, but if they were not available (the usual case in production), it
5509 was lost. Now it is written to syslog before the two lines that record the
5510 failures to open the logs.
5512 65. Users' Exim filters run in subprocesses under the user's uid. It is
5513 possible for a "deliver" command or an alias in a "personal" command to
5514 provoke an address rewrite. If logging of address rewriting is configured,
5515 this fails because the process is not running as root or exim. There may be
5516 a better way of dealing with this, but for the moment (because 4.30 needs
5517 to be released), I have disabled address rewrite logging when running a
5518 filter in a non-root, non-exim process.
5524 1. The buildconfig auxiliary program wasn't quoting the value set for
5525 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5526 not defined. This bug was masked in 4.22 by the effect that was fixed in
5529 2. Some messages that were rejected after a message id was allocated were
5530 shown as "incomplete" by exigrep. It no longer does this for messages that
5531 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5533 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5534 have allow_domain_literals set, the ID did not get logged in the <= line.
5535 Domain literals are now always recognized in Message-ID: header lines.
5537 4. The first argument for a ${extract expansion item is the key name or field
5538 number. Leading and trailing spaces in this item were not being ignored,
5539 causing some misleading effects.
5541 5. When deliver_drop_privilege was set, single queue runner processes started
5542 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5543 same command in the process it spins off) were not dropping privilege.
5545 6. When the daemon running as "exim" started a queue runner, it always
5546 re-executed Exim in the spun-off process. This is a waste of effort when
5547 deliver_drop_privilege is set. The new process now just calls the
5548 queue-runner function directly.
5554 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5557 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5558 ignored. Though the use of -oP was forcing the writing of a pid file, it
5559 was always written to the default place.
5561 3. If the message "no IP address found for host xxxx" is generated during
5562 incoming verification, it is now followed by identification of the incoming
5563 connection (so you can more easily find what provoked it).
5565 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5567 5. Added some features to "harden" Exim a bit more against certain attacks:
5569 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5570 be put in Local/Makefile. This is like the never_users runtime option,
5571 but it cannot be overridden. The default setting is "root".
5573 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5574 prefix string with which any file named in a -C command line option
5577 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5578 is retained for -C and -D only if the caller of Exim is root. Without
5579 it, the exim user may also use -C and -D and retain privilege.
5581 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5582 command line option is disabled.
5584 6. Macro names set by the -D option must start with an upper case letter, just
5585 like macro names defined in the configuration file.
5587 7. Added "dereference=" facility to LDAP.
5589 8. Two instances of the typo "uknown" in the source files are fixed.
5591 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5592 the Configure-Makefile script screwed up while processing it.
5594 10. Incorporated PCRE 4.4.
5596 11. The SMTP synchronization check was not operating right at the start of an
5597 SMTP session. For example, it could not catch a HELO sent before the client
5598 waited for the greeting. There is now a check for outstanding input at the
5599 point when the greeting is written. Because of the duplex, asynchronous
5600 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5601 way, but not yet received, when the check is performed.
5603 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5604 on TCP/IP sockets, because this apparently causes some broken clients to
5607 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5608 unchanged) from the Cygwin maintainer.
5610 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5611 was supported instead of testing for mbx. Effectively a typo.
5613 15. The spa authenticator server code was not checking that the input it
5614 received was valid base64.
5616 16. The debug output line for the "set" modifier in ACLs was not showing the
5617 name of the variable that was being set.
5619 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5621 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5622 lookup on its own. Something else was needed to trigger the lookup. For
5623 example, a match in host_lookup or the need for a host name in a host list.
5624 Now, if $sender_host_name is referenced and the host name has not yet been
5625 looked up, a lookup is performed. If the lookup fails, the variable remains
5626 empty, and $host_lookup_failed is set to "1".
5628 19. Added "eqi" as a case-independent comparison operator.
5630 20. The saslauthd authentication condition could segfault if neither service
5631 nor realm was specified.
5633 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5634 error message that was logged was misleading, and incoming SMTP
5635 connections were dropped. The message is now more accurate, and temporary
5636 errors are given to SMTP connections.
5638 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5639 (except RSET) with a 421 error, until QUIT is received. However, it was
5640 failing to send a response to QUIT.
5642 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5643 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5644 ACL is now run after the helo_try_verify_hosts code.
5646 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5647 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5648 case-independent, so other case variants are also recognized). Apparently
5649 some systems use these upper case variants.
5651 25. If more than two messages were waiting for the same host, and a transport
5652 filter was specified for the transport, Exim sent two messages over the
5653 same TCP/IP connection, and then failed with "socket operation on non-
5654 socket" when it tried to send the third.
5656 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5658 27. The extern definition of crypt16() in expand.c was not being excluded when
5659 the OS had its own crypt16() function.
5661 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5662 as a preferred synonym for return_size_limit, both as an option and as an
5665 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5667 30. Changed the default configuration ACL to relax the local part checking rule
5668 for addresses that are not in any local domains. For these addresses,
5669 slashes and pipe symbols are allowed within local parts, but the sequence
5670 /../ is explicitly forbidden.
5672 31. SPA server authentication was not clearing the challenge buffer before
5675 32. log_message in a "warn" ACL statement was writing to the reject log as
5676 well as to the main log, which contradicts the documentation and doesn't
5677 seem right (because no rejection is happening). So I have stopped it.
5679 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5680 However, I am unable to do any testing of this.
5682 34. Fixed an infelicity in the appendfile transport. When checking directories
5683 for a mailbox, to see if any needed to be created, it was accidentally
5684 using path names with one or more superfluous leading slashes; tracing
5685 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5687 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5688 contents are added to the log line that is written for every discarded
5689 recipient. (Previously a log_message setting was ignored.)
5691 36. The ${quote: operator now quotes the string if it is empty.
5693 37. The install script runs exim in order to find its version number. If for
5694 some reason other than non-existence or emptiness, which it checks, it
5695 could not run './exim', it was installing it with an empty version number,
5696 i.e. as "exim-". This error state is now caught, and the installation is
5699 38. An argument was missing from the function that creates an error message
5700 when Exim fails to connect to the socket for saslauthd authentication.
5701 This could cause Exim to crash, or give a corrupted message.
5703 39. Added isip, isip4, and isip6 to ${if conditions.
5705 40. The ACL variables $acl_xx are now saved with the message, and can be
5706 accessed later in routers, transports, and filters.
5708 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5709 strings in the file are not string-expanded.
5711 42. If a MAIL command specified a SIZE value that was too large to fit into an
5712 int variable, the check against message_size_limit failed. Such values are
5713 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5714 day this will have to be increased, but I don't think I want to be around
5715 when emails are that large.
5722 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5723 iconv() is not standard in FreeBSD.
5725 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5726 IPv6 enabled. The observed symptom was a segmentation fault on return from
5727 the function os_common_find_running_interfaces() in src/os.c.
5729 3. In the check_special_case() function in daemon.c I had used "errno" as an
5730 argument name, which causes warnings on some systems. This was basically a
5731 typo, since it was named "eno" in the comments!
5733 4. The code that waits for the clock to tick (at a resolution of some fraction
5734 of a second) so as to ensure message-id uniqueness was always waiting for
5735 at least one whole tick, when it could have waited for less. [This is
5736 almost certainly not relevant at current processor speeds, where it is
5737 unlikely to ever wait at all. But we try to future-proof.]
5739 5. The function that sleeps for a time interval that includes fractions of a
5740 second contained a race. It did not block SIGALRM between setting the
5741 timer, and suspending (a couple of lines later). If the interval was short
5742 and the sigsuspend() was delayed until after it had expired, the suspension
5743 never ended. On busy systems this could lead to processes getting stuck for
5746 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5747 process, before it forks any delivery processes. The open lookup caching
5748 mechanism meant that the open file or database connection was passed into
5749 the delivery process. The problem was that delivery processes always tidy
5750 up cached lookup data. This could cause a problem for the next delivery
5751 process started by the queue runner, because the external queue runner
5752 process does not know about the closure. So the next delivery process
5753 still has data in the lookup cache. In the case of a file lookup, there was
5754 no problem because closing a file descriptor in a subprocess doesn't affect
5755 the parent. However, if the lookup was caching a connection to a database,
5756 the connection was closed, and the second delivery process was likely to
5757 see errors such as "PGSQL: query failed: server closed the connection
5758 unexpectedly". The problem has been fixed by closing all cached lookups
5759 in a queue runner before running a delivery process.
5761 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5762 seem to have the "const" qualifier which it has on other OS. I've
5765 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5766 *of the same type* (client or server) with the same public name that an
5767 error should be diagnosed.
5769 9. When Exim looked up a host name for an IP address, but failed to find the
5770 original IP address when looking up the host name (a safety check), it
5771 output the message "<ip address> does not match any IP for NULL", which was
5772 confusing, to say the least. The bug was that the host name should have
5773 appeared instead of "NULL".
5775 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5776 user that is built into the binary, and the -C or -D options is used, root
5777 privilege is dropped before the configuration file is read. In addition,
5778 logging is switched to stderr instead of the normal log files. If the
5779 configuration then re-defines the Exim user, the unprivileged environment
5780 is probably not what is expected, so Exim logs a panic warning message (but
5783 However, if deliver_drop_privilege is set, the unprivileged state may well
5784 be exactly what is intended, so the warning has been cut out in that case,
5785 and Exim is allowed to try to write to its normal log files.
5791 1. smtp_return_error_details was not giving details for temporary sender
5792 or receiver verification errors.
5794 2. Diagnose a configuration error if two authenticators have the same public
5797 3. Exim used not to create the message log file for a message until the first
5798 delivery attempt. This could be confusing when incoming messages were held
5799 for policy or load reasons. The message log file is now created at the time
5800 the message is received, and an initial "Received" line is written to it.
5802 4. The automatically generated man page for command line options had a minor
5803 bug that caused no ill effects; however, a more serious problem was that
5804 the procedure for building the man page automatically didn't always
5805 operate. Consequently, release 4.20 contains an out-of-date version. This
5806 shouldn't happen again.
5808 5. When building Exim with embedded Perl support, the script that builds the
5809 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5810 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5812 6. The freeze_tell option was not being used for messages that were frozen on
5813 arrival, either by an ACL or by local_scan().
5815 7. Added the smtp_incomplete_transaction log selector.
5817 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5818 was accepting AUTH without a new EHLO.
5820 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5821 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5824 10. Logging of TCP/IP connections (when configured) now happens in the main
5825 daemon process instead of the child process, so that the TCP/IP connection
5826 count is more accurate (but it can never be perfect).
5828 11. The use of "drop" in a nested ACL was not being handled correctly in the
5829 outer ACL. Now, if condition failure induced by the nested "drop" causes
5830 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5831 or "require"), the connection is dropped.
5833 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5834 that yield "discard" can now be used with an "accept" or a "discard" verb,
5835 but an error is generated for any others (because I can't see a useful way
5836 to define what should happen).
5838 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5839 are now processed in the same way as lines in the Exim configuration file.
5840 In particular, continuation lines are supported.
5842 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5844 15. Added -ti meaning -t -i.
5846 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5847 domains, and warn by logging if others are found.
5849 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5850 returned from GIFCONF when Exim is trying to find the list of interfaces on
5851 a host. The code in os.c has been modified to copy each ifreq to an aligned
5852 structure in all cases.
5854 Also, in some cases, the returned ifreq's were being copied to a 'struct
5855 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5856 means the last couple of bytes of an IPv6 address could be chopped if the
5857 ifreq contained only a normal sockaddr (14 bytes storage).
5859 18. Named domain lists were not supported in the hosts_treat_as_local option.
5860 An entry such as +xxxx was not recognized, and was treated as a literal
5863 19. Ensure that header lines added by a DATA ACL are included in the reject log
5864 if the ACL subsequently rejects the message.
5866 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5867 MD5 (which is deprecated).
5869 21. When testing a filter file using -bf, Exim was writing a message when it
5870 took the sender from a "From " line in the message, but it was not doing so
5871 when it took $return_path from a Return-Path: header line. It now does.
5873 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5874 with a valid header line field name (a series of printing characters
5875 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5877 23. Changed "disc" in the source to "disk" to conform to the documentation and
5878 the book and for uniformity.
5880 24. Ignore Sendmail's -Ooption=value command line item.
5882 25. When execve() failed while trying to run a command in a pipe transport,
5883 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5884 could be confused with a return value of 69 from the command itself. This
5885 has been changed to 127, the value the shell returns if it is asked to run
5886 a non-existent command. The wording for the related log line suggests a
5887 non-existent command as the problem.
5889 26. If received_header_text expands to an empty string, do not add a Received:
5890 header line to the message. (Well, it adds a token one on the spool, but
5891 marks it "old" so that it doesn't get used or transmitted.)
5893 27. Installed eximstats 1.28 (addition of -nt option).
5895 28. There was no check for failure on the call to getsockname() in the daemon
5896 code. This can fail if there is a shortage of resources on the system, with
5897 ENOMEM, for example. A temporary error is now given on failure.
5899 29. Contrary to the C standard, it seems that in some environments, the
5900 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5901 program. Exim now does this explicitly; it affects the formatting of
5902 timestamps using strftime().
5904 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5905 complaints. I've now initialized all the variables, to avoid this.
5907 32. Header lines added by a system filter were not being "seen" during
5908 transport-time rewrites.
5910 33. The info_callback() function passed to OpenSSL is set up with type void
5911 (*)(SSL *, int, int), as described somewhere. However, when calling the
5912 function (actually a macro) that sets it up, the type void(*)() is
5913 expected. I've put in a cast to prevent warnings from picky compilers.
5915 34. If a DNS black list lookup found a CNAME record, but there were no A
5916 records associated with the domain it pointed at, Exim crashed.
5918 35. If a DNS black list lookup returned more than one A record, Exim ignored
5919 all but the first. It now scans all returned addresses if a particular IP
5920 value is being sought. In this situation, the contents of the
5921 $dnslist_value variable are a list of all the addresses, separated by a
5924 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5925 to accept a host name and all its aliases if the forward lookup for any of
5926 them yielded the IP address of the incoming connection. Now it accepts only
5927 those names whose forward lookup yields the correct IP address. Any other
5928 names are discarded. This closes a loophole whereby a rogue DNS
5929 administrator could create reverse DNS records to break through a
5930 wildcarded host restriction in an ACL.
5932 37. If a user filter or a system filter that ran in a subprocess used any of
5933 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5934 the wrong values were passed to the pipe command ($thisaddress had the
5935 value of $0, $0 had the value of $1, etc). This bug was introduced by
5936 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5938 38. Improved the line breaking for long SMTP error messages from ACLs.
5939 Previously, if there was no break point between 40 and 75 characters, Exim
5940 left the rest of the message alone. Two changes have been made: (a) I've
5941 reduced the minimum length to 35 characters; (b) if it can't find a break
5942 point between 35 and 75 characters, it looks ahead and uses the first one
5943 that it finds. This may give the occasional overlong line, but at least the
5944 remaining text gets split now.
5946 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5947 file descriptors might be low, and that setting 1000 would always raise it.
5948 It turns out that in some environments, the limit is already over 1000 and
5949 that lowering it causes trouble. So now Exim takes care not to decrease it.
5951 40. When delivering a message, the value of $return_path is set to $sender_
5952 address at the start of routing (routers may change the value). By an
5953 oversight, this default was not being set up when an address was tested by
5954 -bt or -bv, which affected the outcome if any router or filter referred to
5957 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5958 log only when "message" or "log_message" are set. However, if one of the
5959 conditions was an address verification, or a call to a nested ACL, the
5960 messages generated by the underlying test were being passed through. This
5961 no longer happens. The underlying message is available in $acl_verify_
5962 message for both "message" and "log_message" expansions, so it can be
5963 passed through if needed.
5965 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5966 new expansion $bh_ to give the encoded byte string without charset
5967 translation. Translation happens only if iconv() is available; HAVE_ICONV
5968 indicates this at build time. HEADERS_CHARSET gives the charset to
5969 translate to; headers_charset can change it in the configuration, and
5970 "headers charset" can change it in an individual filter file.
5972 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5973 that creates RFC 2047 encoded "words" labels them as that charset instead
5974 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5975 expansion operator; (ii) when Exim creates a From: line for a local
5976 message; (iii) when a header line is rewritten to include a "phrase" part.
5978 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5979 buggy, causing it to skip the first lines of messages whose message ID
5980 ended in 'D'. This would not have bitten before Exim release 4.14, because
5981 message IDs were unlikely to end in 'D' before then. The effect was to have
5982 incorrect size information for certain domains.
5984 45. #include "config.h" was missing at the start of the crypt16.c module. This
5985 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5988 46. If there was a timeout during a "random" callout check, Exim treated it as
5989 a failure of the random address, and carried on sending RSET and the real
5990 address. If the delay was just some slowness somewhere, the response to the
5991 original RCPT would be taken as a response to RSET and so on, causing
5992 mayhem of various kinds.
5994 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5995 when I implemented it. It didn't allow for the fact that some option values
5996 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5997 the right test for positive values.
5999 48. Domain names in DNS records are case-independent. Exim always looks them up
6000 in lower case. Some resolvers return domain names in exactly the case they
6001 appear in the zone file, that is, they may contain uppercase letters. Not
6002 all resolvers do this - some return always lower case. Exim was treating a
6003 change of case by a resolver as a change of domain, similar to a widening
6004 of a domain abbreviation. This triggered its re-routing code and so it was
6005 trying to route what was effectively the same domain again. This normally
6006 caused routing to fail (because the router wouldn't handle the domain
6007 twice). Now Exim checks for this case specially, and just changes the
6008 casing of the domain that it ultimately uses when it transmits the message
6011 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6014 50. If a filter generated a file delivery with a non-absolute name (possible if
6015 no home directory exists for the router), the forbid_file option was not
6018 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6019 the existing equality matching.
6021 52. Exim was using ints instead of ino_t variables in some places where it was
6022 dealing with inode numbers.
6024 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6025 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6026 called TMPDIR, and if it finds it is different, it changes its value.
6028 54. The smtp_printf() function is now made available to local_scan() so
6029 additional output lines can be written before returning. There is also an
6030 smtp_fflush() function to enable the detection of a dropped connection.
6031 The variables smtp_input and smtp_batched_input are exported to
6034 55. Changed the default runtime configuration: the message "Unknown user"
6035 has been removed from the ACL, and instead placed on the localuser router,
6036 using the cannot_route_message feature. This means that any verification
6037 failures that generate their own messages won't get overridden. Similarly,
6038 the "Unrouteable address" message that was in the ACL for unverifiable
6039 relay addresses has also been removed.
6041 56. Added hosts_avoid_esmtp to the smtp transport.
6043 57. The exicyclog script was not checking for the esoteric option
6044 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6045 will work only if exicyclog is run under the appropriate euid.
6047 58. Following a discussion on the list, the rules by which Exim recognises line
6048 endings on incoming messages have been changed. The -dropcr and drop_cr
6049 options are now no-ops, retained only for backwards compatibility. The
6050 following line terminators are recognized: LF CRLF CR. However, special
6051 processing applies to CR:
6053 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6054 nor a local message in the state where . is a terminator.
6056 (ii) If a bare CR is encountered in a header line, an extra space is added
6057 after the line terminator so as not to end the header. The reasoning
6058 behind this is that bare CRs in header lines are most likely either
6059 to be mistakes, or people trying to play silly games.
6061 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6062 was being incorrectly given as 18 bytes larger than it should have been.
6063 This is a VOB (very old bug).
6065 60. This may never have affected anything current, but just in case it has:
6066 When the local host is found other than at the start of a list of hosts,
6067 the local host, those with the same MX, and any that follow, are discarded.
6068 When the list in question was part of a longer list of hosts, the following
6069 hosts (not currently being processed) were also being discarded. This no
6070 longer happens. I'm not sure if this situation could ever has previously
6073 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6076 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6077 header. This is something that is recommended in a new Internet Draft, and
6078 is something that is documented as being done by Sendmail. There are two
6079 possible values. For messages generated by the autoreply transport, Exim
6082 Auto-Submitted: auto-replied
6084 whereas for all other generated messages (e.g. bounces) it adds
6086 Auto-Submitted: auto-generated
6088 63. The "personal" condition in filters now includes a test for the
6089 Auto-Submitted: header. If it contains the string "auto-" the message it
6090 not considered personal.
6092 64. Added rcpt_include_affixes as a generic transport option.
6094 65. Added queue_only_override (default true).
6096 66. Added the syslog_duplication option.
6098 67. If what should have been the first header line of a message consisted of
6099 a space followed by a colon, Exim was mis-interpreting it as a header line.
6100 It isn't of course - it is syntactically invalid and should therefore be
6101 treated as the start of the message body. The misbehaviour could have
6102 caused a number of strange effects, including loss of data in subsequent
6103 header lines, and spool format errors.
6105 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6106 client host had authenticated. This control can now be exercised by an ACL
6107 for more flexibility.
6109 69. By default, callouts do not happen when testing with -bh. There is now a
6110 variant, -bhc, which does actually run the callout code, including
6111 consulting and updating the callout cache.
6113 70. Added support for saslauthd authentication, courtesy of Alexander
6116 71. If statvfs() failed on the spool or log directories while checking their
6117 size for availability, Exim confusingly gave the error "space shortage".
6118 Furthermore, in debugging mode it crashed with a floating point exception.
6119 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6120 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6121 serious problem, Exim now writes to the main and panic logs when this
6122 happens, with details of the failure. It then refuses to accept the
6123 incoming message, giving the message "spool directory problem" or "log
6124 directory problem" with a 421 code for SMTP messages.
6126 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6127 0, 1, and 2 exist, because some OS complain for execs without them (see
6128 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6129 descriptors. However, the code omitted to check that the open succeeded,
6130 causing mysterious errors if for some reason the permissions on /dev/null
6131 got screwed. Now Exim writes a message to the main and panic logs, and
6132 bombs out if it can't open /dev/null.
6134 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6135 interact so that it is all more flexible. It is supposed to remain
6136 backwards compatible. Also added extra_local_interfaces.
6138 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6139 to bomb out with an assertion failure - to the client this appears as a
6140 connection drop. This problem occurs in the part of the code that was taken
6141 from the Samba project. Fortunately, the assertion is in a very simple
6142 function, so I have fixed this by reproducing the function inline in the
6143 one place where it is called, and arranging for authentication to fail
6144 instead of killing the process with assert().
6146 75. The SPA client code was not working when the server requested OEM rather
6147 than Unicode encoding.
6149 76. Added code to make require_files with a specific uid setting more usable in
6150 the case where statting the file as root fails - usually a non-root-mounted
6151 NFS file system. When this happens and the failure is EACCES, Exim now
6152 forks a subprocess and does the per-uid checking as the relevant uid.
6154 77. Added process_log_path.
6156 78. If log_file_path was not explicitly set, a setting of check_log_space or
6157 check_log_inodes was ignored.
6159 79. If a space check for the spool or log partitions fails, the incident is now
6160 logged. Of course, in the latter case the data may get lost...
6162 80. Added the %p formatting code to string_format() so that it can be used to
6163 print addresses in debug_print(). Adjusted all the address printing in the
6164 debugging in store.c to use %p rather than %d.
6166 81. There was a concern that a line of code in smtp_in.c could overflow a
6167 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6168 initially expressed, the concern was not well-founded, because trailing
6169 spaces are removed early. However, if the trailing spaces were followed by
6170 a NULL, they did not get removed, so the overflow was possible. Two fixes
6173 (a) I re-wrote the offending code in a cleaner fashion.
6174 (b) If an incoming SMTP command contains a NULL character, it is rejected
6177 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6178 runs initgroups(), so that if the Exim user is in any additional groups,
6179 they will be used during message reception.
6185 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob