1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
15 TL/02 The BSD's have an arc4random API. One of the functions to induce
16 adding randomness was arc4random_stir(), but it has been removed in
17 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
18 function when detected.
20 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
21 cause callback expansion.
23 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
24 syntax errors in an expansion can be treated as a string instead of
25 logging or causing an error, due to the internal use of bool_lax
26 instead of bool when processing it.
28 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
29 server certificates when making smtp deliveries.
31 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
33 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
35 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
37 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
38 Merged patch from Sebastian Wiedenroth.
40 JH/05 Fix results-pipe from transport process. Several recipients, combined
41 with certificate use, exposed issues where response data items split
42 over buffer boundaries were not parsed properly. This eventually
43 resulted in duplicates being sent. This issue only became common enough
44 to notice due to the introduction of conection certificate information,
45 the item size being so much larger. Found and fixed by Wolfgang Breyha.
47 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
48 size buffer was used, resulting in syntax errors when an expansion
51 JH/07 Add support for directories of certificates when compiled with a GnuTLS
52 version 3.3.6 or later.
54 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
55 is EXPERIMENTAL_EVENT, the main-configuration and transport options
56 both become "event_action", the variables become $event_name, $event_data
57 and $event_defer_errno. There is a new variable $verify_mode, usable in
58 routers, transports and related events. The tls:cert event is now also
59 raised for inbound connections, if the main configuration event_action
62 TL/06 In test suite, disable OCSP for old versions of openssl which contained
63 early OCSP support, but no stapling (appears to be less than 1.0.0).
65 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
66 server certificate names available under the smtp transport option
67 "tls_verify_cert_hostname" now do not permit multi-component wildcard
70 JH/10 Time-related extraction expansions from certificates now use the main
71 option "timezone" setting for output formatting, and are consistent
72 between OpenSSL and GnuTLS compilations. Bug 1541.
74 JH/11 Fix a crash in mime ACL when meeting a zero-length parameter in the
77 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
78 include certificate info, eximon was claiming there were spoolfile
81 JH/13 Buf 1521: Fix ldap lookup for single-attr request, multiple-attr return.
83 JH/14 Log delivery-related information more consistently, using the sequence
84 "H=<name> [<ip>]" wherever possible.
89 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
90 checkers that were complaining about end of non-void function with no
93 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
94 This was a regression intruduced in 4.83 by another bugfix.
96 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
98 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
99 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
105 TF/01 Correctly close the server side of TLS when forking for delivery.
107 When a message was received over SMTP with TLS, Exim failed to clear up
108 the incoming connection properly after forking off the child process to
109 deliver the message. In some situations the subsequent outgoing
110 delivery connection happened to have the same fd number as the incoming
111 connection previously had. Exim would try to use TLS and fail, logging
112 a "Bad file descriptor" error.
114 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
115 utilities have not been installed.
117 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
118 temporary space as the ACL may create new global variables.
120 TL/01 LDAP support uses per connection or global context settings, depending
121 upon the detected version of the libraries at build time.
123 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
124 to extract and use the src ip:port in logging and expansions as if it
125 were a direct connection from the outside internet. PPv2 support was
126 updated based on HAProxy spec change in May 2014.
128 JH/02 Add ${listextract {number}{list}{success}{fail}}.
130 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
131 Properly escape header and check for NULL return.
133 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
136 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
138 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
139 characters in header names, implemented as a verify condition.
140 Contributed by Michael Fischer v. Mollard.
142 TL/05 Rename SPF condition results err_perm and err_temp to standardized
143 results permerror and temperror. Previous values are deprecated but
144 still accepted. In a future release, err_perm and err_temp will be
145 completely removed, which will be a backward incompatibility if the
146 ACL tests for either of these two old results. Patch contributed by
147 user bes-internal on the mailing list.
149 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
151 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
152 selectors, in both main and reject logs.
154 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
157 JH/07 Add malware type "sock" for talking to simple daemon.
159 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
161 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
162 routers/transports under cutthrough routing.
164 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
165 numbers. Touch up "bool" conditional to keep the same definition.
167 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
169 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
171 JH/12 Expand items in router/transport headers_add or headers_remove lists
172 individually rather than the list as a whole. Bug 1452.
174 Required for reasonable handling of multiple headers_ options when
175 they may be empty; requires that headers_remove items with embedded
176 colons must have them doubled (or the list-separator changed).
178 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
179 view the policy declared in the DMARC record. Currently, $dmarc_status
180 is a combined value of both the record presence and the result of the
183 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
185 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
186 dnslookup router and the smtp transport (applying to the forward
189 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
190 of ldap servers used for a specific lookup. Patch provided by Heiko
193 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
194 New variable $lookup_dnssec_authenticated for observability.
196 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
197 Patch submitted by Lars Timman.
199 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
201 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
202 Requires trusted mode and valid format message id, aborts otherwise.
203 Patch contributed by Heiko Schlichting.
205 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
206 certextract with support for various fields. Bug 1358.
208 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
209 is requested by default, modifiable by smtp transport option
212 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
213 operate on certificate variables to give certificate fingerprints
214 Also new ${sha256:cert_variable}.
216 JH/23 The PRDR feature is moved from being Experimental into the mainline.
218 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
219 Christian Aistleitner.
221 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
223 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
224 file. Patch from Wolfgang Breyha.
226 JH/25 Expand the coverage of the delivery $host and $host_address to
227 client authenticators run in verify callout. Bug 1476.
229 JH/26 Port service names are now accepted for tls_on_connect_ports, to
230 align with daemon_smtp_ports. Bug 72.
232 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
233 support and error reporting did not work properly.
235 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
236 and is readable. Patch from Andrew Colin Kissa.
238 TL/14 Enhance documentation of ${run expansion and how it parses the
239 commandline after expansion, particularly in the case when an
240 unquoted variable expansion results in an empty value.
242 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
244 PP/02 Fix internal collision of T_APL on systems which support RFC3123
245 by renaming away from it. Addresses GH issue 15, reported by
248 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
250 TL/15 SECURITY: prevent double expansion in math comparison functions
251 (can expand unsanitized data). Not remotely exploitable.
258 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
260 PP/02 Make -n do something, by making it not do something.
261 When combined with -bP, the name of an option is not output.
263 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
266 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
267 $sender_host_name and config options to manage this, and basic check
270 PP/05 DSCP support for outbound connections and control modifier for inbound.
272 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
273 (Only plugin which currently uses this is kerberos4, which nobody should
274 be using, but we should make it available and other future plugins might
275 conceivably use it, even though it would break NAT; stuff *should* be
276 using channel bindings instead).
278 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
279 name; added for Sendmail compatibility; requires admin caller.
280 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
281 just ignore it); requires trusted caller.
282 Also parse but ignore: -Ac -Am -X<logfile>
285 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
287 TL/02 Add +smtp_confirmation as a default logging option.
289 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
290 Patch by Magnus Holmgren from 2007-02-20.
292 TL/04 Bugzilla 1281 - Spec typo.
293 Bugzilla 1283 - Spec typo.
294 Bugzilla 1290 - Spec grammar fixes.
296 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
298 TL/06 Add Experimental DMARC support using libopendmarc libraries.
300 TL/07 Fix an out of order global option causing a segfault. Reported to dev
301 mailing list by by Dmitry Isaikin.
303 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
305 JH/02 Support "G" suffix to numbers in ${if comparisons.
307 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
309 NM/01 Bugzilla 1197 - Spec typo
310 Bugzilla 1196 - Spec examples corrections
312 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
314 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
315 gnutls_enable_pkcs11, but renamed to more accurately indicate its
318 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
319 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
321 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
322 "acl {{name}{arg}...}", and optional args on acl condition
325 JH/05 Permit multiple router/transport headers_add/remove lines.
327 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
329 JH/07 Avoid using a waiting database for a single-message-only transport.
330 Performance patch from Paul Fisher. Bugzilla 1262.
332 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
335 JH/09 Add $headers_added variable, with content from use of ACL modifier
336 add_header (but not yet added to the message). Bugzilla 199.
338 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
339 Pulled from Bugzilla 817 by Wolfgang Breyha.
341 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
343 (nb: this is the same fix as in Exim 4.80.1)
345 JH/11 Add A= logging on delivery lines, and a client_set_id option on
348 JH/12 Add optional authenticated_sender logging to A= and a log_selector
351 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
353 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
354 advertise SMTP AUTH mechanism to us, instead of a generic
355 protocol violation error. Also, make Exim more robust to bad
356 data from the Dovecot auth socket.
358 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
360 When a queue runner is handling a message, Exim first routes the
361 recipient addresses, during which it prunes them based on the retry
362 hints database. After that it attempts to deliver the message to
363 any remaining recipients. It then updates the hints database using
366 So if a recipient address works intermittently, it can get repeatedly
367 deferred at routing time. The retry hints record remains fresh so the
368 address never reaches the final cutoff time.
370 This is a fairly common occurrence when a user is bumping up against
371 their storage quota. Exim had some logic in its local delivery code
372 to deal with this. However it did not apply to per-recipient defers
373 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
375 This change adds a proper retry rule check during routing so that the
376 final cutoff time is checked against the message's age. We only do
377 this check if there is an address retry record and there is not a
378 domain retry record; this implies that previous attempts to handle
379 the address had the retry_use_local_parts option turned on. We use
380 this as an approximation for the destination being like a local
381 delivery, as in LMTP.
383 I suspect this new check makes the old local delivery cutoff check
384 redundant, but I have not verified this so I left the code in place.
386 TF/02 Correct gecos expansion when From: is a prefix of the username.
388 Test 0254 submits a message to Exim with the header
392 When I ran the test suite under the user fanf2, Exim expanded
393 the header to contain my full name, whereas it should have added
394 a Resent-Sender: header. It erroneously treats any prefix of the
395 username as equal to the username.
397 This change corrects that bug.
399 GF/01 DCC debug and logging tidyup
400 Error conditions log to paniclog rather than rejectlog.
401 Debug lines prefixed by "DCC: " to remove any ambiguity.
403 TF/03 Avoid unnecessary rebuilds of lookup-related code.
405 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
406 Bug spotted by Jeremy Harris; was flawed since initial commit.
407 Would have resulted in OCSP responses post-SNI triggering an Exim
408 NULL dereference and crash.
410 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
412 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
413 Bug detection, analysis and fix by Samuel Thibault.
414 Bugzilla 1331, Debian bug #698092.
416 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
418 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
419 Server implementation by Todd Lyons, client by JH.
420 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
421 config variable "prdr_enable" controls whether the server
422 advertises the facility. If the client requests PRDR a new
423 acl_data_smtp_prdr ACL is called once for each recipient, after
424 the body content is received and before the acl_smtp_data ACL.
425 The client is controlled by bolth of: a hosts_try_prdr option
426 on the smtp transport, and the server advertisement.
427 Default client logging of deliveries and rejections involving
428 PRDR are flagged with the string "PRDR".
430 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
431 fclose(). Diagnosis by Todd Lyons.
433 PP/17 Update configure.default to handle IPv6 localhost better.
434 Patch by Alain Williams (plus minor tweaks).
437 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
438 This is now consistent with GnuTLS, and is now documented: the
439 previous undocumented portable approach to treating the option as
440 unset was to force an expansion failure. That still works, and
441 an empty string is now equivalent.
443 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
444 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
445 not performing validation itself.
447 PP/20 Added force_command boolean option to pipe transport.
448 Patch from Nick Koston, of cPanel Inc.
450 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
453 TF/04 Added udpsend ACL modifer and hexquote expansion operator
455 PP/21 Fix eximon continuous updating with timestamped log-files.
456 Broken in a format-string cleanup in 4.80, missed when I repaired the
457 other false fix of the same issue.
458 Report and fix from Heiko Schlichting.
461 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
462 Report from Prashanth Katuri.
464 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
465 It's SecureTransport, so affects any MacOS clients which use the
466 system-integrated TLS libraries, including email clients.
468 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
469 using a MIME ACL for non-SMTP local injection.
470 Report and assistance in diagnosis by Warren Baker.
472 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
474 JH/16 Fix comparisons for 64b. Bugzilla 1385.
476 TL/09 Add expansion variable $authenticated_fail_id to keep track of
477 last id that failed so it may be referenced in subsequent ACL's.
479 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
482 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
483 ldap library initialization, allowing self-signed CA's to be
484 used. Also properly sets require_cert option later in code by
485 using NULL (global ldap config) instead of ldap handle (per
486 session). Bug diagnosis and testing by alxgomz.
488 TL/12 Enhanced documentation in the ratelimit.pl script provided in
489 the src/util/ subdirectory.
491 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
492 renamed to Transport Post Delivery Action by Jeremy Harris, as
495 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
496 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
497 redis_servers = needs to be configured which will be used by the redis
498 lookup. Patch from Warren Baker, of The Packet Hub.
500 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
502 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
503 hostname or reverse DNS when processing a host list. Used suggestions
504 from multiple comments on this bug.
506 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
508 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
509 Missed a few lines, added it to make the runtest require no keyboard
512 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
513 contains upper case chars. Make router use caseful_local_part.
515 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
516 support when GnuTLS has been built with p11-kit.
522 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
524 This, or similar/improved, will also be change PP/11 of 4.82.
530 PP/01 Handle short writes when writing local log-files.
531 In practice, only affects FreeBSD (8 onwards).
532 Bugzilla 1053, with thanks to Dmitry Isaikin.
534 NM/01 Bugzilla 949 - Documentation tweak
536 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
539 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
541 PP/02 Implemented gsasl authenticator.
543 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
545 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
546 `pkg-config foo` for cflags/libs.
548 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
549 with rest of GSASL and with heimdal_gssapi.
551 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
552 `pkg-config foo` for cflags/libs for the TLS implementation.
554 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
555 properties get this fed in as external SSF. A number of robustness
556 and debugging improvements to the cyrus_sasl authenticator.
558 PP/08 cyrus_sasl server now expands the server_realm option.
560 PP/09 Bugzilla 1214 - Log authentication information in reject log.
561 Patch by Jeremy Harris.
563 PP/10 Added dbmjz lookup type.
565 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
567 PP/12 MAIL args handles TAB as well as SP, for better interop with
568 non-compliant senders.
569 Analysis and variant patch by Todd Lyons.
571 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
572 Bug report from Lars Müller <lars@samba.org> (via SUSE),
573 Patch from Dirk Mueller <dmueller@suse.com>
575 PP/13 tls_peerdn now print-escaped for spool files.
576 Observed some $tls_peerdn in wild which contained \n, which resulted
577 in spool file corruption.
579 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
580 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
581 or write after TLS renegotiation, which otherwise led to messages
584 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
585 as a tracking header (ie: a signed header comes before the signature).
586 Patch from Wolfgang Breyha.
588 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
589 comma-sep list; embedded commas doubled.
591 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
593 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
595 Report and patch from Dmitry Banschikov.
597 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
598 Removed SSL_clear() after SSL_new() which led to protocol negotiation
599 failures. We appear to now support TLS1.1+ with Exim.
601 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
602 lets Exim select keys and certificates based upon TLS SNI from client.
603 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
604 before an outbound SMTP session. New log_selector, +tls_sni.
606 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
607 NULL dereference. Report and patch from Alun Jones.
609 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
610 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
611 Not seeing resolver debug output on NetBSD, but suspect this is a
612 resolver implementation change.
614 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
615 Left warnings. Added "eximon gdb" invocation mode.
617 PP/21 Defaulting "accept_8bitmime" to true, not false.
619 PP/22 Added -bw for inetd wait mode support.
621 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
622 locate the relevant includes and libraries. Made this the default.
624 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
625 Bugzilla 1246, report and most of solution from Tomasz Kusy.
627 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
628 This may cause build issues on older platforms.
630 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
631 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
632 gnutls_require_mac & gnutls_require_protocols (no longer supported).
633 Added SNI support via GnuTLS too.
634 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
636 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
638 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
639 Applied second patch from Janne, implementing suggestion to default
640 multiple-strings-in-record handling to match SPF spec.
642 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
644 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
645 read-only, out of scope).
646 Patch from Wolfgang Breyha, report from Stuart Northfield.
648 PP/29 Fix three issues highlighted by clang analyser static analysis.
649 Only crash-plausible issue would require the Cambridge-specific
650 iplookup router and a misconfiguration.
651 Report from Marcin Mirosław.
653 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
655 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
656 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
657 As part of this, removing so much warning spew let me fix some minor
658 real issues in debug logging.
660 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
661 assignment on my part. Fixed.
663 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
664 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
665 Janne Snabb (who went above and beyond: thank you).
667 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
668 string otherwise requires a connection and a bunch more work and it's
669 relatively easy to get wrong. Should also expose TLS library linkage
672 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
673 64-bit ${eval} (JH/03).
675 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
676 GNU libc to support some of the 64-bit stuff, should not lead to
677 conflicts. Defined before os.h is pulled in, so if a given platform
678 needs to override this, it can.
680 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
681 protection layer was required, which is not implemented.
682 Bugzilla 1254, patch from Wolfgang Breyha.
684 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
685 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
686 tls_dhparam take prime identifiers. Also unbreak combination of
687 OpenSSL+DH_params+TLSSNI.
689 PP/39 Disable SSLv2 by default in OpenSSL support.
695 PP/01 Solaris build fix for Oracle's LDAP libraries.
696 Bugzilla 1109, patch from Stephen Usher.
698 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
700 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
703 TF/02 Fix a couple more cases where we did not log the error message
704 when unlink() failed. See also change 4.74-TF/03.
706 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
707 lock up or crash if it happened to be inside a call to libc when it
708 got a SIGUSR1 from exiwhat.
710 The SIGUSR1 handler appends the current process status to the process
711 log which is later printed by exiwhat. It used to use the general
712 purpose logging code to do this, but several functions it calls are
713 not safe for signals.
715 The new output code in the SIGUSR1 handler is specific to the process
716 log, and simple enough that it's easy to inspect for signal safety.
717 Removing some special cases also simplifies the general logging code.
718 Removing the spurious timestamps from the process log simplifies
721 TF/04 Improved ratelimit ACL condition.
723 The /noupdate option has been deprecated in favour of /readonly which
724 has clearer semantics. The /leaky, /strict, and /readonly update modes
725 are mutually exclusive. The update mode is no longer included in the
726 database key; it just determines when the database is updated. (This
727 means that when you upgrde Exim will forget old rate measurements.)
729 Exim now checks that the per_* options are used with an update mode that
730 makes sense for the current ACL. For example, when Exim is processing a
731 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
732 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
733 must specify per_mail/readonly. If you omit the update mode it defaults to
734 /leaky where that makes sense (as before) or /readonly where required.
736 The /noupdate option is now undocumented but still supported for
737 backwards compatibility. It is equivalent to /readonly except that in
738 ACLs where /readonly is required you may specify /leaky/noupdate or
739 /strict/noupdate which are treated the same as /readonly.
741 A useful new feature is the /count= option. This is a generalization
742 of the per_byte option, so that you can measure the throughput of other
743 aggregate values. For example, the per_byte option is now equivalent
744 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
746 The per_rcpt option has been generalized using the /count= mechanism
747 (though it's more complicated than the per_byte equivalence). When it is
748 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
749 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
750 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
751 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
752 Note that using per_rcpt with a non-readonly update mode in more than
753 one ACL will cause the recipients to be double-counted. (The per_mail
754 and per_byte options don't have this problem.)
756 The handling of very low rates has changed slightly. If the computed rate
757 is less than the event's count (usually one) then this event is the first
758 after a long gap. In this case the rate is set to the same as this event's
759 count, so that the first message of a spam run is counted properly.
761 The major new feature is a mechanism for counting the rate of unique
762 events. The new per_addr option counts the number of different
763 recipients that someone has sent messages to in the last time period. It
764 behaves like per_rcpt if all the recipient addresses are different, but
765 duplicate recipient addresses do not increase the measured rate. Like
766 the /count= option this is a general mechanism, so the per_addr option
767 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
768 example, measure the rate that a client uses different sender addresses
769 with the options per_mail/unique=$sender_address. There are further
770 details in the main documentation.
772 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
774 TF/06 Removed a few PCRE remnants.
776 TF/07 Automatically extract Exim's version number from tags in the git
777 repository when doing development or release builds.
779 PP/02 Raise smtp_cmd_buffer_size to 16kB.
780 Bugzilla 879. Patch from Paul Fisher.
782 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
783 Heavily based on revision 40f9a89a from Simon Arlott's tree.
786 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
788 PP/05 Variable $av_failed, true if the AV scanner deferred.
789 Bugzilla 1078. Patch from John Horne.
791 PP/06 Stop make process more reliably on build failure.
792 Bugzilla 1087. Patch from Heiko Schlittermann.
794 PP/07 Make maildir_use_size_file an _expandable_ boolean.
795 Bugzilla 1089. Patch from Heiko Schlittermann.
797 PP/08 Handle ${run} returning more data than OS pipe buffer size.
798 Bugzilla 1131. Patch from Holger Weiß.
800 PP/09 Handle IPv6 addresses with SPF.
801 Bugzilla 860. Patch from Wolfgang Breyha.
803 PP/10 GnuTLS: support TLS 1.2 & 1.1.
805 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
808 PP/11 match_* no longer expand right-hand-side by default.
809 New compile-time build option, EXPAND_LISTMATCH_RHS.
810 New expansion conditions, "inlist", "inlisti".
812 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
814 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
816 PP/14 fix log_write() format string regression from TF/03.
817 Bugzilla 1152. Patch from Dmitry Isaikin.
823 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
825 PP/02 Harmonised TLS library version reporting; only show if debugging.
826 Layout now matches that introduced for other libraries in 4.74 PP/03.
828 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
830 PP/04 New "dns_use_edns0" global option.
832 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
835 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
836 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
838 TK/01 Updated PolarSSL code to 0.14.2.
839 Bugzilla 1097. Patch from Andreas Metzler.
841 PP/07 Catch divide-by-zero in ${eval:...}.
844 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
847 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
848 format-string attack -- SECURITY: remote arbitrary code execution.
850 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
851 time unintentionally subject to list matching rules, letting the header
852 cause arbitrary Exim lookups (of items which can occur in lists, *not*
853 arbitrary string expansion). This allowed for information disclosure.
855 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
856 INT_MIN/-1 -- value coerced to INT_MAX.
862 NM/01 Workround for PCRE version dependency in version reporting
865 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
866 This fixes portability to compilers other than gcc, notably
867 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
869 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
870 makefiles for portability to HP-UX and POSIX correctness.
872 PP/01 Permit LOOKUP_foo enabling on the make command-line.
873 Also via indented variable definition in the Makefile.
874 (Debugging by Oliver Heesakkers).
876 PP/02 Restore caching of spamd results with expanded spamd_address.
877 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
879 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
880 Improves build reliability. Fix from: Frank Elsner
882 NM/02 Fix wide character breakage in the rfc2047 coding
883 Fixes bug 1064. Patch from Andrey N. Oktyabrski
885 NM/03 Allow underscore in dnslist lookups
886 Fixes bug 1026. Patch from Graeme Fowler
888 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
889 Code patches from Adam Ciarcinski of NetBSD.
891 NM/04 Fixed exiqgrep to cope with mailq missing size issue
894 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
895 is logged, to avoid truncation. Patch from John Horne.
897 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
898 Patch from Jakob Hirsch.
900 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
901 SQL string expansion failure details.
902 Patch from Andrey Oktyabrski.
904 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
905 Patch from Simon Arlott.
907 PP/09 New lookups functionality failed to compile on old gcc which rejects
908 extern declarations in function scope.
909 Patch from Oliver Fleischmann
911 PP/10 Use sig_atomic_t for flags set from signal handlers.
912 Check getgroups() return and improve debugging.
913 Fixed developed for diagnosis in bug 927 (which turned out to be
916 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
917 Patch from Mark Zealey.
919 PP/12 Bugzilla 1056: Improved spamd server selection.
920 Patch from Mark Zealey.
922 PP/13 Bugzilla 1086: Deal with maildir quota file races.
923 Based on patch from Heiko Schlittermann.
925 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
926 Patch from Uwe Doering, sign-off by Michael Haardt.
928 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
929 variable declaration deep within a block. Bug and patch from
932 PP/15 lookups-Makefile IRIX compatibilty coercion.
934 PP/16 Make DISABLE_DKIM build knob functional.
936 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
937 Patch by Simon Arlott
939 TF/03 Fix valgrind.h portability to C89 compilers that do not support
940 variable argument macros. Our copy now differs from upstream.
946 TF/01 Failure to get a lock on a hints database can have serious
947 consequences so log it to the panic log.
949 TF/02 Log LMTP confirmation messages in the same way as SMTP,
950 controlled using the smtp_confirmation log selector.
952 TF/03 Include the error message when we fail to unlink a spool file.
954 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
955 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
956 for maintaining out-of-tree patches for some time.
958 PP/01 Bugzilla 139: Documentation and portability issues.
959 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
960 Handle per-OS dynamic-module compilation flags.
962 PP/02 Let /dev/null have normal permissions.
963 The 4.73 fixes were a little too stringent and complained about the
964 permissions on /dev/null. Exempt it from some checks.
965 Reported by Andreas M. Kirchwitz.
967 PP/03 Report version information for many libraries, including
968 Exim version information for dynamically loaded libraries. Created
969 version.h, now support a version extension string for distributors
970 who patch heavily. Dynamic module ABI change.
972 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
973 privilege escalation vulnerability whereby the Exim run-time user
974 can cause root to append content of the attacker's choosing to
977 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
980 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
981 If dropping privileges for untrusted macros, we disabled normal logging
982 on the basis that it would fail; for the Exim run-time user, this is not
983 the case, and it resulted in successful deliveries going unlogged.
984 Fixed. Reported by Andreas Metzler.
990 PP/01 Date: & Message-Id: revert to normally being appended to a message,
991 only prepend for the Resent-* case. Fixes regression introduced in
992 Exim 4.70 by NM/22 for Bugzilla 607.
994 PP/02 Include check_rfc2047_length in configure.default because we're seeing
995 increasing numbers of administrators be bitten by this.
997 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
999 PP/03 Bugzilla 994: added openssl_options main configuration option.
1001 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1003 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1005 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1007 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1008 without a peer certificate, leading to a segfault because of an
1009 assumption that peers always have certificates. Be a little more
1010 paranoid. Problem reported by Martin Tscholak.
1012 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1013 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1014 NB: ClamAV planning to remove STREAM in "middle of 2010".
1015 CL also introduces -bmalware, various -d+acl logging additions and
1016 more caution in buffer sizes.
1018 PP/09 Implemented reverse_ip expansion operator.
1020 PP/10 Bugzilla 937: provide a "debug" ACL control.
1022 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1024 PP/12 Bugzilla 973: Implement --version.
1026 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1028 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1030 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1032 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1033 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1034 ignore trailing whitespace.
1036 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1038 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1041 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1042 Notification from Dr Andrew Aitchison.
1044 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1045 ExtendedDetectionInfo response format.
1046 Notification from John Horne.
1048 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1051 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1052 XSL and documented dependency on system catalogs, with examples of how
1055 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1058 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1059 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1060 configuration file which is writeable by the Exim user or group.
1062 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1063 of configuration files to cover files specified with the -C option if
1064 they are going to be used with root privileges, not just the default
1067 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1068 option (effectively making it always true).
1070 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1071 files to be used while preserving root privileges.
1073 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1074 that rogue child processes cannot use them.
1076 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1077 run-time user, instead of root.
1079 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1080 Exim run-time user without dropping privileges.
1082 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1083 result string, instead of calling string_vformat() twice with the same
1086 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1087 for other users. Others should always drop root privileges if they use
1088 -C on the command line, even for a whitelisted configure file.
1090 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1092 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1098 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1099 $data_path, and $header_path variables; fixed documentation bugs and
1102 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1103 exipick to access non-standard spools, including the "frozen" queue
1106 NM/01 Bugzilla 965: Support mysql stored procedures.
1107 Patch from Alain Williams
1109 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1111 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1112 Patch from Andreas Metzler
1114 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1115 Patch from Kirill Miazine
1117 NM/05 Bugzilla 671: Added umask to procmail example.
1119 JJ/03 installed exipick 20100323.0, fixing doc bug
1121 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1122 directory. Notification and patch from Dan Rosenberg.
1124 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1126 TK/02 Improve log output when DKIM signing operation fails.
1128 MH/01 Treat the transport option dkim_domain as a colon separated
1129 list, not as a single string, and sign the message with each element,
1130 omitting multiple occurences of the same signer.
1132 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1133 Bugzilla 985, 986. Patch by Simon Arlott
1135 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1136 Patch by Simon Arlott
1138 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1139 MBX locking. Notification from Dan Rosenberg.
1145 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1147 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1149 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1151 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1153 NM/04 Bugzilla 915: Flag broken perl installation during build.
1159 TK/01 Added patch by Johannes Berg that expands the main option
1160 "spamd_address" if it starts with a dollar sign.
1162 TK/02 Write list of recipients to X-Envelope-Sender header when building
1163 the mbox-format spool file for content scanning (suggested by Jakob
1166 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1167 (http://www.dcc-servers.net/) support via dccifd. Activated by
1168 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1170 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1171 by Mark Daniel Reidel <mr@df.eu>.
1173 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1174 When building exim an external PCRE library is now needed -
1175 PCRE is a system library on the majority of modern systems.
1176 See entry on PCRE_LIBS in EDITME file.
1178 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1179 conversation. Added nologin parameter to request.
1180 Patch contributed by Kirill Miazine.
1182 TF/01 Do not log submission mode rewrites if they do not change the address.
1184 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1186 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1187 log files in place. Contributed by Roberto Lima.
1189 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1191 TF/03 Bugzilla 615: When checking the local_parts router precondition
1192 after a local_part_suffix or local_part_prefix option, Exim now
1193 does not use the address's named list lookup cache, since this
1194 contains cached lookups for the whole local part.
1196 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1197 Robert Millan. Documentation is in experimental-spec.txt.
1199 TF/04 Bugzilla 668: Fix parallel build (make -j).
1201 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1203 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1204 Patch provided by Jan Srzednicki.
1206 TF/05 Leading white space used to be stripped from $spam_report which
1207 wrecked the formatting. Now it is preserved.
1209 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1210 that they are available at delivery time.
1212 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1214 TF/08 TLS error reporting now respects the incoming_interface and
1215 incoming_port log selectors.
1217 TF/09 Produce a more useful error message if an SMTP transport's hosts
1218 setting expands to an empty string.
1220 NM/06 Bugzilla 744: EXPN did not work under TLS.
1221 Patch provided by Phil Pennock.
1223 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1224 Patch provided by Richard Godbee.
1226 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1227 acl_smtp_notquit, added index entry.
1229 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1230 Patch provided by Eugene Bujak.
1232 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1233 accept(). Patch provided by Maxim Dounin.
1235 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1236 Patch provided by Phil Pennock.
1238 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1240 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1241 Patch provided by Brad "anomie" Jorsch.
1243 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1244 Patch provided by Dean Brooks.
1246 TK/05 Add native DKIM support (does not depend on external libraries).
1248 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1249 Patch provided by Graeme Fowler.
1251 NM/16 Bugzilla 851: Documentation example syntax fix.
1253 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1255 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1258 NM/19 Bugzilla 745: TLS version reporting.
1259 Patch provided by Phil Pennock.
1261 NM/20 Bugzilla 167: bool: condition support.
1262 Patch provided by Phil Pennock.
1264 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1265 clients. Patch provided by Phil Pennock.
1267 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1268 Patch provided by Brad "anomie" Jorsch.
1270 NM/23 Bugzilla 687: Fix misparses in eximstats.
1271 Patch provided by Heiko Schlittermann.
1273 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1274 Patch provided by Heiko Schlittermann.
1276 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1277 plus update to original patch.
1279 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1281 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1282 Patch provided by David Brownlee.
1284 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1286 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1288 NM/30 Bugzilla 888: TLS documentation bugfixes.
1290 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1292 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1293 Unlike the original bugzilla I have changed all shell scripts in src tree.
1295 NM/33 Bugzilla 898: Transport filter timeout fix.
1296 Patch by Todd Rinaldo.
1298 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1299 Patch by Serge Demonchaux.
1301 NM/35 Bugzilla 39: Base64 decode bug fixes.
1302 Patch by Jakob Hirsch.
1304 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1306 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1308 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1310 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1316 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1317 ALT-N's libdkim that I have put here:
1318 http://duncanthrax.net/exim-experimental/
1320 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1321 were called 'true' and it seems that C99 defines that as a reserved
1322 keyword to be used with 'bool' variable types. That means you could
1323 not include C99-style headers which use bools without triggering
1324 build errors in sieve.c.
1326 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1327 as mailq or other aliases. Changed the --help handling significantly
1328 to do whats expected. exim_usage() emits usage/help information.
1330 SC/01 Added the -bylocaldomain option to eximstats.
1332 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1334 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1336 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1342 PH/01 Another patch from the Sieve maintainer.
1344 PH/02 When an IPv6 address is converted to a string for single-key lookup
1345 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1346 dots are used instead of colons so that keys in lsearch files need not
1347 contain colons. This was done some time before quoting was made available
1348 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1349 (notated using the quote facility) so as to distinguish them from IPv4
1350 keys. This meant that lookups for IP addresses in host lists did not work
1351 for iplsearch lookups.
1353 This has been fixed by arranging for IPv6 addresses to be expressed with
1354 colons if the lookup type is iplsearch. This is not incompatible, because
1355 previously such lookups could never work.
1357 The situation is now rather anomolous, since one *can* have colons in
1358 ordinary lsearch keys. However, making the change in all cases is
1359 incompatible and would probably break a number of configurations.
1361 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1364 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1365 conversion specification without a maximum field width, thereby enabling
1366 a rogue spamd server to cause a buffer overflow. While nobody in their
1367 right mind would setup Exim to query an untrusted spamd server, an
1368 attacker that gains access to a server running spamd could potentially
1369 exploit this vulnerability to run arbitrary code as the Exim user.
1371 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1372 $primary_hostname instead of what libspf2 thinks the hosts name is.
1374 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1375 a directory entry by the name of the lookup key. Previously, if a
1376 symlink pointed to a non-existing file or a file in a directory that
1377 Exim lacked permissions to read, a lookup for a key matching that
1378 symlink would fail. Now it is enough that a matching directory entry
1379 exists, symlink or not. (Bugzilla 503.)
1381 PH/03 The body_linecount and body_zerocount variables are now exported in the
1384 PH/04 Added the $dnslist_matched variable.
1386 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1387 This means they are set thereafter only if the connection becomes
1390 PH/06 Added the client_condition to authenticators so that some can be skipped
1391 by clients under certain conditions.
1393 PH/07 The error message for a badly-placed control=no_multiline_responses left
1394 "_responses" off the end of the name.
1396 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1398 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1399 (without spaces) instead of just copying the configuration text.
1401 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1403 PH/11 Added $max_received_linelength.
1405 PH/12 Added +ignore_defer and +include_defer to host lists.
1407 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1408 way in which PCRE > 7.0 is built.
1410 PH/14 Implemented queue_only_load_latch.
1412 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1413 MAIL command. The effect was to mangle the value on 64-bit systems.
1415 PH/16 Another patch from the Sieve maintainer.
1417 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1419 PH/18 If a system quota error occurred while trying to create the file for
1420 a maildir delivery, the message "Mailbox is full" was not appended to the
1421 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1422 only to a quota excession during the actual writing of the file.
1424 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1425 characters?) which causes problems in log lines. The DN values are now
1426 passed through string_printing() before being added to log lines.
1428 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1429 and InterBase are left for another time.)
1431 PH/21 Added message_body_newlines option.
1433 PH/22 Guard against possible overflow in moan_check_errorcopy().
1435 PH/23 POSIX allows open() to be a macro; guard against that.
1437 PH/24 If the recipient of an error message contained an @ in the local part
1438 (suitably quoted, of course), incorrect values were put in $domain and
1439 $local_part during the evaluation of errors_copy.
1445 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1446 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1449 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1450 issue a MAIL command.
1452 PH/02 In an ACL statement such as
1454 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1456 if a client was not listed at all, or was listed with a value other than
1457 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1458 the condition was not true (as it should be), so access was not denied.
1459 The bug was that the ! inversion was incorrectly passed on to the second
1460 item. This has been fixed.
1462 PH/03 Added additional dnslists conditions == and =& which are different from
1463 = and & when the dns lookup returns more than one IP address.
1465 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1466 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1468 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1469 FSYNC, which compiles an option called disable_fsync that allows for
1470 bypassing fsync(). The documentation is heavily laced with warnings.
1472 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1474 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1475 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1476 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1477 including adding "make clean"; (3) Added -fPIC when compiling the test
1478 dynamically loaded module, to get rid of a warning.
1480 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1481 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1482 = 0s. The bug is otherwise harmless.
1484 PH/07 There was a bug in the dovecot authenticator such that the value of
1485 $auth1 could be overwritten, and so not correctly preserved, after a
1486 successful authentication. This usually meant that the value preserved by
1487 the server_setid option was incorrect.
1489 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1491 PH/09 Installed PCRE release 7.0.
1493 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1494 run for batched SMTP input. It is now run at the start of every message
1495 in the batch. While fixing this I discovered that the process information
1496 (output by running exiwhat) was not always getting set for -bs and -bS
1497 input. This is fixed, and it now also says "batched" for BSMTP.
1499 PH/11 Added control=no_pipelining.
1501 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1502 patch, slightly modified), and move the expansion of helo_data till after
1503 the connection is made in the smtp transport (so it can use these
1506 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1508 PH/14 Added log_selector = +pid.
1510 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1512 PH/16 Add ${if forany and ${if forall.
1514 PH/17 Added dsn_from option to vary the From: line in DSNs.
1516 PH/18 Flush SMTP output before performing a callout, unless control =
1517 no_callout_flush is set.
1519 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1520 was true (the default) a successful delivery failed to delete the retry
1521 item, thus causing premature timeout of the address. The bug is now
1524 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1526 PH/21 Long custom messages for fakedefer and fakereject are now split up
1527 into multiline reponses in the same way that messages for "deny" and
1528 other ACL rejections are.
1530 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1531 with slight modification.
1533 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1534 draft, changing the syntax and factoring some duplicate code".
1536 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1537 for deliveries of the second and subsequent messages over the same SMTP
1540 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1541 ${reduce, with only minor "tidies".
1543 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1545 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1546 expansion side effects.
1548 PH/27 When a message times out after an over-quota error from an Exim-imposed
1549 quota, the bounce message says "mailbox is full". This message was not
1550 being given when it was a system quota that was exceeded. It now should
1553 MH/03 Made $recipients available in local_scan(). local_scan() already has
1554 better access to the recipient list through recipients_list[], but
1555 $recipients can be useful in postmaster-provided expansion strings.
1557 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1558 in the case of a MAIL command with additional options following the
1559 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1560 were accidentally chopped off.
1562 PH/29 SMTP synchronization checks are implemented when a command is read -
1563 there is a check that no more input is waiting when there shouldn't be
1564 any. However, for some commands, a delay in an ACL can mean that it is
1565 some time before the response is written. In this time, more input might
1566 arrive, invalidly. So now there are extra checks after an ACL has run for
1567 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1568 pipelining has not been advertised.
1570 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1572 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1573 returned in the SMTP response when smtp_return_error_details was false.
1574 This has been fixed.
1576 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1577 instead of the C I/O that was originally supplied, because problems were
1578 reported on Solaris.
1580 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1581 Exim which did not show up earlier: it was assuming that a call to
1582 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1583 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1584 was a macro that became an assignment, so it seemed to work. This has
1585 changed to a proper function call with a void return, hence the compile
1586 error. Exim's code has been fixed.
1588 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1591 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1593 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1595 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1596 because it turns out to be too big or there is a timeout) there is no
1597 "Completed" line in the log. When some messages of this type were
1598 selected by exigrep, they were listed as "not completed". Others were
1599 picked up by some special patterns. I have improved the selection
1600 criteria to be more general.
1602 PH/38 The host_find_failed option in the manualroute router can now be set
1603 to "ignore", to completely ignore a host whose IP address cannot be
1604 found. If all hosts are ignored, the behaviour is controlled by the new
1605 host_all_ignored option.
1607 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1608 homing or because of multiple MX records with /mx) generated more than
1609 one IP address, and the following item turned out to be the local host,
1610 all the secondary addresses of the first item were incorrectly removed
1611 from the list, along with the local host and any following hosts (which
1612 is what is supposed to happen).
1614 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1615 whoever called Exim into the -H file. In the case of the daemon it was
1616 behaving confusingly. When first started, it used values for whoever
1617 started the daemon, but after a SIGHUP it used the Exim user (because it
1618 calls itself on a restart). I have changed the code so that it now always
1621 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1622 message are rejected with the same error (e.g. no authentication or bad
1623 sender address), and a DATA command is nevertheless sent (as can happen
1624 with PIPELINING or a stupid MUA), the error message that was given to the
1625 RCPT commands is included in the rejection of the DATA command. This is
1626 intended to be helpful for MUAs that show only the final error to their
1629 PH/42 Another patch from the Sieve maintainer.
1631 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1632 Eximstats - Fixed some broken HTML links and added missing column headers
1634 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1635 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1637 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1639 PH/43 Yet another patch from the Sieve maintainer.
1641 PH/44 I found a way to check for a TCP/IP connection going away before sending
1642 the response to the final '.' that terminates a message, but only in the
1643 case where the client has not sent further data following the '.'
1644 (unfortunately, this is allowed). However, in many cases there won't be
1645 any further data because there won't be any more messages to send. A call
1646 to select() can be used: if it shows that the input is "ready", there is
1647 either input waiting, or the socket has been closed. An attempt to read
1648 the next input character can distinguish the two cases. Previously, Exim
1649 would have sent an OK response which the client would never have see.
1650 This could lead to message repetition. This fix should cure that, at
1651 least in a lot of common cases.
1653 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1654 advertised in response to EHLO.
1660 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1661 fixed by 4.65/MH/01 (is this a record?) are fixed:
1663 (i) An empty string was always treated as zero by the numeric comparison
1664 operators. This behaviour has been restored.
1666 (ii) It is documented that the numeric comparison operators always treat
1667 their arguments as decimal numbers. This was broken in that numbers
1668 starting with 0 were being interpreted as octal.
1670 While fixing these problems I realized that there was another issue that
1671 hadn't been noticed. Values of message_size_limit (both the global option
1672 and the transport option) were treated as octal if they started with 0.
1673 The documentation was vague. These values are now always treated as
1674 decimal, and I will make that clear in the documentation.
1680 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1681 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1684 MH/01 Don't check that the operands of numeric comparison operators are
1685 integers when their expansion is in "skipping" mode (fixes bug
1686 introduced by 4.64-PH/07).
1688 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1689 child addresses, Exim now panics and dies. Previously, because the count
1690 is held in a short int, deliveries were likely to be lost. As such a
1691 large number of recipients for a single message is ridiculous
1692 (performance will be very, very poor), I have chosen to impose a limit
1693 rather than extend the field.
1699 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1700 leftover -K file (the existence of which was triggered by #402).
1701 While we were at it, introduced process PID as part of the -K
1702 filename. This should rule out race conditions when creating
1705 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1706 processing considerably. Previous code took too long for large mails,
1707 triggering a timeout which in turn triggers #401.
1709 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1710 in the DK code in transports.c. sendfile() is not really portable,
1711 hence the _LINUX specificness.
1713 TF/01 In the add_headers option to the mail command in an Exim filter,
1714 there was a bug that Exim would claim a syntax error in any
1715 header after the first one which had an odd number of characters
1718 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1719 callout verification, Exim cached a "reject" for the entire domain. This
1720 is correct for most verifications, but it is not correct for a recipient
1721 verification with use_sender or use_postmaster set, because in that case
1722 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1723 case of MAIL FROM:<> rejection from other early rejections (e.g.
1724 rejection of HELO). When verifying a recipient using a non-null MAIL
1725 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1726 Whatever the result of the callout, the value of the domain cache is
1727 left unchanged (for any other kind of callout, getting as far as trying
1728 RCPT means that the domain itself is ok).
1730 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1733 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1734 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1735 session, the daemon ignores ECONNECT errors and logs others; it now
1736 ignores EPIPE as well.
1738 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1739 (quoted-printable decoding).
1741 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1742 later the small subsequent patch to fix an introduced bug.
1744 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1746 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1748 PH/08 An error is now given if message_size_limit is specified negative.
1750 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1751 to be given (somewhat) arbitrary names.
1753 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1756 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1757 miscellaneous code fixes
1759 PH/10 Added the log_reject_target ACL modifier to specify where to log
1762 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1763 hostname. This is wrong, because it relates to the incoming message (and
1764 probably the interface on which it is arriving) and not to the outgoing
1765 callout (which could be using a different interface). This has been
1766 changed to use the value of the helo_data option from the smtp transport
1767 instead - this is what is used when a message is actually being sent. If
1768 there is no remote transport (possible with a router that sets up host
1769 addresses), $smtp_active_hostname is used.
1771 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1772 tweaks were necessary in order to get it to work (see also 21 below):
1773 (a) The code assumed that strncpy() returns a negative number on buffer
1774 overflow, which isn't the case. Replaced with Exim's string_format()
1776 (b) There were several signed/unsigned issues. I just did the minimum
1777 hacking in of casts. There is scope for a larger refactoring.
1778 (c) The code used strcasecmp() which is not a standard C function.
1779 Replaced with Exim's strcmpic() function.
1780 (d) The code set only $1; it now sets $auth1 as well.
1781 (e) A simple test gave the error "authentication client didn't specify
1782 service in request". It would seem that Dovecot has changed its
1783 interface. Fortunately there's a specification; I followed it and
1784 changed what the client sends and it appears to be working now.
1786 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1789 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1790 address A is aliased to B and C, where B exists and C does not. Without
1791 -v the output is "A verified" because verification stops after a
1792 successful redirection if more than one address is generated. However,
1793 with -v the child addresses are also verified. Exim was outputting "A
1794 failed to verify" and then showing the successful verification for C,
1795 with its parentage. It now outputs "B failed to verify", showing B's
1796 parentage before showing the successful verification of C.
1798 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1799 look up a TXT record in a specific list after matching in a combined
1802 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1803 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1804 they consult the DNS. I had assumed they would set it the way they
1805 wanted; and indeed my experiments on Linux seem to show that in some
1806 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1807 To be on the safe side, however, I have now made the interface to
1808 host_find_byname() similar to host_find_bydns(), with an argument
1809 containing the DNS resolver options. The host_find_byname() function now
1810 sets these options at its start, just as host_find_bydns() does. The smtp
1811 transport options dns_qualify_single and dns_search_parents are passed to
1812 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1813 of host_find_byname() use the default settings of RES_DEFNAMES
1814 (qualify_single) but not RES_DNSRCH (search_parents).
1816 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1817 spool_read_header() do less string testing, by means of a preliminary
1818 switch on the second character of optional "-foo" lines. (This is
1819 overdue, caused by the large number of possibilities that now exist.
1820 Originally there were few.) While I was there, I also converted the
1821 str(n)cmp tests so they don't re-test the leading "-" and the first
1822 character, in the hope this might squeeze out yet more improvement.
1824 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1825 flag allowing group syntax was set by the header_syntax check but not
1826 turned off, possible causing trouble later; (2) The flag was not being
1827 set at all for the header_verify test, causing "group"-style headers to
1828 be rejected. I have now set it in this case, and also caused header_
1829 verify to ignore an empty address taken from a group. While doing this, I
1830 came across some other cases where the code for allowing group syntax
1831 while scanning a header line wasn't quite right (mostly, not resetting
1832 the flag correctly in the right place). These bugs could have caused
1833 trouble for malformed header lines. I hope it is now all correct.
1835 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1836 with the "reply" argument non-NULL. The code, however (which originally
1837 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1838 but it didn't always do it. This confused somebody who was copying the
1839 code for some other use. I have removed all the tests.
1841 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1842 feature that was used to support insecure browsers during the U.S. crypto
1843 embargo. It requires special client support, and Exim is probably the
1844 only MTA that supported it -- and would never use it because real RSA is
1845 always available. This code has been removed, because it had the bad
1846 effect of slowing Exim down by computing (never used) parameters for the
1847 RSA_EXPORT functionality.
1849 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1850 authenticator to fail if there's a tab character in the incoming data
1851 (there should never be unless someone is messing about, as it's supposed
1852 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1853 if the connection is using TLS or if the remote IP is the same as the
1854 local IP, and the "valid-client-cert option" if a client certificate has
1857 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1858 authenticators. This can be used for authorization after authentication
1859 succeeds. (In the case of plaintext, it servers for both authentication
1862 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1863 if any retry times were supplied.
1865 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1866 connection, where there is no client IP address to check. In this
1867 situation, the verify now always succeeds.
1869 PH/25 Applied John Jetmore's -Mset patch.
1871 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1873 PH/27 In a string expansion for a processed (not raw) header when multiple
1874 headers of the same name were present, leading whitespace was being
1875 removed from all of them, but trailing whitespace was being removed only
1876 from the last one. Now trailing whitespace is removed from each header
1877 before concatenation. Completely empty headers in a concatenation (as
1878 before) are ignored.
1880 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1881 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1883 PH/29 [Removed. This was a change that I later backed out, and forgot to
1884 correct the ChangeLog entry (that I had efficiently created) before
1885 committing the later change.]
1887 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1888 address errors (4xx response to RCPT) over the same connection as other
1889 messages routed to the same hosts. Such deliveries are always "forced",
1890 so retry times are not inspected. This resulted in far too many retries
1891 for the affected addresses. The effect occurred only when there were more
1892 hosts than the hosts_max_try setting in the smtp transport when it had
1893 the 4xx errors. Those hosts that it had tried were not added to the list
1894 of hosts for which the message was waiting, so if all were tried, there
1895 was no problem. Two fixes have been applied:
1897 (i) If there are any address or message errors in an SMTP delivery, none
1898 of the hosts (tried or untried) are now added to the list of hosts
1899 for which the message is waiting, so the message should not be a
1900 candidate for sending over the same connection that was used for a
1901 successful delivery of some other message. This seems entirely
1902 reasonable: after all the message is NOT "waiting for some host".
1903 This is so "obvious" that I'm not sure why it wasn't done
1904 previously. Hope I haven't missed anything, but it can't do any
1905 harm, as the worst effect is to miss an optimization.
1907 (ii) If, despite (i), such a delivery is accidentally attempted, the
1908 routing retry time is respected, so at least it doesn't keep
1909 hammering the server.
1911 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1912 in ${readsocket because some servers need this prod.
1914 PH/32 Added some extra debug output when updating a wait-xxx database.
1916 PH/33 The hint "could be header name not terminated by colon", which has been
1917 given for certain expansion errors for a long time, was not being given
1918 for the ${if def:h_colon_omitted{... case.
1920 PH/34 The spec says: "With one important exception, whenever a domain list is
1921 being scanned, $domain contains the subject domain." There was at least
1922 one case where this was not true.
1924 PH/35 The error "getsockname() failed: connection reset by peer" was being
1925 written to the panic log as well as the main log, but it isn't really
1926 panic-worthy as it just means the connection died rather early on. I have
1927 removed the panic log writing for the ECONNRESET error when getsockname()
1930 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1931 runs only) independently of the message's sender address. This meant
1932 that, if the 4xx error was in fact related to the sender, a different
1933 message to the same recipient with a different sender could confuse
1934 things. In particualar, this can happen when sending to a greylisting
1935 server, but other circumstances could also provoke similar problems.
1936 I have changed the default so that the retry time for these errors is now
1937 based a combination of the sender and recipient addresses. This change
1938 can be overridden by setting address_retry_include_sender=false in the
1941 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1942 remote server are returned as part of bounce messages. This was not
1943 happening for LMTP over a pipe (the lmtp transport), but now it is the
1944 same for both kinds of LMTP.
1946 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1947 in header lines that were in fact CNAMEs. This is no longer the case.
1949 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1950 and queue runs started by the daemon processed all messages. This has
1951 been fixed so that -R and -S can now usefully be given with -q<time>.
1953 PH/40 Import PCRE release 6.7 (fixes some bugs).
1955 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1957 PH/42 Give an error if -q is specified more than once.
1959 PH/43 Renamed the variables $interface_address and $interface_port as
1960 $received_ip_address and $received_port, to make it clear that these
1961 values apply to message reception, and not to the outgoing interface when
1962 a message is delivered. (The old names remain recognized, of course.)
1964 PH/44 There was no timeout on the connect() call when using a Unix domain
1965 socket in the ${readsocket expansion. There now is.
1967 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1968 be meaningful with "accept".
1970 SC/01 Eximstats V1.43
1971 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1973 SC/02 Eximstats V1.44
1974 Use a glob alias rather than an array ref in the generated
1975 parser. This improves both readability and performance.
1977 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1978 Collect SpamAssassin and rejection statistics.
1979 Don't display local sender or destination tables unless
1980 there is data to show.
1981 Added average volumes into the top table text output.
1983 SC/04 Eximstats V1.46
1984 Collect data on the number of addresses (recipients)
1985 as well as the number of messages.
1987 SC/05 Eximstats V1.47
1988 Added 'Message too big' to the list of mail rejection
1989 reasons (thanks to Marco Gaiarin).
1991 SC/06 Eximstats V1.48
1992 Mainlog lines which have GMT offsets and are too short to
1993 have a flag are now skipped.
1995 SC/07 Eximstats V1.49 (Alain Williams)
1996 Added the -emptyok flag.
1998 SC/08 Eximstats V1.50
1999 Fixes for obtaining the IP address from reject messages.
2001 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2002 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2003 whitesspace changes from 4.64-PH/27
2005 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2008 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2009 are found, allow negative numbers in numeric criteria)
2011 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2013 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2014 to match changes made in 4.64-PH/43
2016 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2018 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2019 "domain" is actually an IP address, and force a failure. This locks out
2020 those revolvers/nameservers that support "A-for-A" lookups, in
2021 contravention of the specifications.
2023 PH/48 When a host name was looked up from an IP address, and the subsequent
2024 forward lookup of the name timed out, the host name was left in
2025 $sender_host_name, contrary to the specification.
2027 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2028 restricted to single-key lookups, Exim was not diagnosing an error if
2029 * or *@ was used with a query-style lookup.
2031 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2033 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2034 long ago, but noone interested enough thought of it. Let's just say that
2035 the "1.1" means that there are some new functions that weren't there at
2036 some point in the past.
2038 PH/51 Error processing for expansion failure of helo_data from an smtp
2039 transport during callout processing was broken.
2041 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2042 tested/used via the -bh/-bhc/-bs options.
2044 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2045 bug, fixed in subsequent PCRE releases).
2047 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2048 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2050 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2056 SC/01 Use a glob alias rather than an array ref in eximstats generated
2057 parser. This improves both readability and performance.
2059 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2060 Don't display local sender or destination tables in eximstats unless
2061 there is data to show.
2062 Added average volumes into the eximstats top table text output.
2064 SC/03 Collect data on the number of addresses (recipients) as well
2065 as the number of messages in eximstats.
2067 TF/01 Correct an error in the documentation for the redirect router. Exim
2068 does (usually) call initgroups() when daemonizing.
2070 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2071 with consistent privilege compared to when running as a daemon.
2073 TF/03 Note in the spec that $authenticated_id is not set for local
2074 submissions from trusted users.
2076 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2077 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2079 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2080 by adding some example configuration directives to the default
2081 configuration file. A little bit of work is required to uncomment the
2082 directives and define how usernames and passwords are checked, but
2083 there is now a framework to start from.
2085 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2086 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2087 without this. I don't know how relevant this is to other LDAP libraries.
2089 PH/02 Add the verb name to the "unknown ACL verb" error.
2091 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2093 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2095 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2096 directory not expanded when it should be if an expanded home directory
2097 was set for the address (which is overridden by the transport).
2099 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2102 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2103 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2104 because it is too late at that time, and has no effect.
2106 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2107 security issue with \' (bugzilla #107). I could not use the
2108 PQescapeStringConn() function, because it needs a PGconn value as one of
2111 PH/08 When testing addresses using -bt, indicate those final addresses that
2112 are duplicates that would not cause an additional delivery. At least one
2113 person was confused, thinking that -bt output corresponded to deliveries.
2114 (Suppressing duplicates isn't a good idea as you lose the information
2115 about possibly different redirections that led to the duplicates.)
2117 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2118 systems where poll() doesn't work, in particular OS X.
2120 PH/10 Added more information to debugging output for retry time not reached.
2122 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2123 operations in malware.c.
2125 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2128 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2129 syslog_duplication set false, log lines that would normally be written
2130 both the the main log and to the reject log were not written to syslog at
2133 PH/14 In the default configuration, change the use of "message" in ACL warn
2134 statements to "add_header".
2136 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2137 not followed by a command (e.g. "seen endif").
2139 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2140 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2143 PH/17 Added extra conditions to the default value of delay_warning_condition
2147 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2148 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2149 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2152 The Auto-Submitted: and various List- headers are standardised, whereas I
2153 don't think Precedence: ever was.
2155 PH/18 Refactored debugging code in route_finduser() to show more information,
2156 in particular, the error code if getpwnam() issues one.
2158 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2159 This is apparently needed in addition to the PH/07 change above to avoid
2160 any possible encoding problems.
2162 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2163 but not after initializing Perl.
2165 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2166 output them only if debugging. By default they are written stderr,
2167 apparently, which is not desirable.
2169 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2172 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2175 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2177 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2178 authenticated or an ident call has been made. Suppress the default
2179 values for $authenticated_id and $authenticated_sender (but permit -oMai
2180 and -oMas) when testing with -bh.
2182 PH/24 Re-jigged the order of the tests in the default configuration so that the
2183 tests for valid domains and recipients precede the DNS black list and CSA
2184 tests, on the grounds that those ones are more expensive.
2186 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2187 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2188 This bug exists in every version of Exim that I still have, right back to
2191 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2192 However, an attempt to turn on case-sensitivity in a regex key by
2193 including (?-i) didn't work because the subject string was already
2194 lowercased, and the effects were non-intuitive. It turns out that a
2195 one-line patch can be used to allow (?-i) to work as expected.
2201 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2202 other effects) broke the use of negated acl sub-conditions.
2204 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2207 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2208 "Deny" causes Exim to reject the incoming connection with a 554 error.
2209 Unfortunately, if there is a major crisis, such as a disk failure,
2210 tcp-wrappers gives "deny", whereas what one would like would be some
2211 kind of temporary error. A kludge has been added to help with this.
2212 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2213 554 error is used if errno is still zero or contains ENOENT (which occurs
2214 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2217 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2219 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2220 errors. Otherwise a message that provokes a temporary error (when other
2221 messages do not) can cause a whole host to time out.
2223 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2224 addresses were routed directly to files or pipes from a redirect router.
2225 File deliveries just didn't batch; pipe deliveries might have suffered
2228 PH/06 A failure to get a lock for a hints database would erroneously always say
2229 "Failed to get write lock", even when it was really a read lock.
2231 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2232 of 0600. This has been changed to use the value of the lockfile_mode
2233 option (which defaults to 0600).
2235 PH/08 Applied small patch from the Sieve maintainer.
2237 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2238 folder from quota calculations, a direct delivery into this folder messed
2239 up the contents of the maildirsize file. This was because the regex was
2240 used only to exclude .Trash (or whatever) when the size of the mailbox
2241 was calculated. There was no check that a delivery was happening into an
2242 excluded directory. This bug has been fixed by ignoring all quota
2243 processing for deliveries into excluded directories.
2245 PH/10 Added the maildirfolder_create_regex option to appendfile.
2251 PH/01 The code for finding all the local interface addresses on a FreeBSD
2252 system running IPv6 was broken. This may well have applied to all BSD
2253 systems, as well as to others that have similar system calls. The broken
2254 code found IPv4 interfaces correctly, but gave incorrect values for the
2255 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2256 that it would not match correctly against @[] and not recognize the IPv6
2259 PH/02 The ipliteral router was not recognizing addresses of the form user@
2260 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2262 PH/03 Added disable_ipv6.
2264 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2265 decoded form, because it is most often used to construct To: headers
2266 lines in autoreplies, and the decoded form may well be syntactically
2267 invalid. However, $reply_address has leading white space removed, and all
2268 newlines turned into spaces so that the autoreply transport does not
2271 PH/05 If group was specified without a user on a router, and no group or user
2272 was specified on a transport, the group from the router was ignored.
2274 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2275 for visible compile-time settings that can be used to change these
2276 numbers, for those that want even more. Backwards compatibility with old
2277 spool files has been maintained. However, going back to a previous Exim
2278 release will lost any variables that are in spool files.
2280 PH/07 Two small changes when running in the test harness: increase delay when
2281 passing a TCP/IP connection to a new process, in case the original
2282 process has to generate a bounce, and remove special handling of
2283 127.0.0.2 (sic), which is no longer necessary.
2285 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2286 be the same on different OS.
2288 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2291 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2292 whether --show-vars was specified or not
2294 JJ/02 exipick: Added support for new ACL variable spool format introduced
2297 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2298 syntactically invalid From: or Reply-to: line, and a filter used this to
2299 generate an autoreply, and therefore failed to obtain an address for the
2300 autoreply, Exim could try to deliver to a non-existent relative file
2301 name, causing unrelated and misleading errors. What now happens is that
2302 it logs this as a hard delivery error, but does not attempt to create a
2305 PH/11 The exinext utility has a -C option for testing purposes, but although
2306 the given file was scanned by exinext itself; it wasn't being passed on
2307 when Exim was called.
2309 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2310 an end-of-file indication when reading a command response.
2312 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2313 compiled. In many other places in Exim, IPv6 addresses are always
2314 recognized, so I have changed this. It also means that IPv4 domain
2315 literals of the form [IPV4:n.n.n.n] are now always recognized.
2317 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2318 used if the router is not running as root, for example, when verifying at
2319 ACL time, or when using -bh. The debugging output from this situation was
2320 non-existent - all you got was a failure to exec. I have made two
2323 (a) Failures to set uid/gid, the current directory, or a process leader
2324 in a subprocess such as that created by queryprogram now generate
2325 suitable debugging ouput when -d is set.
2327 (b) The queryprogram router detects when it is not running as root,
2328 outputs suitable debugging information if -d is set, and then runs
2329 the subprocess without attempting to change uid/gid.
2331 PH/15 Minor change to Makefile for building test_host (undocumented testing
2334 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2335 additional section of a DNS packet that returns MX or SRV records.
2336 Instead, it always explicitly searches for A/AAAA records. This avoids
2337 major problems that occur when a DNS server includes only records of one
2338 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2339 fixed another bug: if SRV records were looked up and the corresponding
2340 address records were *not* found in the additional section, the port
2341 values from the SRV records were lost.
2343 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2344 using the correct key (the original address) when searching the retry
2345 rules in order to find which one to use for generating the retry hint.
2347 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2348 adding the default one. Similarly, if it contains a Reply-To: header, the
2349 errors_reply_to option, if set, is not used.
2351 PH/19 When calculating a retry time, Exim used to measure the "time since
2352 failure" by looking at the "first failed" field in the retry record. Now
2353 it does not use this if it is later than than the arrival time of the
2354 message. Instead it uses the arrival time. This makes for better
2355 behaviour in cases where some deliveries succeed, thus re-setting the
2356 "first failed" field. An example is a quota failure for a huge message
2357 when small messages continue to be delivered. Without this change, the
2358 "time since failure" will always be short, possible causing more frequent
2359 delivery attempts for the huge message than are intended.
2360 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2362 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2363 $1, $2, $3) because the numerical variables can be reset during some
2364 expansion items (e.g. "match"), thereby losing the authentication data.
2366 PH/21 Make -bV show the size of off_t variables so that the test suite can
2367 decide whether to run tests for quotas > 2G.
2369 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2370 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2371 filecount value is greater than 2G or if a quota value is greater than 2G
2372 on a system where the size of off_t is not greater than 4, a panic error
2375 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2376 never match. The debug and -bh output now contains an explicit error
2377 message indicating a malformed IPv4 address or mask.
2379 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2380 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2381 PH/23 above applies.
2383 PH/25 Do not write to syslog when running in the test harness. The only
2384 occasion when this arises is a failure to open the main or panic logs
2385 (for which there is an explicit test).
2387 PH/26 Added the /no_tell option to "control=freeze".
2389 PH/27 If a host name lookup failed very early in a connection, for example, if
2390 the IP address matched host_lookup and the reverse lookup yielded a name
2391 that did not have a forward lookup, an error message of the form "no IP
2392 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2393 could be logged. Now it outputs the IP address instead of "NULL".
2395 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2396 allows the sender and the authenticated sender to be set when
2397 submitting a message from within Exim. Since child_open_exim() is
2398 documented for local_scan(), the new function should be too.
2400 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2401 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2402 results in an empty string is now treated as unset.
2404 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2406 PH/31 Added sender_verify_fail logging option.
2408 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2409 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2410 tidied the source and removed it altogether.
2412 PH/33 When a queue run was abandoned because the load average was too high, a
2413 log line was always written; now it is written only if the queue_run log
2414 selector is set. In addition, the log line for abandonment now contains
2415 information about the queue run such as the pid. This is always present
2416 in "start" and "stop" lines but was omitted from the "abandon" line.
2418 PH/34 Omit spaces between a header name and the colon in the error message that
2419 is given when verify = headers_syntax fails (if there are lots of them,
2420 the message gets confusing).
2422 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2423 names, as there are now some PTR records that contain slashes. This check
2424 is only to protect against broken name servers that fall over on strange
2425 characters, so the fact that it applies to all lookups doesn't matter.
2427 PH/36 Now that the new test suite is complete, we can remove some of the
2428 special code in Exim that was needed for the old test suite. For example,
2429 sorting DNS records because real resolvers return them in an arbitrary
2430 order. The new test suite's fake resolver always returns records in the
2433 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2434 bounces) except when queue_only is set, to avoid logging races between
2435 the different processes.
2437 PH/38 Panic-die if .include specifies a non-absolute path.
2439 PH/39 A tweak to the "H" retry rule from its user.
2441 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2442 a label. They prevented compilation on older perls.
2444 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2445 a warning to be raised on newish perls.
2447 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2448 on queue. Changes to match documented behaviour of showing count of
2449 messages matching specified criteria.
2451 PH/40 Changed the default ident timeout from 30s to 5s.
2453 PH/41 Added support for the use of login_cap features, on those BSD systems
2454 that have them, for controlling the resources used by pipe deliveries.
2456 PH/42 The content-scanning code uses fopen() to create files in which to put
2457 message data. Previously it was not paying any attention to the mode of
2458 the files. Exim runs with umask(0) because the rest of the code creates
2459 files with open(), and sets the required mode explicitly. Thus, these
2460 files were ending up world-writeable. This was not a big issue, because,
2461 being within the spool directory, they were not world-accessible. I have
2462 created a function called modefopen, which takes an additional mode
2463 argument. It sets umask(777), creates the file, chmods it to the required
2464 mode, then resets the umask. All the relevant calls to fopen() in the
2465 content scanning code have been changed to use this function.
2467 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2468 to 24 hours. This avoids potential overflow problems when processing G
2469 and H retry rules. I suspect nobody ever tinkers with this value.
2471 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2473 PH/45 When the plaintext authenticator is running as a client, the server's
2474 challenges are checked to ensure they are valid base64 strings. By
2475 default, the authentication attempt is cancelled if an invalid string is
2476 received. Setting client_ignore_invalid_base64 true ignores these errors.
2477 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2478 they are received. Thus, the responses can be made to depend on the
2479 challenges. If an invalid string is ignored, an empty string is placed in
2482 PH/46 Messages that are created by the autoreply transport now contains a
2483 References: header, in accordance with RFCs 2822 and 3834.
2485 PH/47 Added authenticated_sender_force to the smtp transport.
2487 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2489 PH/49 Installed latest patch from the Sieve maintainer.
2491 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2492 was also set, the appendfile transport was unnecessarily scanning a
2493 directory of message files (e.g. for maildir delivery) to find the count
2494 of files (along with the size), even though it did not need this
2495 information. It now does the scan only if it needs to find either the
2496 size of the count of files.
2498 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2500 PH/52 Two bugs concerned with error handling when the smtp transport is
2503 (i) Exim was not creating retry information for temporary errors given
2504 for individual recipients after the DATA command when the smtp transport
2505 was used in LMTP mode. This meant that they could be retried too
2506 frequently, and not timed out correctly.
2508 (ii) Exim was setting the flag that allows error details to be returned
2509 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2510 recipients that were returned after the DATA command.
2512 PH/53 This is related to PH/52, but is more general: for any failing address,
2513 when detailed error information was permitted to be returned to the
2514 sender, but the error was temporary, then after the final timeout, only
2515 "retry timeout exceeded" was returned. Now it returns the full error as
2516 well as "retry timeout exceeded".
2518 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2519 do this, and (what is worse) MTAs that accept it.
2521 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2522 will now be deprecated.
2524 PH/56 New os.c-cygwin from the Cygwin maintainer.
2526 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2527 formats (previously only available in exim formats via -bpr, -bpru,
2528 and -bpra. Now also available in native and exiqgrep formats)
2530 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2531 with very large, slow to parse queues
2533 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2535 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2537 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2538 responses to authentication challenges, though it was showing the
2539 challenges; (ii) I've removed the CR characters from the debug output for
2542 PH/58 Allow for the insertion of a newline as well as a space when a string
2543 is turned into more than one encoded-word during RFC 2047 encoding. The
2544 Sieve code now uses this.
2546 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2547 data_4xx, lost_connection, tls_required.
2549 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2550 message was being sent as an SMTP response.
2552 PH/61 Add -l and -k options to exicyclog.
2554 PH/62 When verifying, if an address was redirected to one new address, so that
2555 verification continued, and the new address failed or deferred after
2556 having set something in $address_data, the value of $address_data was not
2557 passed back to the ACL. This was different to the case when no
2558 redirection occurred. The value is now passed back in both cases.
2560 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2561 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2562 use login_cap.h, so on its own it isn't the distinguishing feature. The
2563 new name refers directly to the setclassresources() function.
2565 PH/65 Added configuration files for NetBSD3.
2567 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2569 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2570 is preferred over IPv4.
2572 PH/68 The bounce_return_message and bounce_return_body options were not being
2573 honoured for bounces generated during the reception of non-SMTP messages.
2574 In particular, this applied to messages rejected by the ACL. This bug has
2575 been fixed. However, if bounce_return_message is true and bounce_return_
2576 body is false, the headers that are returned for a non-SMTP message
2577 include only those that have been read before the error was detected.
2578 (In the case of an ACL rejection, they have all been read.)
2580 PH/69 The HTML version of the specification is now built in a directory called
2581 spec_html instead of spec.html, because the latter looks like a path with
2582 a MIME-type, and this confuses some software.
2584 PH/70 Catch two compiler warnings in sieve.c.
2586 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2587 function verify_get_ident() calls ip_connect() to connect a socket, but
2588 if the "connect()" function timed out, ip_connect() used to close the
2589 socket. However, verify_get_ident() also closes the socket later, and in
2590 between Exim writes to the log, which may get opened at this point. When
2591 the socket was closed in ip_connect(), the log could get the same file
2592 descriptor number as the socket. This naturally causes chaos. The fix is
2593 not to close the socket in ip_connect(); the socket should be closed by
2594 the function that creates it. There was only one place in the code where
2595 this was missing, in the iplookup router, which I don't think anybody now
2596 uses, but I've fixed it anyway.
2598 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2599 well as to direct DNS lookups. Otherwise the handling of names in host
2600 lists is inconsistent and therefore confusing.
2606 PH/01 Two changes to the default runtime configuration:
2608 (1) Move the checks for relay_from_hosts and authenticated clients from
2609 after to before the (commented out) DNS black list checks.
2611 (2) Add control=submission to the relay_from_hosts and authenticated
2612 clients checks, on the grounds that messages accepted by these
2613 statements are most likely to be submissions.
2615 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2617 (1) Generate an error if the third argument for the ${prvs expansion is
2620 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2623 (3) Reset the variables that are obtained from the first argument of
2624 ${prvscheck and used in the second argument before leaving the code,
2625 because their memory is reclaimed, so using them afterwards may do
2628 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2629 one (it's much easier than Tom thought :-).
2631 (5) Because of (4), we can now allow for the use of $prvscheck_result
2632 inside the third argument.
2634 PH/03 For some reason, the default setting of PATH when running a command from
2635 a pipe transport was just "/usr/bin". I have changed it to
2638 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2639 anything to be listed in the output from -bV.
2641 PH/05 When a filter generated an autoreply, the entire To: header line was
2642 quoted in the delivery log line, like this:
2644 => >A.N.Other <ano@some.domain> <original@ddress> ...
2646 This has been changed so that it extracts the operative address. There
2647 may be more than one such address. If so, they are comma-separated, like
2650 => >ano@some.domain,ona@other.domain <original@ddress> ...
2652 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2653 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2654 not being looked up in the rDNS to get a host name, Exim was showing the
2655 IP address twice in Received: lines, even though the IP addresses were
2656 identical. For example:
2658 Received: from [1.2.3.4] (helo=[1.2.3.4])
2660 However, if the real host name was known, it was omitting the HELO data
2661 if it matched the actual IP address. This has been tidied up so that it
2662 doesn't show the same IP address twice.
2664 PH/07 When both +timestamp and +memory debugging was on, the value given by
2665 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2666 called by the debug printing, thereby overwriting the timestamp buffer.
2667 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2669 PH/08 When the original message was included in an autoreply transport, it
2670 always said "this is a copy of the message, including all the headers",
2671 even if body_only or headers_only was set. It now gives an appropriate
2674 PH/09 Applied a patch from the Sieve maintainer which:
2676 o fixes some comments
2677 o adds the (disabled) notify extension core
2678 o adds some debug output for the result of if/elsif tests
2679 o points to the current vacation draft in the documentation
2680 and documents the missing references header update
2684 o fixes a bug in processing the envelope test (when testing
2685 multiple envelope elements, the last element determinted the
2688 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2689 Electronic Mail") by including:
2691 Auto-submitted: auto-generated
2693 in the messages that it generates (bounce messages and others, such as
2694 warnings). In the case of bounce messages for non-SMTP mesages, there was
2695 also a typo: it was using "Auto_submitted" (underscore instead of
2696 hyphen). Since every message generated by Exim is necessarily in response
2697 to another message, thes have all been changed to:
2699 Auto-Submitted: auto-replied
2701 in accordance with these statements in the RFC:
2703 The auto-replied keyword:
2705 - SHOULD be used on messages sent in direct response to another
2706 message by an automatic process,
2708 - MUST NOT be used on manually-generated messages,
2710 - MAY be used on Delivery Status Notifications (DSNs) and Message
2711 Disposition Notifications (MDNs),
2713 - MUST NOT be used on messages generated by automatic or periodic
2714 processes, except for messages which are automatic responses to
2717 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2718 to the default Received: header definition.
2720 PH/12 Added log selector acl_warn_skipped (default on).
2722 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2723 variables because (a) they are in the wrong storage pool and (b) even if
2724 they were copied, it wouldn't work properly because of the caching.
2726 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2727 checking when decoding. Apparently there are clients that generate
2728 overlong encoded strings. Why am I not surprised?
2730 PH/15 If the first argument of "${if match_address" was not empty, but did not
2731 contain an "@" character, Exim crashed. Now it writes a panic log message
2732 and treats the condition as false.
2734 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2736 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2737 extension "envelope-auth". The code is finished and in agreement with
2738 other implementations, but there is no documentation so far and in fact,
2739 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2740 not changing the active code.
2742 Print executed "if" and "elsif" statements when debugging is used. This
2743 helps a great deal to understand what a filter does.
2745 Document more things not specified clearly in RFC3028. I had all this
2746 sorted out, when out of a sudden new issues came to my mind. Oops."
2748 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2751 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2752 canonical form (as documented). However, after a host name lookup from
2753 the IP address, check_host() was doing a simple string comparison with
2754 addresses acquired from the DNS when checking that the found name did
2755 have the original IP as one of its addresses. Since any found IPv6
2756 addresses are likely to be in abbreviated form, the comparison could
2757 fail. Luckily, there already exists a function for doing the comparison
2758 by converting both addresses to binary, so now that is used instead of
2759 the text comparison.
2761 PH/20 There was another similar case to PH/19, when a complete host name was
2762 given in a host list; looking up its IP address could give an abbreviated
2763 form, whereas the current host's name might or might not be abbreviated.
2764 The same fix has been applied.
2770 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2771 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2774 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2775 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2777 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2779 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2780 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2781 Also, for "Auto-Submitted", treat anything other than "no" as
2782 non-personal, in accordance with RFC 3834. (Previously it treated
2783 anything starting "auto-" as non-personal.)
2785 TF/01 The control=submission/name=... option had a problem with syntax
2786 errors if the name included a slash character. The /name= option
2787 now slurps the rest of the string, so it can include any characters
2788 but it must come last in the list of options (after /sender_retain
2791 PH/05 Some modifications to the interface to the fake nameserver for the new
2799 TK/01 Added the "success_on_redirect" address verification option. See
2800 NewStuff for rationale and an example.
2802 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2804 PH/02 Patch to exigrep to allow it to work on syslog lines.
2806 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2807 fread() to skip over the body file's header line, because in Cygwin the
2808 header line is locked and is inaccessible.
2810 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2811 co-exist for some time) to make it clear that it is the Exim ID that is
2812 referenced, not the Message-ID: header line.
2814 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2815 string_format() function, because snprintf() does not exist on all
2818 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2819 ${stat: expansion item.
2821 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2822 protocol synchronization error", to keep the pedants happy.
2824 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2825 well as for IRIX systems, when gcc is being used. See the host.c source
2828 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2830 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2833 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2834 transport and to the smtp transport in LMTP mode.
2836 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2838 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2839 run a filter in a subprocess. This could lead to confusion in subsequent
2840 lookups in the parent process. There should also be a search_tidyup() at
2841 the end of the subprocess.
2843 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2844 only if the host matched helo_try_verify_hosts, which caused the
2845 verification to occur when the EHLO/HELO command was issued. The ACL just
2846 tested the remembered result. Now, if a previous verification attempt has
2847 not happened, "verify = helo" does it there and then.
2849 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2851 TK/03 Fix log output including CR from clamd.
2853 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2854 exist provoked a memory error which could cause a segfault.
2856 PH/15 Installed PCRE 6.2
2858 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2860 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2861 of the problem. Specifically, suggested +O2 rather than +O1 for the
2864 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2866 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2867 record, the port was not correctly logged when the outgoing_port log
2868 selector was set (it logged the transort's default port).
2870 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2871 fallback_hosts, and "hosts" in the smtp transport.
2873 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2874 host errors such as "Connection refused".
2876 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2877 authentication with radiusclient 0.4.9:
2879 - Error returned from rc_read_config was caught wrongly
2880 - Username/password not passed on to radius server due to wrong length.
2882 The presumption is that some radiusclient API changes for 4.51/PH/17
2883 were not taken care of correctly. The code is still untested by me (my
2884 Linux distribution still has 0.3.2 of radiusclient), but it was
2885 contributed by a Radius user.
2887 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2888 expanding the "port" option of the smtp transport.
2890 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2891 while reading a MIME header. Thanks to Tom Hughes for a patch.
2893 PH/24 Include config.h inside local_scan.h so that configuration settings are
2896 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2897 that in an ACL for RCPT (for example), you can examine exactly what was
2900 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2901 commands, but it was not correctly comparing the address with the actual
2902 client host address. Thus, it would show the EHLO address in Received:
2903 header lines when this was not necessary.
2905 PH/27 Added the % operator to ${eval:}.
2907 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2908 it should be ignoring failures (because with -C, for example, it has lost
2909 privilege). It wasn't ignoring creation failures other than "already
2912 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2915 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2916 because an input line was too long, either on its own, or by virtue of
2917 too many continuations, the temporary file was not being removed, and the
2918 return code was incorrect.
2920 PH/31 Missing "BOOL" in function definition in filtertest.c.
2922 PH/32 Applied Sieve patches from the maintainer.
2924 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2926 PH/33 Added "verify = not_blind".
2928 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2929 Local/Makefile (with some defaults set). These are used in built scripts
2930 such as exicyclog, but they have never been used in the exim_install
2931 script (though there are many overriding facilities there). I have
2932 arranged that the exim_install script now takes note of these two
2935 PH/35 Installed configuration files for Dragonfly.
2937 PH/36 When a locally submitted message by a trusted user did not contain a
2938 From: header, and the sender address was obtained from -f or from an SMTP
2939 MAIL command, and the trusted user did not use -F to supply a sender
2940 name, $originator_name was incorrectly used when constructing a From:
2941 header. Furthermore, $originator_name was used for submission mode
2942 messages from external hosts without From: headers in a similar way,
2943 which is clearly wrong.
2945 PH/37 Added control=suppress_local_fixups.
2947 PH/38 When log_selector = +received_sender was set, and the addition of the
2948 sender made the log line's construction buffer exactly full, or one byte
2949 less than full, an overflow happened when the terminating "\n" was
2952 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2953 when the result of a list match is failure because a DNS lookup failed.
2955 PH/40 RM_COMMAND is now used in the building process.
2957 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2958 the "build-* directories that it finds.
2960 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2961 address in a domain literal was a prefix of an interface address.
2963 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2964 when verifying a sender address, unless rewrite_headers is false.
2966 PH/44 Wrote a long comment about why errors_to addresses are verified as
2967 recipients, not senders.
2969 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2970 the ratelimit ACL was added.
2972 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2974 PH/46 Added extra information about PostgreSQL errors to the error string.
2976 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2977 avoiding the need to install special zones in a real server. This is
2978 backwards compatible; if it can't find the fake resolver, it drops back.
2979 Thus, both old and new test suites can be run.
2981 TF/02 Added util/ratelimit.pl
2983 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2984 clock is set back in time.
2986 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2987 Candler <B.Candler@pobox.com>.
2989 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2990 OK for addresses that are the result of redirections.
2992 PH/48 A number of further additions for the benefit of the new test suite,
2993 including a fake gethostbyname() that interfaces to the fake DNS resolver
2996 TF/06 The fix for widen_domains has also been applied to qualify_single and
2997 search_parents which are the other dnslookup options that can cause
3000 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3003 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3005 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3006 service" if it encountered an MX record with an empty target hostname.
3007 The message is now "an MX or SRV record indicated no SMTP service".
3009 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3010 if the DNS of the sending site is misconfigured. This is quite a
3011 common situation. This change restores the behaviour of treating a
3012 helo verification defer as a failure.
3014 PH/51 If self=fail was set on a router, the bounce message did not include the
3015 actual error message.
3021 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3023 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3024 command itself ended in error, the underlying message about the transport
3025 filter timeout was being overwritten with the pipe command error. Now the
3026 underlying error message should be appended to the second error message.
3028 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3030 PH/02 Reduce the amount of output that "make" produces by default. Full output
3031 can still be requested.
3033 PH/03 The warning log line about a condition test deferring for a "warn" verb
3034 was being output only once per connection, rather than after each
3035 occurrence (because it was using the same function as for successful
3036 "warn" verbs). This seems wrong, so I have changed it.
3038 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3039 it should not have, which might have caused a crash in the right
3040 circumstances, but probably never did.
3042 PH/04 Installed a modified version of Tony Finch's patch to make submission
3043 mode fix the return path as well as the Sender: header line, and to
3044 add a /name= option so that you can make the user's friendly name appear
3047 TF/03 Added the control = fakedefer ACL modifier.
3049 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3050 Mark Lowes for thorough testing.
3052 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3054 TK/03 Merged latest SRS patch from Miles Wilton.
3056 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3057 with the definition in sysexits.h (which is #included earlier).
3058 Fortunately, Exim does not actually use EX_OK. The code used to try to
3059 preserve the sysexits.h value, by assumimg that macro definitions were
3060 scanned for macro replacements. I have been disabused of this notion,
3061 so now the code just undefines EX_OK before #including unistd.h.
3063 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3064 in the smtp transport. When a block could not be written in a single
3065 write() function, the timeout was being re-applied to each part-write.
3066 This seems wrong - if the receiver was accepting one byte at a time it
3067 would take for ever. The timeout is now adjusted when this happens. It
3068 doesn't have to be particularly precise.
3070 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3071 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3073 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3074 without a domain if the check to <postmaster@domain> fails.
3076 SC/01 Eximstats: added -xls and the ability to specify output files
3077 (patch written by Frank Heydlauf).
3079 SC/02 Eximstats: use FileHandles for outputing results.
3081 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3083 SC/04 Eximstats: fixed display of large numbers with -nvr option
3085 SC/05 Eximstats: fixed merging of reports with empty tables.
3087 SC/06 Eximstats: added the -include_original_destination flag
3089 SC/07 Eximstats: removed tabs and trailing whitespace.
3091 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3093 TK/06 MBOX spool code: Add real "From " MBOX separator line
3094 so the .eml file is really in mbox format (even though
3095 most programs do not really care). Patch from Alex Miller.
3097 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3098 The latter is generated from $received_to and is only set if the
3099 message has one envelope recipient. SA can use these headers,
3100 obviously out-of-the-box. Patch from Alex Miller.
3102 PH/08 The ${def test on a variable was returning false if the variable's
3103 value was "0", contrary to what the specification has always said!
3104 The result should be true unless the variable is empty.
3106 PH/09 The syntax error of a character other than { following "${if
3107 def:variable_name" (after optional whitespace) was not being diagnosed.
3108 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3109 accidental colon was present, for example, could give incorrect results.
3111 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3112 result is used (not including appendfile, where other changes are about
3115 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3116 This involved changing a lot of size variables from int to off_t. It
3117 should work with maildirs and everything.
3119 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3120 spamd dying while we are connected to it.
3122 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3125 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3126 function for BDB 4.3.
3128 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3130 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3131 the sources for PCRE have been re-arranged and more files are now
3134 PH/15 The code I had for printing potentially long long variables in PH/11
3135 above was not the best (it lost precision). The length of off_t variables
3136 is now inspected at build time, and an appropriate printing format (%ld
3137 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3138 to be "long long int" or "long int". This is needed for the internal
3139 formatting function string_vformat().
3141 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3142 the configuration file to be ":syslog", then the script "guesses" where
3143 the logs files are, rather than using the compiled in default. In our
3144 case the guess is not the same as the compiled default, so the script
3145 suddenly stopped working when I started to use syslog. The patch checks
3146 to see if log_file_path is "". If so, it attempts to read it from exim
3147 with no configuration file to get the compiled in version, before it
3148 falls back to the previous guessing code."
3150 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3151 implementing BATV in an Exim configuration. See NewStuff for the gory
3154 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3155 Makefile that are specific to HP-UX.
3157 PH/18 If the "use_postmaster" option was set for a recipient callout together
3158 with the "random" option, the postmaster address was used as the MAIL
3159 FROM address for the random test, but not for the subsequent recipient
3160 test. It is now used for both.
3162 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3163 patch removes a few documentation additions to RFC 3028, because the
3164 latest draft now contains them. It adds the new en;ascii-case comparator
3165 and a new error check for 8bit text in MIME parts. Comparator and
3166 require names are now matched exactly. I enabled the subaddress
3167 extension, but it is not well tested yet (read: it works for me)."
3169 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3170 rework some of the code of TK/09 above to avoid the hardwired use of
3171 "%lld" and "long long". Replaced the call to snprintf() with a call to
3174 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3175 records point to non-existent hosts", "retry timeout exceeded", and
3176 "retry time not reached for any host after a long failure period".
3178 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3179 experimental DomainKeys support:
3181 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3182 (2) On an error such as an illegally used "control", the wrong name for
3183 the control was given.
3185 These problems did NOT occur unless DomainKeys support was compiled.
3187 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3189 PH/24 Added ${if match_ip condition.
3191 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3192 databases so that it will be absolutely obvious if a crash occurs in the
3193 DB library. This is a regular occurrence (often caused by mis-matched
3196 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3197 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3198 picked up on a user's system that detects such things. There doesn't seem
3199 to be a gcc warning option for this - only an attribute that has to be
3200 put on the function's prototype. It seems that in Fedora Core 4 they have
3201 set this on a number of new functions. No doubt there will be more in due
3204 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3205 specify a transport. However, if an address that was verified by such a
3206 router was the subject of a callout, Exim crashed because it tried to
3207 read the rcpt_include_affixes from the non-existent transport. Now it
3208 just assumes that the setting of that option is false. This bug was
3209 introduced by 4.51/PH/31.
3211 PH/28 Changed -d+all to exclude +memory, because that information is very
3212 rarely of interest, but it makes the output a lot bigger. People tend to
3213 do -d+all out of habit.
3215 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3216 code in os-type was giving problems when libc.so lives in lib64, like on
3219 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3220 aren't the modern standard, and it seems that some systems' include files
3221 don't always have them. Exim was already checking for some of the newer
3222 ones like T_AAAA, and defining it itself. I've added checks for all the
3223 record types that Exim uses.
3225 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3226 not automatically generating a new one, as it is supposed to. This
3227 prevented TLS from working. If the file did exist, but contained invalid
3228 data, a new version was generated, as expected. It was only the case of a
3229 non-existent file that was broken.
3231 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3232 with a change in libdomainkeys > 0.64.
3234 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3235 from DNS. If the selector record carries the flag, it now has
3236 precedence over the domain-wide flag.
3238 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3240 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3241 the use of a port name that isn't defined in /etc/services) occurred, the
3242 message was deferred as in a normal delivery, and thus remained on the
3243 spool, instead of being failed because of the mua_wrapper setting. This
3244 is now fixed, and I tidied up some of the mua_wrapper messages at the
3247 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3248 the messages in a hash of arrays rather than using individual hashes.
3249 This is a bit cleaner and results in dramatic memory savings, albeit
3250 at a slight CPU cost.
3252 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3253 as requested by Marc Sherman.
3255 SC/10 Eximstats: added histograms for user specified patterns as requested
3258 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3260 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3261 fopen() in the content-scanning modules that did not already have it.
3267 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3268 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3270 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3272 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3274 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3275 to test Sieve filters that use "vacation".
3277 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3278 that changes the way the GnuTLS parameters are stored in the cache file.
3279 The new format can be generated externally. For backward compatibility,
3280 if the data in the cache doesn't make sense, Exim assumes it has read an
3281 old-format file, and it generates new data and writes a new file. This
3282 means that you can't go back to an older release without removing the
3285 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3286 work if there are any delivery delays because "one_time" forces the
3287 parent to be marked "delivered", so its unseen clone is never tried
3288 again. For this reason, Exim now forbids the simultaneous setting of
3291 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3292 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3293 entry if you want to know the details. The fix, however, neglected to
3294 consider the case when local delivery batching is involved. The test for
3295 "previously delivered" was not happening when checking to see if an
3296 address could be batched with a previous (undelivered) one; under
3297 certain circumstances this could lead to multiple deliveries to the same
3300 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3301 in its include files, and this causes problems building Exim.
3303 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3304 header_syntax) but Exim was just ignoring anything given after a slash.
3305 In particular, this caused confusion with an attempt to use "verify =
3306 reverse_host_lookup/defer_ok". An error is now given when options are
3307 supplied for verify items that do not have them. (Maybe reverse_host_
3308 lookup should have a defer_ok option, but that's a different point.)
3310 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3311 defined by RFC 821) to 2048, because there were problems with some AUTH
3312 commands, and RFC 1869 says the size should be increased for extended
3313 SMTP commands that take arguments.
3315 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3318 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3319 "unknown" error; now it says that the functionality isn't in the binary.
3321 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3322 an address' error message when a string expansion fails (syntax or
3323 whatever). Otherwise the password may appear in the log. Following change
3324 PH/42 below, there is no longer a chance of it appearing in a bounce
3327 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3329 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3330 of its addresses was ever tried. (Bugzilla bug #2.)
3332 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3333 the result incorrectly in the debug output. (It correctly added a newline
3334 to what was transported.)
3336 TF/01 Added $received_time.
3338 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3339 commented out examples of how to interface to a virus scanner and to
3340 SpamAssassin. Also added commented examples of av_scanner and
3341 spamd_address settings.
3343 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3344 and controls are allowed in which ACLs. There were a couple of minor
3345 errors. Some of the entries in the conditions table (which is a table of
3346 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3347 negation of where the condition IS allowed.
3349 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3351 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3352 header file does not have a version number, so I've had to invent a new
3353 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3354 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3355 radiusclient), but it was contributed by a Radius user.
3357 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3358 files or directories, for OpenSSL.
3360 PH/19 When an Exim process that is running as root has to create an Exim log
3361 file, it does so in a subprocess that runs as exim:exim so as to get the
3362 ownership right at creation (otherwise, other Exim processes might see
3363 the file with the wrong ownership). There was no test for failure of this
3364 fork() call, which would lead to the process getting stuck as it waited
3365 for a non-existent subprocess. Forks do occasionally fail when resources
3366 run out. I reviewed all the other calls to fork(); they all seem to check
3369 PH/20 When checking for unexpected SMTP input at connect time (before writing
3370 the banner), Exim was not dealing correctly with a non-positive return
3371 from the read() function. If the client had disconnected by this time,
3372 the result was a log entry for a synchronization error with an empty
3373 string after "input=" when read() returned zero. If read() returned -1
3374 (an event I could not check), uninitialized data bytes were printed.
3375 There were reports of junk text (parts of files, etc) appearing after
3378 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3380 PH/22 Added support for macro redefinition, and (re)definition in between
3381 driver and ACL definitions.
3383 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3384 forgetting to use the resulting value; it was using the unexpanded value.
3386 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3387 hadn't been configured. The fix is from Juergen Kreileder, who
3388 understands it better than I do:
3390 "Here's what I see happening with three configured cyrus_sasl
3391 authenticators configured (plain, login, cram-md5):
3393 On startup auth_cyrus_sasl_init() gets called for each of these.
3394 This means three calls to sasl_listmech() without a specified mech_list.
3395 => SASL tests which mechs of all available mechs actually work
3396 => three warnings about OTP not working
3397 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3399 With the patch, sasl_listmech() also gets called three times. But now
3400 SASL's mech_list option is set to the server_mech specified in the the
3401 authenticator. Or in other words, the answer from sasl_listmech()
3402 gets limited to just the mech you're testing for (which is different
3404 => the return list contains just 'plain' or 'login', 'cram-md5' or
3405 nothing depending on the value of ob->server_mech.
3407 I've just tested the patch: Authentication still works fine,
3408 unavailable mechs specified in the exim configuration are still
3409 caught, and the auth.log warnings about OTP are gone."
3411 PH/25 When debugging is enabled, the contents of the command line are added
3412 to the debugging output, even when log_selector=+arguments is not
3415 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3416 answer is "GNU", and only if the return is "GNU/something" is the answer
3419 PH/27 $acl_verify_message is now set immediately after the failure of a
3420 verification in an ACL, and so is available in subsequent modifiers. In
3421 particular, the message can be preserved by coding like this:
3423 warn !verify = sender
3424 set acl_m0 = $acl_verify_message
3426 Previously, $acl_verify_message was set only while expanding "message"
3427 and "log_message" when a very denied access.
3429 PH/28 Modified OS/os.c-Linux with
3431 -#ifndef OS_LOAD_AVERAGE
3432 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3434 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3435 nomenclature these days.)
3437 PH/29 Installed patch from the Sieve maintainer that adds the options
3438 sieve_useraddress and sieve_subaddress to the redirect router.
3440 PH/30 In these circumstances:
3441 . Two addresses routed to the same list of hosts;
3442 . First host does not offer TLS;
3443 . First host accepts first address;
3444 . First host gives temporary error to second address;
3445 . Second host offers TLS and a TLS session is established;
3446 . Second host accepts second address.
3447 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3448 and peerdn, if requested) that were in fact used only for the second
3451 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3452 attention to any local part prefix or suffix that was matched by the
3453 router that accepted the address. It now behaves in the same way as it
3454 does for delivery: the affixes are removed from the local part unless
3455 rcpt_include_affixes is set on the transport.
3457 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3458 timeout during the DATA phase of an incoming message.
3460 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3461 applied a patch sanctioned by the Sieve maintainer.
3463 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3464 the uid or gid is negative. A case of a negative gid caused this to be
3465 noticed. The fix allows for either to be negative.
3467 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3468 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3471 PH/36 The change PH/12 above was broken. Fixed it.
3473 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3474 the grounds that routing the same address twice would always produce the
3475 same answer. This might have been true once, but it is certainly no
3476 longer true now. Routing a child address may depend on the previous
3477 routing that produced that child. Some complicated redirection strategies
3478 went wrong when messages had multiple recipients, and made Exim's
3479 behaviour dependent on the order in which the addresses were given.
3481 I have moved the duplicate checking until after the routing is complete.
3482 Exim scans the addresses that are assigned to local and remote
3483 transports, and removes any duplicates. This means that more work will be
3484 done, as duplicates will always all be routed, but duplicates are
3485 presumably rare, so I don't expect this is of any significance.
3487 For deliveries to pipes, files, and autoreplies, the duplicate checking
3488 still happens during the routing process, since they are not going to be
3491 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3492 It corrects a timeout issue with spamd. This is Ian's comment: "The
3493 background is that sometimes spamd either never reads data from a
3494 connection it has accepted, or it never writes response data. The exiscan
3495 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3496 blindly assumes that writes won't block so it may never time out."
3498 PH/39 Allow G after quota size as well as K and M.
3500 PH/40 The value set for $authenticated_id in an authenticator may not contain
3501 binary zeroes or newlines because the value is written to log lines and
3502 to spool files. There was no check on this. Now the value is run through
3503 the string_printing() function so that such characters are converted to
3504 printable escape sequences.
3506 PH/41 $message_linecount is a new variable that contains the total number of
3507 lines in the message. Compare $body_linecount, which is the count for the
3510 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3511 bounce and delay warning messages, except in certain special cases, which
3514 (a) An SMTP error message from a remote host;
3515 (b) A message specified in a :fail: redirection;
3516 (c) A message specified in a "fail" command in a system filter;
3517 (d) A message specified in a FAIL return from the queryprogram router;
3518 (e) A message specified by the cannot_route_message router option.
3520 In these cases only, Exim does include the error details in bounce and
3521 warning messages. There are also a few cases where bland messages such
3522 as "unrouteable address" or "local delivery error" are given.
3524 PH/43 $value is now also set for the "else" part of a ${run expansion.
3526 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3527 being worked on, but at least Exim now implements the latest version to
3530 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3531 process to complete was treated as a delivery failure, a timeout while
3532 writing the message to the pipe was logged, but erroneously treated as a
3533 successful delivery. Such timeouts include transport filter timeouts. For
3534 consistency with the overall process timeout, these timeouts are now
3535 treated as errors, giving rise to delivery failures by default. However,
3536 there is now a new Boolean option for the pipe transport called
3537 timeout_defer, which, if set TRUE, converts the failures into defers for
3538 both kinds of timeout. A transport filter timeout is now identified in
3541 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3542 systems where "make" and "gmake" are different, calling "gmake" at top
3543 level broke things. I've arranged for the value of $(MAKE) to be passed
3544 from the Makefile to this script so that it can call the same version of
3548 A note about Exim versions 4.44 and 4.50
3549 ----------------------------------------
3551 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3552 changes of various kinds. As a consequence, a big documentation update was
3553 needed. This delayed the release for rather longer than seemed good, especially
3554 in the light of a couple of (minor) security issues. Therefore, the changes
3555 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3556 release. So 4.44 and 4.50 are in effect two different branches that both start
3559 I have left the 4.50 change log unchanged; it contains all the changes since
3560 4.43. The change log for 4.44 is below; many of its items are identical to
3561 those for 4.50. This seems to be the most sensible way to preserve the
3562 historical information.
3568 1. Minor wording change to the doc/README.SIEVE file.
3570 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3571 computation of the current number of files was incorrect.
3573 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3574 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3577 4. Give more explanation in the error message when the command for a transport
3578 filter fails to execute.
3580 5. There are several places where Exim runs a non-Exim command in a
3581 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3582 was being done only for the command run by the queryprogram router. It is
3583 now done for all such subprocesses. The other cases are: ${run, transport
3584 filters, and the commands run by the lmtp and pipe transports.
3586 6. Added CONFIGURE_GROUP build-time option.
3588 7. Some older OS have a limit of 256 on the maximum number of file
3589 descriptors. Exim was using setrlimit() to set 1000 as a large value
3590 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3591 systems. I've change it so that if it can't get 1000, it tries for 256.
3593 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3594 was an oversight, and furthermore, ever since the addition of extra
3595 controls (e.g. 4.43/32), the checks on when to allow different forms of
3596 "control" were broken. There should now be diagnostics for all cases when a
3597 control that does not make sense is encountered.
3599 9. Added the /retain_sender option to "control=submission".
3601 10. $recipients is now available in the predata ACL (oversight).
3603 11. Tidy the search cache before the fork to do a delivery from a message
3604 received from the command line. Otherwise the child will trigger a lookup
3605 failure and thereby defer the delivery if it tries to use (for example) a
3606 cached ldap connection that the parent has called unbind on.
3608 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3609 of $address_data from the recipient verification was clobbered by the
3610 sender verification.
3612 13. The value of address_data from a sender verification is now available in
3613 $sender_address_data in subsequent conditions in the ACL statement.
3615 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3617 15. Added a new option "connect=<time>" to callout options, to set a different
3620 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3621 was its contents. (It was OK if the option was not defined at all.)
3623 17. A "Completed" log line is now written for messages that are removed from
3624 the spool by the -Mrm option.
3626 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3627 information about exactly what failed.
3629 19. Added -dd to debug only the daemon process.
3631 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3632 handles timeouts, both on the server side and network timeouts. Renamed the
3633 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3635 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3636 It is now set to "smtps".
3638 22. $host_address is now set to the target address during the checking of
3639 ignore_target_hosts.
3641 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3642 being passed; this would have caused $sender_host_name to have been used if
3643 matching the list had actually called for a host name (not very likely,
3644 since this list is usually IP addresses). A host name is now passed as
3647 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3648 code that specifies a non-restarting handler (typically sigaction() in
3649 modern systems) in an attempt to fix a rare and obscure crash bug.
3651 25. Narrowed the window for a race in the daemon that could cause it to ignore
3652 SIGCHLD signals. This is not a major problem, because they are used only to
3653 wake it up if nothing else does.
3655 26. A malformed maildirsize file could cause Exim to calculate negative values
3656 for the mailbox size or file count. Odd effects could occur as a result.
3657 The maildirsize information is now recalculated if the size or filecount
3660 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3661 support for a long time. Removed HAVE_SYS_VFS_H.
3663 28. Installed the latest version of exipick from John Jetmore.
3665 29. In an address list, if the pattern was not a regular expression, an empty
3666 subject address (from a bounce message) matched only if the pattern was an
3667 empty string. Non-empty patterns were not even tested. This was the wrong
3668 because it is perfectly reasonable to use an empty address as part of a
3669 database query. An empty address is now tested by patterns that are
3670 lookups. However, all the other forms of pattern expect the subject to
3671 contain a local part and a domain, and therefore, for them, an empty
3672 address still always fails if the pattern is not itself empty.
3674 30. Exim went into a mad DNS loop when attempting to do a callout where the
3675 host was specified on an smtp transport, and looking it up yielded more
3676 than one IP address.
3678 31. Re-factored the code for checking spool and log partition space into a
3679 function that finds that data and another that does the check. The former
3680 is then used to implement four new variables: $spool_space, $log_space,
3681 $spool_inodes, and $log_inodes.
3683 32. The RFC2047 encoding function was originally intended for short strings
3684 such as real names; it was not keeping to the 75-character limit for
3685 encoded words that the RFC imposes. It now respects the limit, and
3686 generates multiple encoded words if necessary. To be on the safe side, I
3687 have increased the buffer size for the ${rfc2047: expansion operator from
3690 33. It is now permitted to omit both strings after an "if" condition; if the
3691 condition is true, the result is "true". As before, when the second string
3692 is omitted, a false condition yields an empty string. This makes it less
3693 cumbersome to write custom ACL and router conditions.
3695 34. Failure to deliver a bounce message always caused it to be frozen, even if
3696 there was an errors_to setting on the router. The errors_to setting is now
3699 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3700 canonical form (fully expanded) before being placed in
3701 $sender_host_address.
3703 36. The table in the code that translates DNS record types into text (T_A to
3704 "A" for instance) was missing entries for NS and CNAME. It is just possible
3705 that this could have caused confusion if both these types were looked up
3706 for the same domain, because the text type is used as part of Exim's
3707 per-process caching. But the chance of anyone hitting this buglet seems
3710 37. The dnsdb lookup has been extended in a number of ways.
3712 (1) There is a new type, "zns", which walks up the domain tree until it
3713 finds some nameserver records. It should be used with care.
3715 (2) There is a new type, "mxh", which is like "mx" except that it returns
3716 just the host names, not the priorities.
3718 (3) It is now possible to give a list of domains (or IP addresses) to be
3719 looked up. The behaviour when one of the lookups defers can be
3720 controlled by a keyword.
3722 (4) It is now possible to specify the separator character for use when
3723 multiple records are returned.
3725 38. The dnslists ACL condition has been extended: it is now possible to supply
3726 a list of IP addresses and/or domains to be looked up in a particular DNS
3729 39. Added log_selector=+queue_time_overall.
3731 40. When running the queue in the test harness, wait just a tad after forking a
3732 delivery process, to get repeatability of debugging output.
3734 41. Include certificate and key file names in error message when GnuTLS fails
3735 to set them up, because the GnuTLS error message doesn't include the name
3736 of the failing file when there is a problem reading it.
3738 42. Allow both -bf and -bF in the same test run.
3740 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3742 44. The "Exiscan patch" is now merged into the mainline Exim source.
3744 45. Sometimes the final signoff response after QUIT could fail to get
3745 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3746 < 0 before doing a fflush(). This bug looks as though it goes back to the
3747 introduction of TLS in release 3.20, but "sometimes" must have been rare
3748 because the tests only now provoked it.
3750 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3751 (this can affect the format of dates).
3753 47. exim_tidydb, when checking for the continued existence of a message for
3754 which it has found a message-specific retry record, was not finding
3755 messages that were in split spool directories. Consequently, it was
3756 deleting retry records that should have stayed in existence.
3758 48. Steve fixed some bugs in eximstats.
3760 49. The SPA authentication driver was not abandoning authentication and moving
3761 on to the next authenticator when an expansion was forced to fail,
3762 contradicting the general specification for all authenticators. Instead it
3763 was generating a temporary error. It now behaves as specified.
3765 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3766 (the order specifies the preference for clients). The order is now AES256,
3767 AES128, 3DES, ARCFOUR128.
3769 51. Small patch to Sieve code - explicitly set From: when generating an
3772 52. Exim crashed if a remote delivery caused a very long error message to be
3773 recorded - for instance if somebody sent an entire SpamAssassin report back
3774 as a large number of 550 error lines. This bug was coincidentally fixed by
3775 increasing the size of one of Exim's internal buffers (big_buffer) that
3776 happened as part of the Exiscan merge. However, to be on the safe side, I
3777 have made the code more robust (and fixed the comments that describe what
3780 53. Now that there can be additional text after "Completed" in log lines (if
3781 the queue_time_overall log selector is set), a one-byte patch to exigrep
3782 was needed to allow it to recognize "Completed" as not the last thing in
3785 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3786 patch that reportedly fixes this has been added. I am not expert enough to
3787 create a test for it. This is what the patch creator wrote:
3789 "I found a little strange behaviour of ldap code when working with
3790 Windows 2003 AD Domain, where users was placed in more than one
3791 Organization Units. When I tried to give exim partial DN, the exit code
3792 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3793 But simultaneously result of request was absolutely normal ldap result,
3794 so I produce this patch..."
3796 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3797 REFERENCE, so I have modified the code to exclude the patch when that macro
3800 55. Some experimental protocols are using DNS PTR records for new purposes. The
3801 keys for these records are domain names, not reversed IP addresses. The
3802 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3803 leaves it alone. Component reversal etc. now happens only for IP addresses.
3806 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3808 57. Double the size of the debug message buffer (to 2048) so that more of very
3809 long debug lines gets shown.
3811 58. The exicyclog utility now does better if the number of log files to keep
3812 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3814 59. Two changes related to the smtp_active_hostname option:
3816 (1) $smtp_active_hostname is now available as a variable.
3817 (2) The default for smtp_banner uses $smtp_active_hostname instead
3818 of $primary_hostname.
3820 60. The host_aton() function is supposed to be passed a string that is known
3821 to be a valid IP address. However, in the case of IPv6 addresses, it was
3822 not checking this. This is a hostage to fortune. Exim now panics and dies
3823 if the condition is not met. A case was found where this could be provoked
3824 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3825 components; fortuitously, this particular loophole had already been fixed
3826 by change 4.50/55 above.
3828 If there are any other similar loopholes, the new check in host_aton()
3829 itself should stop them being exploited. The report I received stated that
3830 data on the command line could provoke the exploit when Exim was running as
3831 exim, but did not say which command line option was involved. All I could
3832 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3833 running as the user.
3836 61. There was a buffer overflow vulnerability in the SPA authentication code
3837 (which came originally from the Samba project). I have added a test to the
3838 spa_base64_to_bits() function which I hope fixes it.
3841 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3842 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3844 63. The daemon start-up calls getloadavg() while still root for those OS that
3845 need the first call to be done as root, but it missed one case: when
3846 deliver_queue_load_max is set with deliver_drop_privilege. This is
3847 necessary for the benefit of the queue runner, because there is no re-exec
3848 when deliver_drop_privilege is set.
3850 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3851 This has been fixed.
3853 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3854 was in use, was not putting the data itself into the right store pool;
3855 consequently, it could be overwritten for a subsequent message in the same
3856 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3859 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3861 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3862 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3863 the function were treating the return as a boolean value, which happened to
3864 work because 0=false and not-0=true, but is not correct code.
3866 68. The host_aton() function was not handling scoped IPv6 addresses (those
3867 with, for example, "%eth0" on the end) correctly.
3869 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3870 negated items (that is, ~something) in unsigned ints. Some compilers
3871 apparently mutter when there is no cast.
3873 70. If an address verification called from an ACL failed, and did not produce a
3874 user-specific message (i.e. there was only a "system" message), nothing was
3875 put in $acl_verify_message. In this situation, it now puts the system
3878 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3879 session; change 4.31/43 added the unwanted input to the log line - except
3880 that it did not do this in the start of session case. It now does.
3882 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3883 This is wrong and can cause the other end to generate a synchronization
3884 error if it is another Exim or anything else that does the synchronization
3885 check. A QUIT command is no longer sent after a timeout.
3887 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3888 during host lookups.
3890 74. The defer_ok option of callout verification was not working if it was used
3891 when verifying addresses in header lines, that is, for this case:
3893 verify = header_sender/callout=defer_ok
3895 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3896 those file descriptors could be used for SMTP connections. If anything
3897 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3898 could be sent to the SMTP client, causing chaos. The daemon now opens
3899 stdin, stdout, and stderr to /dev/null when it puts itself into the
3902 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3903 log by default. The user can override this with suitable Perl magic.
3905 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3906 the log message when discard triggers, was not working for the DATA ACL or
3907 for the non-SMTP ACL.
3909 78. Error message wording change in sieve.c.
3911 79. If smtp_accept_max_per_host was set, the number of connections could be
3912 restricted to fewer than expected, because the daemon was trying to set up
3913 a new connection before checking whether the processes handling previous
3914 connections had finished. The check for completed processes is now done
3915 earlier. On busy systems, this bug wouldn't be noticed because something
3916 else would have woken the daemon, and it would have reaped the completed
3919 80. If a message was submitted locally by a user whose login name contained one
3920 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3921 It caused a spool format error. I have fixed the spool reading code. A
3922 related problem was that the "from" clause in the Received: line became
3923 illegal because of the space(s). It is now covered by ${quote_local_part.
3925 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3928 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3930 83. Patch from Sieve maintainer to fix unterminated string problem in
3931 "vacation" handling.
3933 84. Some minor changes to the Linux configuration files to help with other
3934 OS variants using glibc.
3936 85. One more patch for Sieve to update vacation handling to latest spec.
3939 ----------------------------------------------------
3940 See the note above about the 4.44 and 4.50 releases.
3941 ----------------------------------------------------
3947 1. Change 4.43/35 introduced a bug that caused file counts to be
3948 incorrectly computed when quota_filecount was set in an appendfile
3951 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3952 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3955 3. Give more explanation in the error message when the command for a transport
3956 filter fails to execute.
3958 4. There are several places where Exim runs a non-Exim command in a
3959 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3960 was being done only for the command run by the queryprogram router. It is
3961 now done for all such subprocesses. The other cases are: ${run, transport
3962 filters, and the commands run by the lmtp and pipe transports.
3964 5. Some older OS have a limit of 256 on the maximum number of file
3965 descriptors. Exim was using setrlimit() to set 1000 as a large value
3966 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3967 systems. I've change it so that if it can't get 1000, it tries for 256.
3969 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3970 was an oversight, and furthermore, ever since the addition of extra
3971 controls (e.g. 4.43/32), the checks on when to allow different forms of
3972 "control" were broken. There should now be diagnostics for all cases when a
3973 control that does not make sense is encountered.
3975 7. $recipients is now available in the predata ACL (oversight).
3977 8. Tidy the search cache before the fork to do a delivery from a message
3978 received from the command line. Otherwise the child will trigger a lookup
3979 failure and thereby defer the delivery if it tries to use (for example) a
3980 cached ldap connection that the parent has called unbind on.
3982 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3983 of $address_data from the recipient verification was clobbered by the
3984 sender verification.
3986 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3987 was its contents. (It was OK if the option was not defined at all.)
3989 11. A "Completed" log line is now written for messages that are removed from
3990 the spool by the -Mrm option.
3992 12. $host_address is now set to the target address during the checking of
3993 ignore_target_hosts.
3995 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3996 being passed; this would have caused $sender_host_name to have been used if
3997 matching the list had actually called for a host name (not very likely,
3998 since this list is usually IP addresses). A host name is now passed as
4001 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4002 code that specifies a non-restarting handler (typically sigaction() in
4003 modern systems) in an attempt to fix a rare and obscure crash bug.
4005 15. Narrowed the window for a race in the daemon that could cause it to ignore
4006 SIGCHLD signals. This is not a major problem, because they are used only to
4007 wake it up if nothing else does.
4009 16. A malformed maildirsize file could cause Exim to calculate negative values
4010 for the mailbox size or file count. Odd effects could occur as a result.
4011 The maildirsize information is now recalculated if the size or filecount
4014 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4015 support for a long time. Removed HAVE_SYS_VFS_H.
4017 18. Updated exipick to current release from John Jetmore.
4019 19. Allow an empty sender to be matched against a lookup in an address list.
4020 Previously the only cases considered were a regular expression, or an
4023 20. Exim went into a mad DNS lookup loop when doing a callout where the
4024 host was specified on the transport, if the DNS lookup yielded more than
4027 21. The RFC2047 encoding function was originally intended for short strings
4028 such as real names; it was not keeping to the 75-character limit for
4029 encoded words that the RFC imposes. It now respects the limit, and
4030 generates multiple encoded words if necessary. To be on the safe side, I
4031 have increased the buffer size for the ${rfc2047: expansion operator from
4034 22. Failure to deliver a bounce message always caused it to be frozen, even if
4035 there was an errors_to setting on the router. The errors_to setting is now
4038 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4039 canonical form (fully expanded) before being placed in
4040 $sender_host_address.
4042 24. Updated eximstats to version 1.33
4044 25. Include certificate and key file names in error message when GnuTLS fails
4045 to set them up, because the GnuTLS error message doesn't include the name
4046 of the failing file when there is a problem reading it.
4048 26. Expand error message when OpenSSL has problems setting up cert/key files.
4051 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4052 (this can affect the format of dates).
4054 28. exim_tidydb, when checking for the continued existence of a message for
4055 which it has found a message-specific retry record, was not finding
4056 messages that were in split spool directories. Consequently, it was
4057 deleting retry records that should have stayed in existence.
4059 29. eximstats updated to version 1.35
4060 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4061 1.35 - bugfix such that pie charts by volume are generated correctly
4063 30. The SPA authentication driver was not abandoning authentication and moving
4064 on to the next authenticator when an expansion was forced to fail,
4065 contradicting the general specification for all authenticators. Instead it
4066 was generating a temporary error. It now behaves as specified.
4068 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4069 (the order specifies the preference for clients). The order is now AES256,
4070 AES128, 3DES, ARCFOUR128.
4072 31. Small patch to Sieve code - explicitly set From: when generating an
4075 32. Exim crashed if a remote delivery caused a very long error message to be
4076 recorded - for instance if somebody sent an entire SpamAssassin report back
4077 as a large number of 550 error lines. This bug was coincidentally fixed by
4078 increasing the size of one of Exim's internal buffers (big_buffer) that
4079 happened as part of the Exiscan merge. However, to be on the safe side, I
4080 have made the code more robust (and fixed the comments that describe what
4083 33. Some experimental protocols are using DNS PTR records for new purposes. The
4084 keys for these records are domain names, not reversed IP addresses. The
4085 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4086 leaves it alone. Component reversal etc. now happens only for IP addresses.
4089 34. The host_aton() function is supposed to be passed a string that is known
4090 to be a valid IP address. However, in the case of IPv6 addresses, it was
4091 not checking this. This is a hostage to fortune. Exim now panics and dies
4092 if the condition is not met. A case was found where this could be provoked
4093 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4094 components; fortuitously, this particular loophole had already been fixed
4095 by change 4.50/55 or 4.44/33 above.
4097 If there are any other similar loopholes, the new check in host_aton()
4098 itself should stop them being exploited. The report I received stated that
4099 data on the command line could provoke the exploit when Exim was running as
4100 exim, but did not say which command line option was involved. All I could
4101 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4102 running as the user.
4105 35. There was a buffer overflow vulnerability in the SPA authentication code
4106 (which came originally from the Samba project). I have added a test to the
4107 spa_base64_to_bits() function which I hope fixes it.
4110 36. The daemon start-up calls getloadavg() while still root for those OS that
4111 need the first call to be done as root, but it missed one case: when
4112 deliver_queue_load_max is set with deliver_drop_privilege. This is
4113 necessary for the benefit of the queue runner, because there is no re-exec
4114 when deliver_drop_privilege is set.
4116 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4117 was in use, was not putting the data itself into the right store pool;
4118 consequently, it could be overwritten for a subsequent message in the same
4119 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4122 38. Sometimes the final signoff response after QUIT could fail to get
4123 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4124 < 0 before doing a fflush(). This bug looks as though it goes back to the
4125 introduction of TLS in release 3.20, but "sometimes" must have been rare
4126 because the tests only now provoked it.
4132 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4133 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4134 Now it returns an integer. A number of calls were still expecting a T/F
4135 return. Fortuitously, in all cases, the tests worked in OK situations,
4136 which is the norm. However, things would have gone wrong on any write
4137 failures on the smtp file descriptor. This function is used when sending
4138 messages over SMTP and also when doing verify callouts.
4140 2. When Exim is called to do synchronous delivery of a locally submitted
4141 message (the -odf or -odi options), it no longer closes stderr before doing
4144 3. Implemented the mua_wrapper option.
4146 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4148 5. Implemented the functions header_remove(), header_testname(),
4149 header_add_at_position(), and receive_remove_recipient(), and exported them
4152 6. If an ACL "warn" statement specified the addition of headers, Exim already
4153 inserted X-ACL-Warn: at the start if there was no header name. However, it
4154 was not making this test for the second and subsequent header lines if
4155 there were newlines in the string. This meant that an invalid header could
4156 be inserted if Exim was badly configured.
4158 7. Allow an ACL "warn" statement to add header lines at the start or after all
4159 the Received: headers, as well as at the end.
4161 8. Added the rcpt_4xx retry error code.
4163 9. Added postmaster_mailfrom=xxx to callout verification option.
4165 10. Added mailfrom=xxxx to the callout verification option, for verify=
4168 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4169 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4171 12. Inserted some casts to stop certain compilers warning when using pointer
4172 differences as field lengths or precisions in printf-type calls (mostly
4173 affecting debugging statements).
4175 13. Added optional readline() support for -be (dynamically loaded).
4177 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4178 same clock tick as a message's arrival, so that its received time was the
4179 same as the "first fail" time on the retry record, and that message
4180 remained on the queue past the ultimate address timeout, every queue runner
4181 would try a delivery (because it was past the ultimate address timeout) but
4182 after another failure, the ultimate address timeout, which should have then
4183 bounced the address, did not kick in. This was a "< instead of <=" error;
4184 in most cases the first failure would have been in the next clock tick
4185 after the received time, and all would be well.
4187 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4188 being recognized when the domain list was tested by the match_domain
4189 condition in an expansion string.
4191 16. Added the ${str2b64: operator.
4193 17. Exim was always calling setrlimit() to set a large limit for the number of
4194 processes, without checking whether the existing limit was already
4195 adequate. (It did check for the limit on file descriptors.) Furthermore,
4196 errors from getrlimit() and setrlimit() were being ignored. Now they are
4197 logged to the main and panic logs, but Exim does carry on, to try to do its
4198 job under whatever limits there are.
4200 18. Imported PCRE 5.0.
4202 19. Trivial typo in log message " temporarily refused connection" (the leading
4205 20. If the log selector return_path_on_delivery was set and an address was
4206 redirected to /dev/null, the delivery process crashed because it assumed
4207 that a return path would always be set for a "successful" delivery. In this
4208 case, the whole delivery is bypassed as an optimization, and therefore no
4211 21. Internal re-arrangement: the function for sending a challenge and reading
4212 a response while authentication was assuming a zero-terminated challenge
4213 string. It's now changed to take a pointer and a length, to allow for
4214 binary data in such strings.
4216 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4218 23. Exim was not respecting finduser_retries when seeking the login of the
4219 uid under which it was called; it was always trying 10 times. (The default
4220 setting of finduser_retries is zero.) Also, it was sleeping after the final
4221 failure, which is pointless.
4223 24. Implemented tls_on_connect_ports.
4225 25. Implemented acl_smtp_predata.
4227 26. If the domain in control=submission is set empty, Exim assumes that the
4228 authenticated id is a complete email address when it generates From: or
4229 Sender: header lines.
4231 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4232 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4233 chown and chgrp in /bin and hostname in /usr/bin.
4235 28. Exim was keeping the "process log" file open after each use, just as it
4236 does for the main log. This opens the possibility of it remaining open for
4237 long periods when the USR1 signal hits a daemon. Occasional processlog
4238 errors were reported, that could have been caused by this. Anyway, it seems
4239 much more sensible not to leave this file open at all, so that is what now
4242 29. The long-running daemon process does not normally write to the log once it
4243 has entered its main loop, and it closes the log before doing so. This is
4244 so that log files can straightforwardly be renamed and moved. However,
4245 there are a couple of unusual error situations where the daemon does write
4246 log entries, and I had neglected to close the log afterwards.
4248 30. The text of an SMTP error response that was received during a remote
4249 delivery was being truncated at 512 bytes. This is too short for some of
4250 the long messages that one sometimes sees. I've increased the limit to
4253 31. It is now possible to make retry rules that apply only when a message has a
4254 specific sender, in particular, an empty sender.
4256 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4257 it possible to be selective about when SMTP synchronization is enforced.
4259 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4261 32. Implemented hosts_connection_nolog.
4263 33. Added an ACL for QUIT.
4265 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4268 35. Added mailbox_size and mailbox_filecount to appendfile.
4270 36. Added control = no_multiline_responses to ACLs.
4272 37. There was a bug in the logic of the code that waits for the clock to tick
4273 in the case where the clock went backwards by a substantial amount such
4274 that the microsecond fraction of "now" was more than the microsecond
4275 fraction of "then" (but the whole seconds number was less).
4277 38. Added support for the libradius Radius client library this is found on
4278 FreeBSD (previously only the radiusclient library was supported).
4284 1. When certain lookups returned multiple values in the form name=value, the
4285 quoting of the values was not always being done properly. Specifically:
4286 (a) If the value started with a double quote, but contained no whitespace,
4288 (b) If the value contained whitespace other than a space character (i.e.
4289 tabs or newlines or carriage returns) it was not quoted.
4290 This fix has been applied to the mysql and pgsql lookups by writing a
4291 separate quoting function and calling it from the lookup code. The fix
4292 should probably also be applied to nisplus, ibase and oracle lookups, but
4293 since I cannot test any of those, I have not disturbed their existing code.
4295 2. A hit in the callout cache for a specific address caused a log line with no
4296 reason for rejecting RCPT. Now it says "Previous (cached) callout
4297 verification failure".
4299 3. There was an off-by-one bug in the queryprogram router. An over-long
4300 return line was truncated at 256 instead of 255 characters, thereby
4301 overflowing its buffer with the terminating zero. As well as fixing this, I
4302 have increased the buffer size to 1024 (and made a note to document this).
4304 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4305 when Exim is waiting for an SMTP response from a remote server, Exim
4306 restarts its select() call on the socket, thereby resetting its timeout.
4307 This is not a problem when such interrupts are rare. Somebody set up a cron
4308 job to run exiwhat every 2 minutes, which is less than the normal select()
4309 timeout (5 or 10 minutes). This meant that the select() timeout never
4310 kicked in because it was always reset. I have fixed this by comparing the
4311 time when an interrupt arrives with the time at the start of the first call
4312 to select(). If more time than the timeout has elapsed, the interrupt is
4313 treated as a timeout.
4315 5. Some internal re-factoring in preparation for the addition of Sieve
4316 extensions (by MH). In particular, the "personal" test is moved to a
4317 separate function, and given an option for scanning Cc: and Bcc: (which is
4318 not set for Exim filters).
4320 6. When Exim created an email address using the login of the caller as the
4321 local part (e.g. when creating a From: or Sender: header line), it was not
4322 quoting the local part when it contained special characters such as @.
4324 7. Installed new OpenBSD configuration files.
4326 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4327 try to make them clearer.
4329 9. Callout options, other than the timeout value, were being ignored when
4330 verifying sender addresses in header lines. For example, when using
4332 verify = header_sender/callout=no_cache
4334 the cache was (incorrectly) being used.
4336 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4337 only the Cygwin environment.
4339 11. When return_path_on_delivery was set as a log selector, if different remote
4340 addresses in the same message used different return paths and parallel
4341 remote delivery occurred, the wrong values would sometimes be logged.
4342 (Whenever a remote delivery process finished, the return path value from
4343 the most recently started remote delivery process was logged.)
4345 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4346 lines when AUTH and/or TLS are in use. This is the "received protocol"
4347 field. Exim used to use "asmtp" for authenticated SMTP, without any
4348 indication (in the protocol name) for TLS use. Now it follows the RFC and
4349 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4350 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4351 names appear in log lines as well as in Received: header lines.
4353 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4354 extensions, and comparison tests, and to fix some bugs.
4356 14. Changes to the "personal" filter test:
4358 (1) The test was buggy in that it was just doing the equivalent of
4359 "contains" tests on header lines. For example, if a user's address was
4360 anne@some.where, the "personal" test would incorrectly be true for
4362 To: susanne@some.where
4364 This test is now done by extracting each address from the header in turn,
4365 and checking the entire address. Other tests that are part of "personal"
4366 are now done using regular expressions (for example, to check local parts
4367 of addresses in From: header lines).
4369 (2) The list of non-personal local parts in From: addresses has been
4370 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4371 taken from the Sieve specification recommendations.
4373 (3) If the message contains any header line starting with "List-" it is
4374 treated as non-personal.
4376 (4) The test for "circular" in the Subject: header line has been removed
4377 because it now seems ill-conceived.
4379 15. Minor typos in src/EDITME comments corrected.
4381 16. Installed latest exipick from John Jetmore.
4383 17. If headers_add on a router specified a text string that was too long for
4384 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4385 of string_sprintf() is now avoided.
4387 18. $message_body_size was not set (it was always zero) when running the DATA
4388 ACL and the local_scan() function.
4390 19. For the "mail" command in an Exim filter, no default was being set for
4391 the once_repeat time, causing a random time value to be used if "once" was
4392 specified. (If the value happened to be <= 0, no repeat happened.) The
4393 default is now 0s, meaning "never repeat". The "vacation" command was OK
4394 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4395 (I found it when inspecting the code).
4397 20. There is now an overall timeout for performing a callout verification. It
4398 defaults to 4 times the callout timeout, which applies to individual SMTP
4399 commands during the callout. The overall timeout applies when there is more
4400 than one host that can be tried. The timeout is checked before trying the
4401 next host. This prevents very long delays if there are a large number of
4402 hosts and all are timing out (e.g. when the network connections are timing
4403 out). The value of the overall timeout can be changed by specifying an
4404 additional sub-option for "callout", called "maxwait". For example:
4406 verify = sender/callout=5s,maxwait=20s
4408 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4409 to the end before writing, but this should make it even safer).
4411 22. Exim was forgetting that it had advertised PIPELINING for the second and
4412 subsequent messages on an SMTP connection. It was also not resetting its
4413 memory on STARTTLS and an internal HELO.
4415 23. When Exim logs an SMTP synchronization error within a session, it now
4416 records whether PIPELINING has been advertised or not.
4418 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4419 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4420 rather than long int.
4422 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4424 26. Added the never_mail option to autoreply.
4430 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4431 crash if the getsockname() call failed; this can happen if a connection is
4432 closed very soon after it is established. The problem was simply in the
4433 order in which certain operations were done, causing Exim to try to write
4434 to the SMTP stream before it had set up the file descriptor. The bug has
4435 been fixed by making things happen in the correct order.
4441 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4442 before the connection was closed, thus losing the rejection response.
4444 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4445 some early Solaris releases, but causes trouble in current releases where
4446 socklen_t is defined.
4448 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4451 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4454 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4455 that releasing the top of it at the end releases what was used for sub-
4456 expansions (unless the block got too big). However, discard this block if
4457 the first thing is a variable or header, so that we can use its block when
4458 it is dynamic (useful for very large $message_headers, for example).
4460 6. Lookups now cache *every* query, not just the most recent. A new, separate
4461 store pool is used for this. It can be recovered when all lookup caches are
4462 flushed. Lookups now release memory at the end of their result strings.
4463 This has involved some general refactoring of the lookup sources.
4465 7. Some code has been added to the store_xxx() functions to reduce the amount
4466 of flapping under certain conditions.
4468 8. log_incoming_interface used to affect only the <= reception log lines. Now
4469 it causes the local interface and port to be added to several more SMTP log
4470 lines, for example "SMTP connection from", and rejection lines.
4472 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4474 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4476 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4477 could be overwritten at the end of the current message (or the start of a
4478 new message if it was set in a HELO ACL). The value is now preserved for
4479 the duration of the SMTP connection.
4481 12. If a transport had a headers_rewrite setting, and a matching header line
4482 contained an unqualified address, that address was qualified, even if it
4483 did not match any rewriting rules. The underlying bug was that the values
4484 of the flags that permit the existence of unqualified sender and recipient
4485 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4486 non-local messages, and by -bnq for local messages) were not being
4487 preserved with the message after it was received.
4489 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4490 "next input=" as part of the text comprising the host identity instead of
4491 the correct text. The code was using the same buffer for two different
4492 strings. However, depending on which order the printing function evaluated
4493 its arguments, the bug did not always show up. Under Linux, for example, my
4494 test suite worked just fine.
4496 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4497 doesn't work with some older versions of Perl. It has been changed to "my",
4498 which in any case is probably the better facility to use.
4500 15. A really picky compiler found some instances of statements for creating
4501 error messages that either had too many or two few arguments for the format
4504 16. The size of the buffer for calls to the DNS resolver has been increased
4505 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4506 for addresses that have a lot of PTR records. This alleviates a problem; it
4507 does not fully solve it.
4509 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4510 buffer now truncates the list and logs the incident, which is the same
4511 action as happens when Exim is looking up a host name and its aliases.
4512 Previously in this situation something unpredictable would happen;
4513 sometimes it was "internal error: store_reset failed".
4515 18. If a server dropped the connection unexpectedly when an Exim client was
4516 using GnuTLS and trying to read a response, the client delivery process
4517 crashed while trying to generate an error log message.
4519 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4520 string, for example:
4522 warn message = H1: something\nH2: something
4524 the text was added as a single header line from Exim's point of view
4525 though it ended up OK in the delivered message. However, searching for the
4526 second and subsequent header lines using $h_h2: did not work. This has been
4527 fixed. Similarly, if a system filter added multiple headers in this way,
4528 the routers could not see them.
4530 20. Expanded the error message when iplsearch is called with an invalid key to
4531 suggest using net-iplsearch in a host list.
4533 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4534 ACLs are no longer actually imposed (and a message to that effect is
4537 22. If a "gecos" field in a passwd entry contained escaped characters, in
4538 particular, if it contained a \" sequence, Exim got it wrong when building
4539 a From: or a Sender: header from that name. A second bug also caused
4540 incorrect handling when an unquoted " was present following a character
4541 that needed quoting.
4543 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4544 was not being matched caselessly.
4546 24. Arranged for all hyphens in the exim.8 source to be escaped with
4549 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4550 properly. Recipient callout cache records were still being keyed to include
4551 the sender, even when use_sender was set false. This led to far more
4552 callouts that were necessary. The sender is no longer included in the key
4553 when use_sender is false.
4555 26. Added "control = submission" modifier to ACLs.
4557 27. Added the ${base62d: operator to decode base 62 numbers.
4559 28. dnsdb lookups can now access SRV records.
4561 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4562 the configuration file.
4564 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4565 (-v) mode. This makes the output for a verbose queue run more intelligible.
4567 31. Added a use_postmaster feature to recipient callouts.
4569 32. Added the $body_zerocount variable, containing the number of binary zero
4570 bytes in the message body.
4572 33. The time of last modification of the "new" subdirectory is now used as the
4573 "mailbox time last read" when there is a quota error for a maildir
4576 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4578 35. Added +ignore_unknown as a special item in host lists.
4580 36. Code for decoding IPv6 addresses in host lists is now included, even if
4581 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4582 address was recognized as an IP address, but was then not correctly decoded
4583 into binary, causing unexpected and incorrect effects when compared with
4590 1. Very minor rewording of debugging text in manualroute to say "list of
4591 hosts" instead of "hostlist".
4593 2. If verify=header_syntax was set, and a header line with an unqualified
4594 address (no domain) and a large number of spaces between the end of the
4595 name and the colon was received, the reception process suffered a buffer
4596 overflow, and (when I tested it) crashed. This was caused by some obsolete
4597 code that should have been removed. The fix is to remove it!
4599 3. When running in the test harness, delay a bit after writing a bounce
4600 message to get a bit more predictability in the log output.
4602 4. Added a call to search_tidyup() just before forking a reception process. In
4603 theory, someone could use a lookup in the expansion of smtp_accept_max_
4604 per_host which, without the tidyup, could leave open a database connection.
4606 5. Added the variables $recipient_data and $sender_data which get set from a
4607 lookup success in an ACL "recipients" or "senders" condition, or a router
4608 "senders" option, similar to $domain_data and $local_part_data.
4610 6. Moved the writing of debug_print from before to after the "senders" test
4613 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4614 problems for message scanning, either using a data ACL, or using
4615 local_scan() because the Received: header was not generated till after they
4616 were called (in order to set the time as the time of reception completion).
4617 I have revised the way this works. The header is now generated after the
4618 body is received, but before the ACL or local_scan() are called. After they
4619 are run, the timestamp in the header is updated.
4625 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4626 before starting a queue runner without re-exec. This happened only when
4627 deliver_drop_privilege was set or when the Exim user was set to root. The
4628 effect of the bug was that timeouts during subsequent deliveries caused
4629 crashes instead of being properly handled. The handler is now left at its
4630 default (and expected) setting.
4632 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4633 message, again when deliver_drop_privilege is set or Exim is run as root.
4634 The bug described in (1) was not present in this case, but the tidying up
4635 of the other signals was missing. I have made the two cases consistent.
4637 3. The ignore_target_hosts setting on a manualroute router was being ignored
4638 for hosts that were looked up using the /MX notation.
4640 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4643 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4644 operated on the sender address. After changing the $sender_address to <>
4645 for the sender address verify, Exim was re-instated it as the original
4646 (before rewriting) address, but remembering that it had rewritten it, so it
4647 wasn't rewriting it again. This bug also had the effect of breaking the
4648 sender address verification caching when the sender address was rewritten.
4650 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4651 This has been changed so that if the ip literal address matches
4652 ignore_target_hosts, the router declines.
4654 7. Added expansion conditions match_domain, match_address, and match_local_
4655 part (NOT match_host).
4657 8. The placeholder for the Received: header didn't have a length field set.
4659 9. Added code to Exim itself and to exim_lock to test for a specific race
4660 condition that could lead to file corruption when using MBX delivery. The
4661 issue is with the lockfile that is created in /tmp. If this file is removed
4662 after a process has opened it but before that process has acquired a lock,
4663 there is the potential for a second process to recreate the file and also
4664 acquire a lock. This could lead to two Exim processes writing to the file
4665 at the same time. The added code performs the same test as UW imapd; it
4666 checks after acquiring the lock that its file descriptor still refers to
4667 the same named file.
4669 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4670 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4673 11. Added the smtp_active_hostname option. If used, this will typically be made
4674 to depend on the incoming interface address. Because $interface_address is
4675 not set up until the daemon has forked a reception process, error responses
4676 that can happen earlier (such as "too many connections") no longer contain
4679 12. If an expansion in a condition on a "warn" statement fails because a lookup
4680 defers, the "warn" statement is abandoned, and the next ACL statement is
4681 processed. Previously this caused the whole ACL to be aborted.
4683 13. Added the iplsearch lookup type.
4685 14. Added ident_timeout as a log selector.
4687 15. Added tls_certificate_verified as a log selector.
4689 16. Added a global option tls_require_ciphers (compare the smtp transport
4690 option of the same name). This controls incoming TLS connections.
4692 17. I finally figured out how to make tls_require_ciphers do a similar thing
4693 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4694 before starting the TLS session.
4696 18. Tabs are now shown as \t in -bP output.
4698 19. If the log selector return_path_on_delivery was set, Exim crashed when
4699 bouncing a message because it had too many Received: header lines.
4701 20. If two routers both had headers_remove settings, and the first one included
4702 a superfluous trailing colon, the final name in the first list and the
4703 first name in the second list were incorrectly joined into one item (with a
4704 colon in the middle).
4710 1. Added -C and -D options to the exinext utility, mainly to make it easier
4711 to include in the automated testing, but these could be helpful when
4712 multiple configurations are in use.
4714 2. The exinext utility was not formatting the output nicely when there was
4715 an alternate port involved in the retry record key, nor when there was a
4716 message id as well (for retries that were specific to a specific message
4717 and a specific host). It was also confused by IPv6 addresses, because of
4718 the additional colons they contain. I have fixed the IPv4 problem, and
4719 patched it up to do a reasonable job for IPv6.
4721 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4722 delivery, the log line now contains "pipelined" if PIPELINING was used.
4724 4. An SMTP transport process used to panic and die if the bind() call to set
4725 an explicit outgoing interface failed. This has been changed; it is now
4726 treated in the same way as a connect() failure.
4728 5. A reference to $sender_host_name in the part of a conditional expansion
4729 that was being skipped was still causing a DNS lookup. This no longer
4732 6. The def: expansion condition was not recognizing references to header lines
4733 that used bh_ and bheader_.
4735 7. Added the _cache feature to named lists.
4737 8. The code for checking quota_filecount in the appendfile transport was
4738 allowing one more file than it should have been.
4740 9. For compatibility with Sendmail, the command line option
4748 and sets the incoming protocol and host name (for trusted callers). The
4749 host name and its colon can be omitted when only the protocol is to be set.
4750 Note the Exim already has two private options, -pd and -ps, that refer to
4751 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4752 "s", but I don't think that's a major issue.
4754 10. A number of refactoring changes to the code, none of which should affect
4757 (a) The number of logging options was getting close to filling up the
4758 32-bit word that was used as a bit map. I have split them into two classes:
4759 those that are passed in the argument to log_write(), and those that are
4760 only ever tested independently outside of that function. These are now in
4761 separate 32-bit words, so there is plenty of room for expansion again.
4762 There is no change in the user interface or the logging behaviour.
4764 (b) When building, for example, log lines, the code previously used a
4765 macro that called string_cat() twice, in order to add two strings. This is
4766 not really sufficiently general. Furthermore, there was one instance where
4767 it was actually wrong because one of the argument was used twice, and in
4768 one call a function was used. (As it happened, calling the function twice
4769 did not affect the overall behaviour.) The macro has been replaced by a
4770 function that can join an arbitrary number of extra strings onto a growing
4773 (c) The code for expansion conditions now uses a table and a binary chop
4774 instead of a serial search (which was left over from when there were very
4775 few conditions). Also, it now recognizes conditions like "pam" even when
4776 the relevant support is not compiled in: a suitably worded error message is
4777 given if an attempt is made to use such a condition.
4779 11. Added ${time_interval:xxxxx}.
4781 12. A bug was causing one of the ddress fields not to be passed back correctly
4782 from remote delivery subprocesses. The field in question was not being
4783 subsequently used, so this caused to problems in practice.
4785 13. Added new log selectors queue_time and deliver_time.
4787 14. Might have fixed a bug in maildirsizefile handling that threw up
4788 "unexpected character" debug warnings, and recalculated the data
4789 unnecessarily. In any case, I expanded the warning message to give more
4792 15. Added the message "Restricted characters in address" to the statements in
4793 the default ACL that block characters like @ and % in local parts.
4795 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4796 Three changes have been made:
4798 (a) There was a serious bug; a negative response to MAIL caused the whole
4799 recipient domain to be cached as invalid, thereby blocking all messages
4800 to all local parts at the same domain, from all senders. This bug has
4801 been fixed. The domain is no longer cached after a negative response to
4802 MAIL if the sender used is not empty.
4804 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4807 (c) A new callout option, "use_sender" has been added for people who want
4808 the modified behaviour.
4814 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4817 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4818 indeed breaks things for older releases.
4820 3. Added additional logging to the case where there is a problem reading data
4821 from a filter that is running in a subprocess using a pipe, in order to
4822 try to track down a specific problem.
4824 4. Testing facility fudge: when running in the test harness and attempting
4825 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4826 getting "No route to host". Convert this to a timeout.
4828 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4831 6. Some OS don't have socklen_t but use size_t instead. This affects the
4832 fifth argument of getsockopt() amongst other things. This is now
4833 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4834 can be set for individual OS. I have set it for SunOS5, OSF1, and
4835 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4836 some earlier ones do not.
4838 7. Change 4.30/15 was not doing the test caselessly.
4840 8. The standard form for an IPv6 address literal was being rejected by address
4841 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4842 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4843 this, as well as the form without the "IPv6" on the front (but only when
4844 address literals are enabled, of course).
4846 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4848 10. Exim crashed if a message with an empty sender address specified by -f
4849 encountered a router with an errors_to setting. This could be provoked only
4850 by a command such as
4854 where an empty string was supplied; "<>" did not hit this bug.
4856 11. Installed PCRE release 4.5.
4858 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4859 remained set. It is now erased.
4861 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4862 times from message ids (which are base 36 rather than the normal 62).
4864 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4865 were being counted as actual protocol errors, and logged if the log
4866 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4867 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4868 and DATA following a set of rejected RCPTs do not count as protocol errors.
4869 In other words, Exim assumes they were pipelined, though this may not
4870 actually be the case. Of course, in all cases the client gets an
4871 appropriate error code.
4873 15. If a lookup fails in an ACL condition, a message about the failure may
4874 be available; it is used if testing the ACL cannot continue, because most
4875 such messages specify what the cause of the deferral is. However, some
4876 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4877 that caused an old message to be retained and used if a later statement
4878 caused a defer, replacing the real cause of the deferral.
4880 16. If an IP address had so many PTR records that the DNS lookup buffer
4881 was not large enough to hold them, Exim could crash while trying to process
4882 the truncated data. It now detects and logs this case.
4884 17. Further to 4.21/58, another change has been made: if (and only if) the
4885 first line of a message (the first header line) ends with CRLF, a bare LF
4886 in a subsequent header line has a space inserted after it, so as not to
4887 terminate the header.
4889 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4890 unnecessarily, used atoi() instead of strtol(), and didn't check the
4891 termination when getting file sizes from file names by regex.
4893 19. Completely re-implemented the support for maildirsize files, in the light
4894 of a number of problems with the previous contributed implementation
4895 (4.30/29). In particular:
4897 . If the quota is zero, the maildirsize file is maintained, but no quota is
4900 . If the maildir directory does not exist, it is created before any attempt
4901 to write a maildirsize file.
4903 . The quota value in the file is just a cache; if the quota is changed in
4904 the transport, the new value overrides.
4906 . A regular expression is available for excluding directories from the
4909 20. The autoreply transport checks the characters in options that define the
4910 message's headers; it allows continued headers, but it was checking with
4911 isspace() after an embedded newline instead of explicitly looking for a
4914 21. If all the "regular" hosts to which an address was routed had passed their
4915 expiry times, and had not reached their retry times, the address was
4916 bounced, even if fallback hosts were defined. Now Exim should go on to try
4919 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4920 equivalent code in the SMTP transport. Some hosts send humungous responses
4921 to HELO/EHLO, more than 1024 it seems.
4923 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4924 gives compiler warnings in some environments. I've now done it "properly",
4927 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4928 (because of problems with the built-in one) was declared to return uschar *
4929 instead of char *, causing compiler failure.
4931 25. Fixed a file descriptor leak when processing alias/forward files.
4933 26. Fixed a minor format string issue in dbfn.c.
4935 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4937 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4938 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4941 29. When a sender address is verified, it is cached, to save repeating the test
4942 when there is more than one recipient in a message. However, when the
4943 verification involves a callout, it is possible for different callout
4944 options to be set for different recipients. It is too complicated to keep
4945 track of this in the cache, so now Exim always runs a verification when a
4946 callout is required, relying on the callout cache for the optimization.
4947 The overhead is duplication of the address routing, but this should not be
4950 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4951 to be verified with callout=postmaster, and the main callout worked but the
4952 postmaster check failed, the verification correctly failed. However, if a
4953 subsequent RCPT command asked for sender verification *without* the
4954 postmaster check, incorrect caching caused this verification also to fail,
4957 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4958 it was not caching the DNS options (qualify_single, search_parents) that
4959 were used when the lookup failed. A subsequent lookup with different
4960 options therefore always gave the same answer, though there were cases
4961 where it should not have. (Example: a "domains = !$mx_any" option on a
4962 dnslookup router: the "domains" option is always processed without any
4963 widening, but the router might have qualify_single set.) Now Exim uses the
4964 cached value only when the same options are set.
4966 32. Added John Jetmore's "exipick" utility to the distribution.
4968 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4969 than a timeout (e.g. a certificate is required, and is not provided), an
4970 Exim server now closes the connection immediately. Previously it waited for
4971 the client to close - but if the client is SSL, it seems that they each
4972 wait for each other, leading to a delay before one of them times out.
4974 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4975 maintained 0.8.x compatibility because I don't think many are using it, and
4976 it is clearly obsolete.
4978 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4981 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4982 client certificate was expired. A simple patch fixes this, though I don't
4983 understand the full logic of why the verify callback is called multiple
4986 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4987 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4988 which causes problems with some clients (such as the Certicom SSL Plus
4989 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4990 disables the coutermeasure allowing Eudora to connect."
4992 38. Exim was not checking that a write() to a log file succeeded. This could
4993 lead to Bad Things if a log got too big, in particular if it hit a file
4994 size limit. Exim now panics and dies if it cannot write to a log file, just
4995 as it does if it cannot open a log file.
4997 39. Modified OS/Makefile-Linux so that it now contains
4999 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5001 The two -D definitions ensure that Exim is compiled with large file
5002 support, which makes it possible to handle log files that are bigger than
5005 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5006 instance) a domain was checked against a named list that involved a lookup,
5007 causing $domain_data to be set, then another domain was checked against the
5008 same list, then the first domain was re-checked, the value of $domain_data
5009 after the final check could be wrong. In particular, if the second check
5010 failed, it could be set empty. This bug probably also applied to
5013 41. The strip_trailing_dot option was not being applied to the address given
5014 with the -f command-line option.
5016 42. The code for reading a message's header from the spool was incrementing
5017 $received_count, but never initializing it. This meant that the value was
5018 incorrect (doubled) while delivering a message in the same process in which
5019 it was received. In the most common configuration of Exim, this never
5020 happens - a fresh exec is done - but it can happen when
5021 deliver_drop_privilege is set.
5023 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5024 it now includes up to 150 characters of the unexpected data in the log
5027 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5028 and building data strings. The size of both of these buffers was 10 000
5029 bytes - far larger than anybody would *ever* want, thought I. Needless to
5030 say, somebody hit the limit. I have increased the maximum line length to
5031 20 000 and the maximum data length of concatenated lines to 100 000. I have
5032 also fixed two bugs, because there was no checking on these buffers. Tsk,
5033 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5034 buffer is too small.
5036 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5037 lsearch lookups. Now it does.
5039 46. When parsing a route_list item in a manualroute router, a fixed-length
5040 buffer was used for the list of hosts. I made this 1024 bytes long,
5041 thinking that nobody would ever have a list of hosts that long. Wrong.
5042 Somebody had a whole pile of complicated expansion conditions, and the
5043 string was silently truncated, leading to an expansion error. It turns out
5044 that it is easier to change to an unlimited length (owing to other changes
5045 that have happened since this code was originally written) than to build
5046 structure for giving a limitation error. The length of the item that
5047 expands into the list of hosts is now unlimited.
5049 47. The lsearch lookup could not handle data where the length of text line was
5050 more than 4095 characters. Such lines were truncated, leading to shortened
5051 data being returned. It should now handle lines of any length.
5053 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5054 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5057 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5058 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5059 no longer happens in comment lines. A list of replacements is now placed
5060 at the head of all of the source files, except those whose only change is
5061 to replace PERL_COMMAND in the very first #! line.
5063 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5064 messages on the queue, with a bottom-up merge sort, using code contributed
5065 by Michael Haardt. This should make operations like -bp somewhat faster on
5066 large queues. It won't affect queue runners, except when queue_run_in_order
5069 51. Installed eximstats 1.31 in the distribution.
5071 52. Added support for SRV lookups to the dnslookup router.
5073 53. If an ACL referred to $message_body or $message_body_end, the value was not
5074 reset for any messages that followed in the same SMTP session.
5076 54. The store-handling optimization for building very long strings was not
5077 differentiating between the different store pools. I don't think this
5078 actually made any difference in practice, but I've tidied it.
5080 55. While running the routers to verify a sender address, $sender_address
5081 was still set to the sender address. This is wrong, because when routing to
5082 send a bounce to the sender, it would be empty. Therefore, I have changed
5083 it so that, while verifying a sender address, $sender_address is set to <>.
5084 (There is no change to what happens when verifying a recipient address.)
5086 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5087 target A or AAAA records (if not already returned) without resetting the
5088 qualify_single or search_parents options of the DNS resolver. These are
5089 inappropriate in this case because the targets of MX and SRV records must
5090 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5091 target that, when qualified, matched something in the local domain. These
5092 two options are now turned off when doing these lookups.
5094 57. It seems that at least some releases of Reiserfs (which does not have the
5095 concept of a fixed number of inodes) returns zero and not -1 for the
5096 number of available inodes. This interacted badly with check_spool_inodes,
5097 which assumed that -1 was the "no such thing" setting. What I have done is
5098 to check that the total number of inodes is greater than zero before doing
5099 the test of how many are available.
5101 58. When a "warn" ACL statement has a log_message modifier, the message is
5102 remembered, and not repeated. This is to avoid a lot of repetition when a
5103 message has many recipients that cause the same warning to be written.
5104 Howewer, Exim was preserving the list of already written lines for an
5105 entire SMTP session, which doesn't seem right. The memory is now reset if a
5106 new message is started.
5108 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5109 debugging output unless log_rewrite was also set.
5111 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5112 of (int)(handle) when we know that handle contains (void *)(-1).
5114 61. The Exim daemon panic-logs an error return when it closes the incoming
5115 connection. However "connection reset by peer" seems to be common, and
5116 isn't really an error worthy of noting specially, so that particular error
5119 62. When Exim is trying to find all the local interfaces, it used to panic and
5120 die if the ioctl to get the interface flags failed. However, it seems that
5121 on at least one OS (Solaris 9) it is possible to have an interface that is
5122 included in the list of interfaces, but for which you get a failure error
5123 for this call. This happens when the interface is not "plumbed" into a
5124 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5125 failure of the "get flags" call assumes that the interface is down.
5127 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5128 makes life easier for people who are doing arithmetic on fields extracted
5129 from dates, where you often get leading zeros that should not be
5130 interpreted as octal.
5132 64. Added qualify_domain to the redirect router, to override the global
5135 65. If a pathologically long header line contained very many addresses (the
5136 report of this problem mentioned 10 000) and each of them was rewritten,
5137 Exim could use up a very large amount of memory. (It kept on making new
5138 copies of the header line as it rewrote, and never released the old ones.)
5139 At the expense of a bit more processing, the header rewriting function has
5140 been changed so that it no longer eats memory in this way.
5142 66. The generation of the Received: header has been moved from the time that a
5143 message starts to be received, to the time that it finishes. The timestamp
5144 in the Received: header should now be very close to that of the <= log
5145 line. There are two side-effects of this change:
5147 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5148 logged header lines no longer include the local Received: line, because
5149 it has not yet been created. The same applies to a copy of the message
5150 that is returned to a non-SMTP sender when a message is rejected.
5152 (b) When a filter file is tested using -bf, no additional Received: header
5153 is added to the test message. After some thought, I decided that this
5156 This change does not affect the value of $received_for. It is still set
5157 after address rewriting, but before local_scan() is called.
5159 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5161 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5162 gave an unhelpful panic error message, and a defer error. I have managed to
5163 change this behaviour so that it now rejects any supplied certificate,
5164 which seems right, as the list of acceptable certificates is empty.
5166 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5167 gave an unhelpful defer error. I have not managed to make this reject any
5168 supplied certificates, but the error message it gives is "no certificate
5169 supplied", which is not helpful.
5171 70. exigrep's output now also includes lines that are not associated with any
5172 message, but which match the given pattern. Implemented by a patch from
5173 Martin Sluka, which also tidied up the Perl a bit.
5175 71. Recipient callout verification, like sender verification, was using <> in
5176 the MAIL FROM command. This isn't really the right thing, since the actual
5177 sender may affect whether the remote host accepts the recipient or not. I
5178 have changed it to use the actual sender in the callout; this means that
5179 the cache record is now keyed on a recipient/sender pair, not just the
5180 recipient address. There doesn't seem to be a real danger of callout loops,
5181 since a callout by the remote host to check the sender would use <>.
5182 [SEE ABOVE: changed after hitting problems.]
5184 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5185 temporary errors. However, in the case of such a code being given after
5186 the end of a data transmission (i.e. after ".") Exim was failing to write
5187 a retry record for the message. (Yes, there was some broken host that was
5188 actually sending 8xx at this point.)
5190 73. An unknown lookup type in a host list could cause Exim to panic-die when
5191 the list was checked. (An example that provoked this was putting <; in the
5192 middle of a list instead of at the start.) If this happened during a DATA
5193 ACL check, a -D file could be left lying around. This kind of configuration
5194 error no longer causes Exim to die; instead it causes a defer errror. The
5195 incident is still logged to the main and panic logs.
5197 74. Buglet left over from Exim 3 conversion. The message "too many messages
5198 in one connection" was written to the rejectlog but not the mainlog, except
5199 when address rewriting (yes!) was being logged.
5201 75. Added write_rejectlog option.
5203 76. When a system filter was run not as root (that is, when system_filter_user
5204 was set), the values of the $n variables were not being returned to the
5205 main process; thus, they were not subsequently available in the $sn
5208 77. Added +return_path_on_delivery log selector.
5210 78. A connection timeout was being treated differently from recipients deferred
5211 when testing hosts_max_try with a message that was older than the host's
5212 retry timeout. (The host should not be counted, thus allowing all hosts to
5213 be tried at least once before bouncing.) This may have been the cause of an
5214 occasionally reported bug whereby a message would remain on the queue
5215 longer than the retry timeout, but would be bounced if a delivery was
5216 forced. I say "may" because I never totally pinned down the problem;
5217 setting up timeout/retry tests is difficult. See also the next item.
5219 79. The ultimate address timeout was not being applied to errors that involved
5220 a combination of host plus message (for example, a timeout on a MAIL
5221 command). When an address resolved to a number of possible hosts, and they
5222 were not all tried for each delivery (e.g. because of hosts_max_try), a
5223 message could remain on the queue longer than the retry timeout.
5225 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5228 81. Fixed an obscure SMTP outgoing bug which required at least the following
5229 conditions: (a) there was another message waiting for the same server;
5230 (b) the server returned 5xx to all RCPT commands in the first message so
5231 that the message was not completed; (c) the server dropped the connection
5232 or gave a negative response to the RSET that Exim sends to abort the
5233 transaction. The observed case was a dropped connection after DATA that had
5234 been sent in pipelining mode. That is, the server had advertised PIPELINING
5235 but was not implementing it correctly. The effect of the bug was incorrect
5236 behaviour, such as trying another host, and this could lead to a crash.
5242 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5243 and daemon.c were passed as pointers to ints; they should have been
5244 pointers to socklen_t variables (which are typically unsigned ints).
5246 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5249 3. Fixed a really odd bug that affected only the testing scheme; patching a
5250 certain fixed string in the binary changed the value of another string that
5251 happened to be identical to the end of the original first string.
5253 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5254 name", it returns that address as the IP address. On some operating
5255 systems (e.g. Solaris), it also passes back the IP address string as the
5256 "host name". However, on others (e.g. Linux), it passes back an empty
5257 string. Exim wasn't checking for this, and was changing the host name to an
5258 empty string, assuming it had been canonicized.
5260 5. Although rare, it is permitted to have more than one PTR record for a given
5261 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5262 all the names associated with an address, because they do in Solaris.
5263 However, it seems that they do not in Linux for data that comes from the
5264 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5265 I found this out when I moved to a new Linux workstation and tried to run
5266 the Exim test suite.
5268 To get round this problem I have changed the code so that it now does its
5269 own call to the DNS to look up PTR records when searching for a host name.
5270 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5271 addresses that are only in /etc/hosts are still found.
5273 This behaviour is, however, controlled by an option called host_lookup_
5274 order, which defaults to "bydns:byaddr". If people want to use the other
5275 order, or indeed, just use one or the other means of lookup, they can
5276 specify it in this variable.
5278 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5279 some operating systems, this comes back from gethostbyaddr() as an empty
5280 string, and this is what Exim used to test for. However, it seems that in
5281 other systems, "." is yielded. Exim now tests for this case too.
5283 7. The values of check_spool_space and check_log_space are now held internally
5284 as a number of kilobytes instead of an absolute number of bytes. If a
5285 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5286 kilobyte. This means that much larger values can be stored.
5288 8. Exim monitor: an attempt to get the action menu when not actually pointing
5289 at a message produces an empty menu entitled "No message selected". This
5290 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5291 no entries in it ("Shell widget menu has zero width and/or height"). So I
5292 have added a single, blank menu entry in this case.
5294 9. Added ${quote_local_part.
5296 10. MIME decoding is now applied to the contents of Subject: header lines when
5299 11. Now that a reference to $sender_host_address automatically causes a reverse
5300 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5301 host lookup before query-style lookups in lists that might use this
5302 variable. This has therefore been abolished, and the "net-" prefix is no
5303 longer necessary for query-style lookups.
5305 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5306 have been a typo for LFLAGS, so it has been changed.
5308 13. The install script calls Exim with "-C /dev/null" in order to find the
5309 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5310 to be output. Howeve, since Exim outputs its version number before the
5311 error, it didn't break the script. It just looked ugly. I fixed this by
5312 always allowing "-C /dev/null" if the caller is root.
5314 14. Ignore overlarge ACL variable number when reading spool file - insurance
5315 against a later release with more variables having written the file.
5317 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5318 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5319 this, as well as the form without the "IPv6" on the front.
5321 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5322 OS/Makefile-Darwin file.
5324 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5325 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5327 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5328 variable. However, some OS use ints for inodes. I've added cast to long int
5329 to get rid of the compiler warning.
5331 19. I had forgotten to lock out "/../" in configuration file names when
5332 ALT_CONFIG_PREFIX was set.
5334 20. Routers used for verification do not need to specify transports. However,
5335 if such a router generated a host list, and callout was configured, Exim
5336 crashed, because it could not find a port number from the (non-existent)
5337 transport. It now assumes port 25 in this circumstance.
5339 21. Added the -t option to exigrep.
5341 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5342 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5343 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5344 them is set, LOOKUP_LSEARCH is forced.
5346 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5348 24. Added sender and host information to the "rejected by local_scan()" log
5349 line; previously there was no indication of these.
5351 25. Added .include_if_exists.
5353 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5354 Linux. It turns out that not all file systems support this. Apparently some
5355 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5356 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5357 not supported on the file descriptor, is now ignored when Exim is trying to
5358 sync a directory. This applies only to Linux.
5360 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5362 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5363 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5366 29. Added support for maildirsize files from supplied patch (modified a bit).
5368 30. The use of :fail: followed by an empty string could lead Exim to respond to
5369 sender verification failures with (e.g.):
5371 550 Verification failed for <xxx>
5372 550 Sender verify failed
5374 where the first response line was missing the '-' that indicates it is not
5375 the final line of the response.
5377 31. The loop for finding the name of the user that called Exim had a hardwired
5378 limit of 10; it now uses the value of finduser_retries, which is used for
5379 all other user lookups.
5381 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5384 33. Exim was neglecting to zero errno before one call of strtol() when
5385 expanding a string and expecting an integer value. On some systems this
5386 resulted in spurious "integer overflow" errors. Also, it was casting the
5387 result into an int without checking.
5389 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5390 did not work. The code looks as if it has *never* worked, though it appears
5391 to have been documented since at least releast 1.62. I have made it work.
5393 35. The "timeout_DNS" error in retry rules, also documented since at least
5394 1.62, also never worked. As it isn't clear exactly what this means, and
5395 clearly it isn't a major issue, I have abolished the feature by treating it
5396 as "timeout", and writing a warning to the main and panic logs.
5398 36. The display of retry rules for -brt wasn't always showing the error code
5401 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5402 timeout_connect_A, timeout_connect_MX.
5404 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5405 to the empty sender.
5407 39. The daemon was not analysing the content of -oX till after it had closed
5408 stderr and disconnected from the controlling terminal. This meant that any
5409 syntax errors were only noted on the panic log, and the return code from
5410 the command was 0. By re-arranging the code a little, I've made the
5411 decoding happen first, so such errors now appear on stderr, and the return
5412 code is 1. However, the actual setting up of the sockets still happens in
5413 the disconnected process, so errors there are still only recorded on the
5416 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5417 connections (as happens on some IP stacks) was logged at start up time as
5418 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5419 it from "IPv6 and IPv4", which means that two separate sockets are being
5422 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5423 says whether AF_INET or AF_INET6 was passed as an argument.
5425 42. Exiwhat output was messed up when time zones were included in log
5428 43. Exiwhat now gives more information about the daemon's listening ports,
5429 and whether -tls-on-connect was used.
5431 44. The "port" option of the smtp transport is now expanded.
5433 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5434 silently ignored. Now an error message is written to the main and panic
5437 46. There's a new ACL modifier called "logwrite" which writes to a log file
5438 as soon as it is encountered.
5440 47. Added $local_user_uid and $local_user_gid at routing time.
5442 48. Exim crashed when trying to verify a sender address that was being
5445 49. Exim was recognizing only a space character after ".include". It now also
5446 recognizes a tab character.
5448 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5449 extracting the relevant information from the specification. The man page no
5450 longer contains scrambled data for the -d option, and I've added a section
5451 at the front about calling Exim under different names.
5453 51. Added "extra_headers" argument to the "mail" command in filter files.
5455 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5458 53. Installed eximstats 1.29.
5460 54. Added transport_filter_timeout as a generic transport option.
5462 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5463 Cc: header lines. This was required by RFC 822, but it not required by RFC
5466 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5467 incoming messages that did not have them. Now it does so only if the
5468 message originates locally, that is, if there is no associated remote host
5469 address. When Resent- header lines are present, this applies to the Resent-
5470 lines rather than the non-Resent- lines.
5472 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5473 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5475 58. Messages for configuration errors now include the name of the main
5476 configuration file - useful now that there may be more than one file in a
5477 list (.included file names were always shown).
5479 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5480 for those rare installations that do not start the daemon as root or run it
5481 setuid root. I've cut out the call to initgroups() if the daemon is not
5484 60. The Exim user and group can now be bound into the binary as text strings
5485 that are looked up at the start of Exim's processing.
5487 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5489 62. Added $mailstore_basename variable.
5491 63. Installed patch to sieve.c from Michael Haardt.
5493 64. When Exim failed to open the panic log after failing to open the main log,
5494 the original message it was trying to log was written to stderr and debug
5495 output, but if they were not available (the usual case in production), it
5496 was lost. Now it is written to syslog before the two lines that record the
5497 failures to open the logs.
5499 65. Users' Exim filters run in subprocesses under the user's uid. It is
5500 possible for a "deliver" command or an alias in a "personal" command to
5501 provoke an address rewrite. If logging of address rewriting is configured,
5502 this fails because the process is not running as root or exim. There may be
5503 a better way of dealing with this, but for the moment (because 4.30 needs
5504 to be released), I have disabled address rewrite logging when running a
5505 filter in a non-root, non-exim process.
5511 1. The buildconfig auxiliary program wasn't quoting the value set for
5512 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5513 not defined. This bug was masked in 4.22 by the effect that was fixed in
5516 2. Some messages that were rejected after a message id was allocated were
5517 shown as "incomplete" by exigrep. It no longer does this for messages that
5518 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5520 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5521 have allow_domain_literals set, the ID did not get logged in the <= line.
5522 Domain literals are now always recognized in Message-ID: header lines.
5524 4. The first argument for a ${extract expansion item is the key name or field
5525 number. Leading and trailing spaces in this item were not being ignored,
5526 causing some misleading effects.
5528 5. When deliver_drop_privilege was set, single queue runner processes started
5529 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5530 same command in the process it spins off) were not dropping privilege.
5532 6. When the daemon running as "exim" started a queue runner, it always
5533 re-executed Exim in the spun-off process. This is a waste of effort when
5534 deliver_drop_privilege is set. The new process now just calls the
5535 queue-runner function directly.
5541 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5544 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5545 ignored. Though the use of -oP was forcing the writing of a pid file, it
5546 was always written to the default place.
5548 3. If the message "no IP address found for host xxxx" is generated during
5549 incoming verification, it is now followed by identification of the incoming
5550 connection (so you can more easily find what provoked it).
5552 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5554 5. Added some features to "harden" Exim a bit more against certain attacks:
5556 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5557 be put in Local/Makefile. This is like the never_users runtime option,
5558 but it cannot be overridden. The default setting is "root".
5560 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5561 prefix string with which any file named in a -C command line option
5564 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5565 is retained for -C and -D only if the caller of Exim is root. Without
5566 it, the exim user may also use -C and -D and retain privilege.
5568 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5569 command line option is disabled.
5571 6. Macro names set by the -D option must start with an upper case letter, just
5572 like macro names defined in the configuration file.
5574 7. Added "dereference=" facility to LDAP.
5576 8. Two instances of the typo "uknown" in the source files are fixed.
5578 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5579 the Configure-Makefile script screwed up while processing it.
5581 10. Incorporated PCRE 4.4.
5583 11. The SMTP synchronization check was not operating right at the start of an
5584 SMTP session. For example, it could not catch a HELO sent before the client
5585 waited for the greeting. There is now a check for outstanding input at the
5586 point when the greeting is written. Because of the duplex, asynchronous
5587 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5588 way, but not yet received, when the check is performed.
5590 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5591 on TCP/IP sockets, because this apparently causes some broken clients to
5594 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5595 unchanged) from the Cygwin maintainer.
5597 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5598 was supported instead of testing for mbx. Effectively a typo.
5600 15. The spa authenticator server code was not checking that the input it
5601 received was valid base64.
5603 16. The debug output line for the "set" modifier in ACLs was not showing the
5604 name of the variable that was being set.
5606 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5608 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5609 lookup on its own. Something else was needed to trigger the lookup. For
5610 example, a match in host_lookup or the need for a host name in a host list.
5611 Now, if $sender_host_name is referenced and the host name has not yet been
5612 looked up, a lookup is performed. If the lookup fails, the variable remains
5613 empty, and $host_lookup_failed is set to "1".
5615 19. Added "eqi" as a case-independent comparison operator.
5617 20. The saslauthd authentication condition could segfault if neither service
5618 nor realm was specified.
5620 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5621 error message that was logged was misleading, and incoming SMTP
5622 connections were dropped. The message is now more accurate, and temporary
5623 errors are given to SMTP connections.
5625 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5626 (except RSET) with a 421 error, until QUIT is received. However, it was
5627 failing to send a response to QUIT.
5629 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5630 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5631 ACL is now run after the helo_try_verify_hosts code.
5633 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5634 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5635 case-independent, so other case variants are also recognized). Apparently
5636 some systems use these upper case variants.
5638 25. If more than two messages were waiting for the same host, and a transport
5639 filter was specified for the transport, Exim sent two messages over the
5640 same TCP/IP connection, and then failed with "socket operation on non-
5641 socket" when it tried to send the third.
5643 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5645 27. The extern definition of crypt16() in expand.c was not being excluded when
5646 the OS had its own crypt16() function.
5648 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5649 as a preferred synonym for return_size_limit, both as an option and as an
5652 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5654 30. Changed the default configuration ACL to relax the local part checking rule
5655 for addresses that are not in any local domains. For these addresses,
5656 slashes and pipe symbols are allowed within local parts, but the sequence
5657 /../ is explicitly forbidden.
5659 31. SPA server authentication was not clearing the challenge buffer before
5662 32. log_message in a "warn" ACL statement was writing to the reject log as
5663 well as to the main log, which contradicts the documentation and doesn't
5664 seem right (because no rejection is happening). So I have stopped it.
5666 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5667 However, I am unable to do any testing of this.
5669 34. Fixed an infelicity in the appendfile transport. When checking directories
5670 for a mailbox, to see if any needed to be created, it was accidentally
5671 using path names with one or more superfluous leading slashes; tracing
5672 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5674 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5675 contents are added to the log line that is written for every discarded
5676 recipient. (Previously a log_message setting was ignored.)
5678 36. The ${quote: operator now quotes the string if it is empty.
5680 37. The install script runs exim in order to find its version number. If for
5681 some reason other than non-existence or emptiness, which it checks, it
5682 could not run './exim', it was installing it with an empty version number,
5683 i.e. as "exim-". This error state is now caught, and the installation is
5686 38. An argument was missing from the function that creates an error message
5687 when Exim fails to connect to the socket for saslauthd authentication.
5688 This could cause Exim to crash, or give a corrupted message.
5690 39. Added isip, isip4, and isip6 to ${if conditions.
5692 40. The ACL variables $acl_xx are now saved with the message, and can be
5693 accessed later in routers, transports, and filters.
5695 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5696 strings in the file are not string-expanded.
5698 42. If a MAIL command specified a SIZE value that was too large to fit into an
5699 int variable, the check against message_size_limit failed. Such values are
5700 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5701 day this will have to be increased, but I don't think I want to be around
5702 when emails are that large.
5709 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5710 iconv() is not standard in FreeBSD.
5712 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5713 IPv6 enabled. The observed symptom was a segmentation fault on return from
5714 the function os_common_find_running_interfaces() in src/os.c.
5716 3. In the check_special_case() function in daemon.c I had used "errno" as an
5717 argument name, which causes warnings on some systems. This was basically a
5718 typo, since it was named "eno" in the comments!
5720 4. The code that waits for the clock to tick (at a resolution of some fraction
5721 of a second) so as to ensure message-id uniqueness was always waiting for
5722 at least one whole tick, when it could have waited for less. [This is
5723 almost certainly not relevant at current processor speeds, where it is
5724 unlikely to ever wait at all. But we try to future-proof.]
5726 5. The function that sleeps for a time interval that includes fractions of a
5727 second contained a race. It did not block SIGALRM between setting the
5728 timer, and suspending (a couple of lines later). If the interval was short
5729 and the sigsuspend() was delayed until after it had expired, the suspension
5730 never ended. On busy systems this could lead to processes getting stuck for
5733 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5734 process, before it forks any delivery processes. The open lookup caching
5735 mechanism meant that the open file or database connection was passed into
5736 the delivery process. The problem was that delivery processes always tidy
5737 up cached lookup data. This could cause a problem for the next delivery
5738 process started by the queue runner, because the external queue runner
5739 process does not know about the closure. So the next delivery process
5740 still has data in the lookup cache. In the case of a file lookup, there was
5741 no problem because closing a file descriptor in a subprocess doesn't affect
5742 the parent. However, if the lookup was caching a connection to a database,
5743 the connection was closed, and the second delivery process was likely to
5744 see errors such as "PGSQL: query failed: server closed the connection
5745 unexpectedly". The problem has been fixed by closing all cached lookups
5746 in a queue runner before running a delivery process.
5748 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5749 seem to have the "const" qualifier which it has on other OS. I've
5752 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5753 *of the same type* (client or server) with the same public name that an
5754 error should be diagnosed.
5756 9. When Exim looked up a host name for an IP address, but failed to find the
5757 original IP address when looking up the host name (a safety check), it
5758 output the message "<ip address> does not match any IP for NULL", which was
5759 confusing, to say the least. The bug was that the host name should have
5760 appeared instead of "NULL".
5762 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5763 user that is built into the binary, and the -C or -D options is used, root
5764 privilege is dropped before the configuration file is read. In addition,
5765 logging is switched to stderr instead of the normal log files. If the
5766 configuration then re-defines the Exim user, the unprivileged environment
5767 is probably not what is expected, so Exim logs a panic warning message (but
5770 However, if deliver_drop_privilege is set, the unprivileged state may well
5771 be exactly what is intended, so the warning has been cut out in that case,
5772 and Exim is allowed to try to write to its normal log files.
5778 1. smtp_return_error_details was not giving details for temporary sender
5779 or receiver verification errors.
5781 2. Diagnose a configuration error if two authenticators have the same public
5784 3. Exim used not to create the message log file for a message until the first
5785 delivery attempt. This could be confusing when incoming messages were held
5786 for policy or load reasons. The message log file is now created at the time
5787 the message is received, and an initial "Received" line is written to it.
5789 4. The automatically generated man page for command line options had a minor
5790 bug that caused no ill effects; however, a more serious problem was that
5791 the procedure for building the man page automatically didn't always
5792 operate. Consequently, release 4.20 contains an out-of-date version. This
5793 shouldn't happen again.
5795 5. When building Exim with embedded Perl support, the script that builds the
5796 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5797 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5799 6. The freeze_tell option was not being used for messages that were frozen on
5800 arrival, either by an ACL or by local_scan().
5802 7. Added the smtp_incomplete_transaction log selector.
5804 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5805 was accepting AUTH without a new EHLO.
5807 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5808 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5811 10. Logging of TCP/IP connections (when configured) now happens in the main
5812 daemon process instead of the child process, so that the TCP/IP connection
5813 count is more accurate (but it can never be perfect).
5815 11. The use of "drop" in a nested ACL was not being handled correctly in the
5816 outer ACL. Now, if condition failure induced by the nested "drop" causes
5817 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5818 or "require"), the connection is dropped.
5820 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5821 that yield "discard" can now be used with an "accept" or a "discard" verb,
5822 but an error is generated for any others (because I can't see a useful way
5823 to define what should happen).
5825 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5826 are now processed in the same way as lines in the Exim configuration file.
5827 In particular, continuation lines are supported.
5829 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5831 15. Added -ti meaning -t -i.
5833 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5834 domains, and warn by logging if others are found.
5836 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5837 returned from GIFCONF when Exim is trying to find the list of interfaces on
5838 a host. The code in os.c has been modified to copy each ifreq to an aligned
5839 structure in all cases.
5841 Also, in some cases, the returned ifreq's were being copied to a 'struct
5842 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5843 means the last couple of bytes of an IPv6 address could be chopped if the
5844 ifreq contained only a normal sockaddr (14 bytes storage).
5846 18. Named domain lists were not supported in the hosts_treat_as_local option.
5847 An entry such as +xxxx was not recognized, and was treated as a literal
5850 19. Ensure that header lines added by a DATA ACL are included in the reject log
5851 if the ACL subsequently rejects the message.
5853 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5854 MD5 (which is deprecated).
5856 21. When testing a filter file using -bf, Exim was writing a message when it
5857 took the sender from a "From " line in the message, but it was not doing so
5858 when it took $return_path from a Return-Path: header line. It now does.
5860 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5861 with a valid header line field name (a series of printing characters
5862 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5864 23. Changed "disc" in the source to "disk" to conform to the documentation and
5865 the book and for uniformity.
5867 24. Ignore Sendmail's -Ooption=value command line item.
5869 25. When execve() failed while trying to run a command in a pipe transport,
5870 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5871 could be confused with a return value of 69 from the command itself. This
5872 has been changed to 127, the value the shell returns if it is asked to run
5873 a non-existent command. The wording for the related log line suggests a
5874 non-existent command as the problem.
5876 26. If received_header_text expands to an empty string, do not add a Received:
5877 header line to the message. (Well, it adds a token one on the spool, but
5878 marks it "old" so that it doesn't get used or transmitted.)
5880 27. Installed eximstats 1.28 (addition of -nt option).
5882 28. There was no check for failure on the call to getsockname() in the daemon
5883 code. This can fail if there is a shortage of resources on the system, with
5884 ENOMEM, for example. A temporary error is now given on failure.
5886 29. Contrary to the C standard, it seems that in some environments, the
5887 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5888 program. Exim now does this explicitly; it affects the formatting of
5889 timestamps using strftime().
5891 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5892 complaints. I've now initialized all the variables, to avoid this.
5894 32. Header lines added by a system filter were not being "seen" during
5895 transport-time rewrites.
5897 33. The info_callback() function passed to OpenSSL is set up with type void
5898 (*)(SSL *, int, int), as described somewhere. However, when calling the
5899 function (actually a macro) that sets it up, the type void(*)() is
5900 expected. I've put in a cast to prevent warnings from picky compilers.
5902 34. If a DNS black list lookup found a CNAME record, but there were no A
5903 records associated with the domain it pointed at, Exim crashed.
5905 35. If a DNS black list lookup returned more than one A record, Exim ignored
5906 all but the first. It now scans all returned addresses if a particular IP
5907 value is being sought. In this situation, the contents of the
5908 $dnslist_value variable are a list of all the addresses, separated by a
5911 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5912 to accept a host name and all its aliases if the forward lookup for any of
5913 them yielded the IP address of the incoming connection. Now it accepts only
5914 those names whose forward lookup yields the correct IP address. Any other
5915 names are discarded. This closes a loophole whereby a rogue DNS
5916 administrator could create reverse DNS records to break through a
5917 wildcarded host restriction in an ACL.
5919 37. If a user filter or a system filter that ran in a subprocess used any of
5920 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5921 the wrong values were passed to the pipe command ($thisaddress had the
5922 value of $0, $0 had the value of $1, etc). This bug was introduced by
5923 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5925 38. Improved the line breaking for long SMTP error messages from ACLs.
5926 Previously, if there was no break point between 40 and 75 characters, Exim
5927 left the rest of the message alone. Two changes have been made: (a) I've
5928 reduced the minimum length to 35 characters; (b) if it can't find a break
5929 point between 35 and 75 characters, it looks ahead and uses the first one
5930 that it finds. This may give the occasional overlong line, but at least the
5931 remaining text gets split now.
5933 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5934 file descriptors might be low, and that setting 1000 would always raise it.
5935 It turns out that in some environments, the limit is already over 1000 and
5936 that lowering it causes trouble. So now Exim takes care not to decrease it.
5938 40. When delivering a message, the value of $return_path is set to $sender_
5939 address at the start of routing (routers may change the value). By an
5940 oversight, this default was not being set up when an address was tested by
5941 -bt or -bv, which affected the outcome if any router or filter referred to
5944 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5945 log only when "message" or "log_message" are set. However, if one of the
5946 conditions was an address verification, or a call to a nested ACL, the
5947 messages generated by the underlying test were being passed through. This
5948 no longer happens. The underlying message is available in $acl_verify_
5949 message for both "message" and "log_message" expansions, so it can be
5950 passed through if needed.
5952 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5953 new expansion $bh_ to give the encoded byte string without charset
5954 translation. Translation happens only if iconv() is available; HAVE_ICONV
5955 indicates this at build time. HEADERS_CHARSET gives the charset to
5956 translate to; headers_charset can change it in the configuration, and
5957 "headers charset" can change it in an individual filter file.
5959 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5960 that creates RFC 2047 encoded "words" labels them as that charset instead
5961 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5962 expansion operator; (ii) when Exim creates a From: line for a local
5963 message; (iii) when a header line is rewritten to include a "phrase" part.
5965 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5966 buggy, causing it to skip the first lines of messages whose message ID
5967 ended in 'D'. This would not have bitten before Exim release 4.14, because
5968 message IDs were unlikely to end in 'D' before then. The effect was to have
5969 incorrect size information for certain domains.
5971 45. #include "config.h" was missing at the start of the crypt16.c module. This
5972 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5975 46. If there was a timeout during a "random" callout check, Exim treated it as
5976 a failure of the random address, and carried on sending RSET and the real
5977 address. If the delay was just some slowness somewhere, the response to the
5978 original RCPT would be taken as a response to RSET and so on, causing
5979 mayhem of various kinds.
5981 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5982 when I implemented it. It didn't allow for the fact that some option values
5983 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5984 the right test for positive values.
5986 48. Domain names in DNS records are case-independent. Exim always looks them up
5987 in lower case. Some resolvers return domain names in exactly the case they
5988 appear in the zone file, that is, they may contain uppercase letters. Not
5989 all resolvers do this - some return always lower case. Exim was treating a
5990 change of case by a resolver as a change of domain, similar to a widening
5991 of a domain abbreviation. This triggered its re-routing code and so it was
5992 trying to route what was effectively the same domain again. This normally
5993 caused routing to fail (because the router wouldn't handle the domain
5994 twice). Now Exim checks for this case specially, and just changes the
5995 casing of the domain that it ultimately uses when it transmits the message
5998 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6001 50. If a filter generated a file delivery with a non-absolute name (possible if
6002 no home directory exists for the router), the forbid_file option was not
6005 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6006 the existing equality matching.
6008 52. Exim was using ints instead of ino_t variables in some places where it was
6009 dealing with inode numbers.
6011 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6012 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6013 called TMPDIR, and if it finds it is different, it changes its value.
6015 54. The smtp_printf() function is now made available to local_scan() so
6016 additional output lines can be written before returning. There is also an
6017 smtp_fflush() function to enable the detection of a dropped connection.
6018 The variables smtp_input and smtp_batched_input are exported to
6021 55. Changed the default runtime configuration: the message "Unknown user"
6022 has been removed from the ACL, and instead placed on the localuser router,
6023 using the cannot_route_message feature. This means that any verification
6024 failures that generate their own messages won't get overridden. Similarly,
6025 the "Unrouteable address" message that was in the ACL for unverifiable
6026 relay addresses has also been removed.
6028 56. Added hosts_avoid_esmtp to the smtp transport.
6030 57. The exicyclog script was not checking for the esoteric option
6031 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6032 will work only if exicyclog is run under the appropriate euid.
6034 58. Following a discussion on the list, the rules by which Exim recognises line
6035 endings on incoming messages have been changed. The -dropcr and drop_cr
6036 options are now no-ops, retained only for backwards compatibility. The
6037 following line terminators are recognized: LF CRLF CR. However, special
6038 processing applies to CR:
6040 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6041 nor a local message in the state where . is a terminator.
6043 (ii) If a bare CR is encountered in a header line, an extra space is added
6044 after the line terminator so as not to end the header. The reasoning
6045 behind this is that bare CRs in header lines are most likely either
6046 to be mistakes, or people trying to play silly games.
6048 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6049 was being incorrectly given as 18 bytes larger than it should have been.
6050 This is a VOB (very old bug).
6052 60. This may never have affected anything current, but just in case it has:
6053 When the local host is found other than at the start of a list of hosts,
6054 the local host, those with the same MX, and any that follow, are discarded.
6055 When the list in question was part of a longer list of hosts, the following
6056 hosts (not currently being processed) were also being discarded. This no
6057 longer happens. I'm not sure if this situation could ever has previously
6060 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6063 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6064 header. This is something that is recommended in a new Internet Draft, and
6065 is something that is documented as being done by Sendmail. There are two
6066 possible values. For messages generated by the autoreply transport, Exim
6069 Auto-Submitted: auto-replied
6071 whereas for all other generated messages (e.g. bounces) it adds
6073 Auto-Submitted: auto-generated
6075 63. The "personal" condition in filters now includes a test for the
6076 Auto-Submitted: header. If it contains the string "auto-" the message it
6077 not considered personal.
6079 64. Added rcpt_include_affixes as a generic transport option.
6081 65. Added queue_only_override (default true).
6083 66. Added the syslog_duplication option.
6085 67. If what should have been the first header line of a message consisted of
6086 a space followed by a colon, Exim was mis-interpreting it as a header line.
6087 It isn't of course - it is syntactically invalid and should therefore be
6088 treated as the start of the message body. The misbehaviour could have
6089 caused a number of strange effects, including loss of data in subsequent
6090 header lines, and spool format errors.
6092 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6093 client host had authenticated. This control can now be exercised by an ACL
6094 for more flexibility.
6096 69. By default, callouts do not happen when testing with -bh. There is now a
6097 variant, -bhc, which does actually run the callout code, including
6098 consulting and updating the callout cache.
6100 70. Added support for saslauthd authentication, courtesy of Alexander
6103 71. If statvfs() failed on the spool or log directories while checking their
6104 size for availability, Exim confusingly gave the error "space shortage".
6105 Furthermore, in debugging mode it crashed with a floating point exception.
6106 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6107 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6108 serious problem, Exim now writes to the main and panic logs when this
6109 happens, with details of the failure. It then refuses to accept the
6110 incoming message, giving the message "spool directory problem" or "log
6111 directory problem" with a 421 code for SMTP messages.
6113 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6114 0, 1, and 2 exist, because some OS complain for execs without them (see
6115 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6116 descriptors. However, the code omitted to check that the open succeeded,
6117 causing mysterious errors if for some reason the permissions on /dev/null
6118 got screwed. Now Exim writes a message to the main and panic logs, and
6119 bombs out if it can't open /dev/null.
6121 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6122 interact so that it is all more flexible. It is supposed to remain
6123 backwards compatible. Also added extra_local_interfaces.
6125 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6126 to bomb out with an assertion failure - to the client this appears as a
6127 connection drop. This problem occurs in the part of the code that was taken
6128 from the Samba project. Fortunately, the assertion is in a very simple
6129 function, so I have fixed this by reproducing the function inline in the
6130 one place where it is called, and arranging for authentication to fail
6131 instead of killing the process with assert().
6133 75. The SPA client code was not working when the server requested OEM rather
6134 than Unicode encoding.
6136 76. Added code to make require_files with a specific uid setting more usable in
6137 the case where statting the file as root fails - usually a non-root-mounted
6138 NFS file system. When this happens and the failure is EACCES, Exim now
6139 forks a subprocess and does the per-uid checking as the relevant uid.
6141 77. Added process_log_path.
6143 78. If log_file_path was not explicitly set, a setting of check_log_space or
6144 check_log_inodes was ignored.
6146 79. If a space check for the spool or log partitions fails, the incident is now
6147 logged. Of course, in the latter case the data may get lost...
6149 80. Added the %p formatting code to string_format() so that it can be used to
6150 print addresses in debug_print(). Adjusted all the address printing in the
6151 debugging in store.c to use %p rather than %d.
6153 81. There was a concern that a line of code in smtp_in.c could overflow a
6154 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6155 initially expressed, the concern was not well-founded, because trailing
6156 spaces are removed early. However, if the trailing spaces were followed by
6157 a NULL, they did not get removed, so the overflow was possible. Two fixes
6160 (a) I re-wrote the offending code in a cleaner fashion.
6161 (b) If an incoming SMTP command contains a NULL character, it is rejected
6164 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6165 runs initgroups(), so that if the Exim user is in any additional groups,
6166 they will be used during message reception.
6172 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob