1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
23 TL/01 When running the test suite, the README says that variables such as
24 no_msglog_check are global and can be placed anywhere in a specific
25 test's script, however it was observed that placement needed to be near
26 the beginning for it to behave that way. Changed the runtest perl
27 script to read through the entire script once to detect and set these
28 variables, reset to the beginning of the script, and then run through
29 the script parsing/test process like normal.
31 TL/02 The BSD's have an arc4random API. One of the functions to induce
32 adding randomness was arc4random_stir(), but it has been removed in
33 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
34 function when detected.
36 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
37 cause callback expansion.
39 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
40 syntax errors in an expansion can be treated as a string instead of
41 logging or causing an error, due to the internal use of bool_lax
42 instead of bool when processing it.
44 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
45 server certificates when making smtp deliveries.
47 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
49 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
51 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
53 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
54 Merged patch from Sebastian Wiedenroth.
56 JH/05 Fix results-pipe from transport process. Several recipients, combined
57 with certificate use, exposed issues where response data items split
58 over buffer boundaries were not parsed properly. This eventually
59 resulted in duplicates being sent. This issue only became common enough
60 to notice due to the introduction of conection certificate information,
61 the item size being so much larger. Found and fixed by Wolfgang Breyha.
63 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
64 size buffer was used, resulting in syntax errors when an expansion
67 JH/07 Add support for directories of certificates when compiled with a GnuTLS
68 version 3.3.6 or later.
70 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
71 is EXPERIMENTAL_EVENT, the main-configuration and transport options
72 both become "event_action", the variables become $event_name, $event_data
73 and $event_defer_errno. There is a new variable $verify_mode, usable in
74 routers, transports and related events. The tls:cert event is now also
75 raised for inbound connections, if the main configuration event_action
78 TL/06 In test suite, disable OCSP for old versions of openssl which contained
79 early OCSP support, but no stapling (appears to be less than 1.0.0).
81 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
82 server certificate names available under the smtp transport option
83 "tls_verify_cert_hostname" now do not permit multi-component wildcard
86 JH/10 Time-related extraction expansions from certificates now use the main
87 option "timezone" setting for output formatting, and are consistent
88 between OpenSSL and GnuTLS compilations. Bug 1541.
90 JH/11 Fix a crash in mime ACL when meeting a zero-length parameter in the
93 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
94 include certificate info, eximon was claiming there were spoolfile
97 JH/13 Buf 1521: Fix ldap lookup for single-attr request, multiple-attr return.
99 JH/14 Log delivery-related information more consistently, using the sequence
100 "H=<name> [<ip>]" wherever possible.
105 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
106 checkers that were complaining about end of non-void function with no
109 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
110 This was a regression intruduced in 4.83 by another bugfix.
112 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
114 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
115 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
121 TF/01 Correctly close the server side of TLS when forking for delivery.
123 When a message was received over SMTP with TLS, Exim failed to clear up
124 the incoming connection properly after forking off the child process to
125 deliver the message. In some situations the subsequent outgoing
126 delivery connection happened to have the same fd number as the incoming
127 connection previously had. Exim would try to use TLS and fail, logging
128 a "Bad file descriptor" error.
130 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
131 utilities have not been installed.
133 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
134 temporary space as the ACL may create new global variables.
136 TL/01 LDAP support uses per connection or global context settings, depending
137 upon the detected version of the libraries at build time.
139 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
140 to extract and use the src ip:port in logging and expansions as if it
141 were a direct connection from the outside internet. PPv2 support was
142 updated based on HAProxy spec change in May 2014.
144 JH/02 Add ${listextract {number}{list}{success}{fail}}.
146 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
147 Properly escape header and check for NULL return.
149 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
152 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
154 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
155 characters in header names, implemented as a verify condition.
156 Contributed by Michael Fischer v. Mollard.
158 TL/05 Rename SPF condition results err_perm and err_temp to standardized
159 results permerror and temperror. Previous values are deprecated but
160 still accepted. In a future release, err_perm and err_temp will be
161 completely removed, which will be a backward incompatibility if the
162 ACL tests for either of these two old results. Patch contributed by
163 user bes-internal on the mailing list.
165 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
167 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
168 selectors, in both main and reject logs.
170 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
173 JH/07 Add malware type "sock" for talking to simple daemon.
175 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
177 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
178 routers/transports under cutthrough routing.
180 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
181 numbers. Touch up "bool" conditional to keep the same definition.
183 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
185 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
187 JH/12 Expand items in router/transport headers_add or headers_remove lists
188 individually rather than the list as a whole. Bug 1452.
190 Required for reasonable handling of multiple headers_ options when
191 they may be empty; requires that headers_remove items with embedded
192 colons must have them doubled (or the list-separator changed).
194 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
195 view the policy declared in the DMARC record. Currently, $dmarc_status
196 is a combined value of both the record presence and the result of the
199 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
201 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
202 dnslookup router and the smtp transport (applying to the forward
205 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
206 of ldap servers used for a specific lookup. Patch provided by Heiko
209 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
210 New variable $lookup_dnssec_authenticated for observability.
212 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
213 Patch submitted by Lars Timman.
215 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
217 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
218 Requires trusted mode and valid format message id, aborts otherwise.
219 Patch contributed by Heiko Schlichting.
221 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
222 certextract with support for various fields. Bug 1358.
224 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
225 is requested by default, modifiable by smtp transport option
228 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
229 operate on certificate variables to give certificate fingerprints
230 Also new ${sha256:cert_variable}.
232 JH/23 The PRDR feature is moved from being Experimental into the mainline.
234 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
235 Christian Aistleitner.
237 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
239 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
240 file. Patch from Wolfgang Breyha.
242 JH/25 Expand the coverage of the delivery $host and $host_address to
243 client authenticators run in verify callout. Bug 1476.
245 JH/26 Port service names are now accepted for tls_on_connect_ports, to
246 align with daemon_smtp_ports. Bug 72.
248 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
249 support and error reporting did not work properly.
251 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
252 and is readable. Patch from Andrew Colin Kissa.
254 TL/14 Enhance documentation of ${run expansion and how it parses the
255 commandline after expansion, particularly in the case when an
256 unquoted variable expansion results in an empty value.
258 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
260 PP/02 Fix internal collision of T_APL on systems which support RFC3123
261 by renaming away from it. Addresses GH issue 15, reported by
264 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
266 TL/15 SECURITY: prevent double expansion in math comparison functions
267 (can expand unsanitized data). Not remotely exploitable.
274 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
276 PP/02 Make -n do something, by making it not do something.
277 When combined with -bP, the name of an option is not output.
279 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
282 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
283 $sender_host_name and config options to manage this, and basic check
286 PP/05 DSCP support for outbound connections and control modifier for inbound.
288 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
289 (Only plugin which currently uses this is kerberos4, which nobody should
290 be using, but we should make it available and other future plugins might
291 conceivably use it, even though it would break NAT; stuff *should* be
292 using channel bindings instead).
294 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
295 name; added for Sendmail compatibility; requires admin caller.
296 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
297 just ignore it); requires trusted caller.
298 Also parse but ignore: -Ac -Am -X<logfile>
301 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
303 TL/02 Add +smtp_confirmation as a default logging option.
305 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
306 Patch by Magnus Holmgren from 2007-02-20.
308 TL/04 Bugzilla 1281 - Spec typo.
309 Bugzilla 1283 - Spec typo.
310 Bugzilla 1290 - Spec grammar fixes.
312 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
314 TL/06 Add Experimental DMARC support using libopendmarc libraries.
316 TL/07 Fix an out of order global option causing a segfault. Reported to dev
317 mailing list by by Dmitry Isaikin.
319 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
321 JH/02 Support "G" suffix to numbers in ${if comparisons.
323 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
325 NM/01 Bugzilla 1197 - Spec typo
326 Bugzilla 1196 - Spec examples corrections
328 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
330 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
331 gnutls_enable_pkcs11, but renamed to more accurately indicate its
334 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
335 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
337 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
338 "acl {{name}{arg}...}", and optional args on acl condition
341 JH/05 Permit multiple router/transport headers_add/remove lines.
343 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
345 JH/07 Avoid using a waiting database for a single-message-only transport.
346 Performance patch from Paul Fisher. Bugzilla 1262.
348 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
351 JH/09 Add $headers_added variable, with content from use of ACL modifier
352 add_header (but not yet added to the message). Bugzilla 199.
354 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
355 Pulled from Bugzilla 817 by Wolfgang Breyha.
357 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
359 (nb: this is the same fix as in Exim 4.80.1)
361 JH/11 Add A= logging on delivery lines, and a client_set_id option on
364 JH/12 Add optional authenticated_sender logging to A= and a log_selector
367 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
369 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
370 advertise SMTP AUTH mechanism to us, instead of a generic
371 protocol violation error. Also, make Exim more robust to bad
372 data from the Dovecot auth socket.
374 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
376 When a queue runner is handling a message, Exim first routes the
377 recipient addresses, during which it prunes them based on the retry
378 hints database. After that it attempts to deliver the message to
379 any remaining recipients. It then updates the hints database using
382 So if a recipient address works intermittently, it can get repeatedly
383 deferred at routing time. The retry hints record remains fresh so the
384 address never reaches the final cutoff time.
386 This is a fairly common occurrence when a user is bumping up against
387 their storage quota. Exim had some logic in its local delivery code
388 to deal with this. However it did not apply to per-recipient defers
389 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
391 This change adds a proper retry rule check during routing so that the
392 final cutoff time is checked against the message's age. We only do
393 this check if there is an address retry record and there is not a
394 domain retry record; this implies that previous attempts to handle
395 the address had the retry_use_local_parts option turned on. We use
396 this as an approximation for the destination being like a local
397 delivery, as in LMTP.
399 I suspect this new check makes the old local delivery cutoff check
400 redundant, but I have not verified this so I left the code in place.
402 TF/02 Correct gecos expansion when From: is a prefix of the username.
404 Test 0254 submits a message to Exim with the header
408 When I ran the test suite under the user fanf2, Exim expanded
409 the header to contain my full name, whereas it should have added
410 a Resent-Sender: header. It erroneously treats any prefix of the
411 username as equal to the username.
413 This change corrects that bug.
415 GF/01 DCC debug and logging tidyup
416 Error conditions log to paniclog rather than rejectlog.
417 Debug lines prefixed by "DCC: " to remove any ambiguity.
419 TF/03 Avoid unnecessary rebuilds of lookup-related code.
421 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
422 Bug spotted by Jeremy Harris; was flawed since initial commit.
423 Would have resulted in OCSP responses post-SNI triggering an Exim
424 NULL dereference and crash.
426 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
428 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
429 Bug detection, analysis and fix by Samuel Thibault.
430 Bugzilla 1331, Debian bug #698092.
432 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
434 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
435 Server implementation by Todd Lyons, client by JH.
436 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
437 config variable "prdr_enable" controls whether the server
438 advertises the facility. If the client requests PRDR a new
439 acl_data_smtp_prdr ACL is called once for each recipient, after
440 the body content is received and before the acl_smtp_data ACL.
441 The client is controlled by bolth of: a hosts_try_prdr option
442 on the smtp transport, and the server advertisement.
443 Default client logging of deliveries and rejections involving
444 PRDR are flagged with the string "PRDR".
446 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
447 fclose(). Diagnosis by Todd Lyons.
449 PP/17 Update configure.default to handle IPv6 localhost better.
450 Patch by Alain Williams (plus minor tweaks).
453 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
454 This is now consistent with GnuTLS, and is now documented: the
455 previous undocumented portable approach to treating the option as
456 unset was to force an expansion failure. That still works, and
457 an empty string is now equivalent.
459 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
460 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
461 not performing validation itself.
463 PP/20 Added force_command boolean option to pipe transport.
464 Patch from Nick Koston, of cPanel Inc.
466 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
469 TF/04 Added udpsend ACL modifer and hexquote expansion operator
471 PP/21 Fix eximon continuous updating with timestamped log-files.
472 Broken in a format-string cleanup in 4.80, missed when I repaired the
473 other false fix of the same issue.
474 Report and fix from Heiko Schlichting.
477 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
478 Report from Prashanth Katuri.
480 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
481 It's SecureTransport, so affects any MacOS clients which use the
482 system-integrated TLS libraries, including email clients.
484 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
485 using a MIME ACL for non-SMTP local injection.
486 Report and assistance in diagnosis by Warren Baker.
488 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
490 JH/16 Fix comparisons for 64b. Bugzilla 1385.
492 TL/09 Add expansion variable $authenticated_fail_id to keep track of
493 last id that failed so it may be referenced in subsequent ACL's.
495 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
498 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
499 ldap library initialization, allowing self-signed CA's to be
500 used. Also properly sets require_cert option later in code by
501 using NULL (global ldap config) instead of ldap handle (per
502 session). Bug diagnosis and testing by alxgomz.
504 TL/12 Enhanced documentation in the ratelimit.pl script provided in
505 the src/util/ subdirectory.
507 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
508 renamed to Transport Post Delivery Action by Jeremy Harris, as
511 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
512 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
513 redis_servers = needs to be configured which will be used by the redis
514 lookup. Patch from Warren Baker, of The Packet Hub.
516 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
518 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
519 hostname or reverse DNS when processing a host list. Used suggestions
520 from multiple comments on this bug.
522 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
524 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
525 Missed a few lines, added it to make the runtest require no keyboard
528 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
529 contains upper case chars. Make router use caseful_local_part.
531 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
532 support when GnuTLS has been built with p11-kit.
538 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
540 This, or similar/improved, will also be change PP/11 of 4.82.
546 PP/01 Handle short writes when writing local log-files.
547 In practice, only affects FreeBSD (8 onwards).
548 Bugzilla 1053, with thanks to Dmitry Isaikin.
550 NM/01 Bugzilla 949 - Documentation tweak
552 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
555 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
557 PP/02 Implemented gsasl authenticator.
559 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
561 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
562 `pkg-config foo` for cflags/libs.
564 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
565 with rest of GSASL and with heimdal_gssapi.
567 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
568 `pkg-config foo` for cflags/libs for the TLS implementation.
570 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
571 properties get this fed in as external SSF. A number of robustness
572 and debugging improvements to the cyrus_sasl authenticator.
574 PP/08 cyrus_sasl server now expands the server_realm option.
576 PP/09 Bugzilla 1214 - Log authentication information in reject log.
577 Patch by Jeremy Harris.
579 PP/10 Added dbmjz lookup type.
581 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
583 PP/12 MAIL args handles TAB as well as SP, for better interop with
584 non-compliant senders.
585 Analysis and variant patch by Todd Lyons.
587 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
588 Bug report from Lars Müller <lars@samba.org> (via SUSE),
589 Patch from Dirk Mueller <dmueller@suse.com>
591 PP/13 tls_peerdn now print-escaped for spool files.
592 Observed some $tls_peerdn in wild which contained \n, which resulted
593 in spool file corruption.
595 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
596 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
597 or write after TLS renegotiation, which otherwise led to messages
600 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
601 as a tracking header (ie: a signed header comes before the signature).
602 Patch from Wolfgang Breyha.
604 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
605 comma-sep list; embedded commas doubled.
607 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
609 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
611 Report and patch from Dmitry Banschikov.
613 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
614 Removed SSL_clear() after SSL_new() which led to protocol negotiation
615 failures. We appear to now support TLS1.1+ with Exim.
617 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
618 lets Exim select keys and certificates based upon TLS SNI from client.
619 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
620 before an outbound SMTP session. New log_selector, +tls_sni.
622 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
623 NULL dereference. Report and patch from Alun Jones.
625 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
626 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
627 Not seeing resolver debug output on NetBSD, but suspect this is a
628 resolver implementation change.
630 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
631 Left warnings. Added "eximon gdb" invocation mode.
633 PP/21 Defaulting "accept_8bitmime" to true, not false.
635 PP/22 Added -bw for inetd wait mode support.
637 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
638 locate the relevant includes and libraries. Made this the default.
640 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
641 Bugzilla 1246, report and most of solution from Tomasz Kusy.
643 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
644 This may cause build issues on older platforms.
646 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
647 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
648 gnutls_require_mac & gnutls_require_protocols (no longer supported).
649 Added SNI support via GnuTLS too.
650 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
652 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
654 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
655 Applied second patch from Janne, implementing suggestion to default
656 multiple-strings-in-record handling to match SPF spec.
658 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
660 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
661 read-only, out of scope).
662 Patch from Wolfgang Breyha, report from Stuart Northfield.
664 PP/29 Fix three issues highlighted by clang analyser static analysis.
665 Only crash-plausible issue would require the Cambridge-specific
666 iplookup router and a misconfiguration.
667 Report from Marcin Mirosław.
669 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
671 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
672 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
673 As part of this, removing so much warning spew let me fix some minor
674 real issues in debug logging.
676 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
677 assignment on my part. Fixed.
679 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
680 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
681 Janne Snabb (who went above and beyond: thank you).
683 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
684 string otherwise requires a connection and a bunch more work and it's
685 relatively easy to get wrong. Should also expose TLS library linkage
688 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
689 64-bit ${eval} (JH/03).
691 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
692 GNU libc to support some of the 64-bit stuff, should not lead to
693 conflicts. Defined before os.h is pulled in, so if a given platform
694 needs to override this, it can.
696 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
697 protection layer was required, which is not implemented.
698 Bugzilla 1254, patch from Wolfgang Breyha.
700 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
701 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
702 tls_dhparam take prime identifiers. Also unbreak combination of
703 OpenSSL+DH_params+TLSSNI.
705 PP/39 Disable SSLv2 by default in OpenSSL support.
711 PP/01 Solaris build fix for Oracle's LDAP libraries.
712 Bugzilla 1109, patch from Stephen Usher.
714 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
716 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
719 TF/02 Fix a couple more cases where we did not log the error message
720 when unlink() failed. See also change 4.74-TF/03.
722 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
723 lock up or crash if it happened to be inside a call to libc when it
724 got a SIGUSR1 from exiwhat.
726 The SIGUSR1 handler appends the current process status to the process
727 log which is later printed by exiwhat. It used to use the general
728 purpose logging code to do this, but several functions it calls are
729 not safe for signals.
731 The new output code in the SIGUSR1 handler is specific to the process
732 log, and simple enough that it's easy to inspect for signal safety.
733 Removing some special cases also simplifies the general logging code.
734 Removing the spurious timestamps from the process log simplifies
737 TF/04 Improved ratelimit ACL condition.
739 The /noupdate option has been deprecated in favour of /readonly which
740 has clearer semantics. The /leaky, /strict, and /readonly update modes
741 are mutually exclusive. The update mode is no longer included in the
742 database key; it just determines when the database is updated. (This
743 means that when you upgrde Exim will forget old rate measurements.)
745 Exim now checks that the per_* options are used with an update mode that
746 makes sense for the current ACL. For example, when Exim is processing a
747 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
748 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
749 must specify per_mail/readonly. If you omit the update mode it defaults to
750 /leaky where that makes sense (as before) or /readonly where required.
752 The /noupdate option is now undocumented but still supported for
753 backwards compatibility. It is equivalent to /readonly except that in
754 ACLs where /readonly is required you may specify /leaky/noupdate or
755 /strict/noupdate which are treated the same as /readonly.
757 A useful new feature is the /count= option. This is a generalization
758 of the per_byte option, so that you can measure the throughput of other
759 aggregate values. For example, the per_byte option is now equivalent
760 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
762 The per_rcpt option has been generalized using the /count= mechanism
763 (though it's more complicated than the per_byte equivalence). When it is
764 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
765 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
766 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
767 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
768 Note that using per_rcpt with a non-readonly update mode in more than
769 one ACL will cause the recipients to be double-counted. (The per_mail
770 and per_byte options don't have this problem.)
772 The handling of very low rates has changed slightly. If the computed rate
773 is less than the event's count (usually one) then this event is the first
774 after a long gap. In this case the rate is set to the same as this event's
775 count, so that the first message of a spam run is counted properly.
777 The major new feature is a mechanism for counting the rate of unique
778 events. The new per_addr option counts the number of different
779 recipients that someone has sent messages to in the last time period. It
780 behaves like per_rcpt if all the recipient addresses are different, but
781 duplicate recipient addresses do not increase the measured rate. Like
782 the /count= option this is a general mechanism, so the per_addr option
783 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
784 example, measure the rate that a client uses different sender addresses
785 with the options per_mail/unique=$sender_address. There are further
786 details in the main documentation.
788 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
790 TF/06 Removed a few PCRE remnants.
792 TF/07 Automatically extract Exim's version number from tags in the git
793 repository when doing development or release builds.
795 PP/02 Raise smtp_cmd_buffer_size to 16kB.
796 Bugzilla 879. Patch from Paul Fisher.
798 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
799 Heavily based on revision 40f9a89a from Simon Arlott's tree.
802 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
804 PP/05 Variable $av_failed, true if the AV scanner deferred.
805 Bugzilla 1078. Patch from John Horne.
807 PP/06 Stop make process more reliably on build failure.
808 Bugzilla 1087. Patch from Heiko Schlittermann.
810 PP/07 Make maildir_use_size_file an _expandable_ boolean.
811 Bugzilla 1089. Patch from Heiko Schlittermann.
813 PP/08 Handle ${run} returning more data than OS pipe buffer size.
814 Bugzilla 1131. Patch from Holger Weiß.
816 PP/09 Handle IPv6 addresses with SPF.
817 Bugzilla 860. Patch from Wolfgang Breyha.
819 PP/10 GnuTLS: support TLS 1.2 & 1.1.
821 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
824 PP/11 match_* no longer expand right-hand-side by default.
825 New compile-time build option, EXPAND_LISTMATCH_RHS.
826 New expansion conditions, "inlist", "inlisti".
828 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
830 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
832 PP/14 fix log_write() format string regression from TF/03.
833 Bugzilla 1152. Patch from Dmitry Isaikin.
839 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
841 PP/02 Harmonised TLS library version reporting; only show if debugging.
842 Layout now matches that introduced for other libraries in 4.74 PP/03.
844 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
846 PP/04 New "dns_use_edns0" global option.
848 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
851 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
852 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
854 TK/01 Updated PolarSSL code to 0.14.2.
855 Bugzilla 1097. Patch from Andreas Metzler.
857 PP/07 Catch divide-by-zero in ${eval:...}.
860 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
863 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
864 format-string attack -- SECURITY: remote arbitrary code execution.
866 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
867 time unintentionally subject to list matching rules, letting the header
868 cause arbitrary Exim lookups (of items which can occur in lists, *not*
869 arbitrary string expansion). This allowed for information disclosure.
871 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
872 INT_MIN/-1 -- value coerced to INT_MAX.
878 NM/01 Workround for PCRE version dependency in version reporting
881 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
882 This fixes portability to compilers other than gcc, notably
883 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
885 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
886 makefiles for portability to HP-UX and POSIX correctness.
888 PP/01 Permit LOOKUP_foo enabling on the make command-line.
889 Also via indented variable definition in the Makefile.
890 (Debugging by Oliver Heesakkers).
892 PP/02 Restore caching of spamd results with expanded spamd_address.
893 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
895 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
896 Improves build reliability. Fix from: Frank Elsner
898 NM/02 Fix wide character breakage in the rfc2047 coding
899 Fixes bug 1064. Patch from Andrey N. Oktyabrski
901 NM/03 Allow underscore in dnslist lookups
902 Fixes bug 1026. Patch from Graeme Fowler
904 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
905 Code patches from Adam Ciarcinski of NetBSD.
907 NM/04 Fixed exiqgrep to cope with mailq missing size issue
910 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
911 is logged, to avoid truncation. Patch from John Horne.
913 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
914 Patch from Jakob Hirsch.
916 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
917 SQL string expansion failure details.
918 Patch from Andrey Oktyabrski.
920 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
921 Patch from Simon Arlott.
923 PP/09 New lookups functionality failed to compile on old gcc which rejects
924 extern declarations in function scope.
925 Patch from Oliver Fleischmann
927 PP/10 Use sig_atomic_t for flags set from signal handlers.
928 Check getgroups() return and improve debugging.
929 Fixed developed for diagnosis in bug 927 (which turned out to be
932 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
933 Patch from Mark Zealey.
935 PP/12 Bugzilla 1056: Improved spamd server selection.
936 Patch from Mark Zealey.
938 PP/13 Bugzilla 1086: Deal with maildir quota file races.
939 Based on patch from Heiko Schlittermann.
941 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
942 Patch from Uwe Doering, sign-off by Michael Haardt.
944 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
945 variable declaration deep within a block. Bug and patch from
948 PP/15 lookups-Makefile IRIX compatibilty coercion.
950 PP/16 Make DISABLE_DKIM build knob functional.
952 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
953 Patch by Simon Arlott
955 TF/03 Fix valgrind.h portability to C89 compilers that do not support
956 variable argument macros. Our copy now differs from upstream.
962 TF/01 Failure to get a lock on a hints database can have serious
963 consequences so log it to the panic log.
965 TF/02 Log LMTP confirmation messages in the same way as SMTP,
966 controlled using the smtp_confirmation log selector.
968 TF/03 Include the error message when we fail to unlink a spool file.
970 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
971 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
972 for maintaining out-of-tree patches for some time.
974 PP/01 Bugzilla 139: Documentation and portability issues.
975 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
976 Handle per-OS dynamic-module compilation flags.
978 PP/02 Let /dev/null have normal permissions.
979 The 4.73 fixes were a little too stringent and complained about the
980 permissions on /dev/null. Exempt it from some checks.
981 Reported by Andreas M. Kirchwitz.
983 PP/03 Report version information for many libraries, including
984 Exim version information for dynamically loaded libraries. Created
985 version.h, now support a version extension string for distributors
986 who patch heavily. Dynamic module ABI change.
988 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
989 privilege escalation vulnerability whereby the Exim run-time user
990 can cause root to append content of the attacker's choosing to
993 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
996 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
997 If dropping privileges for untrusted macros, we disabled normal logging
998 on the basis that it would fail; for the Exim run-time user, this is not
999 the case, and it resulted in successful deliveries going unlogged.
1000 Fixed. Reported by Andreas Metzler.
1006 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1007 only prepend for the Resent-* case. Fixes regression introduced in
1008 Exim 4.70 by NM/22 for Bugzilla 607.
1010 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1011 increasing numbers of administrators be bitten by this.
1013 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1015 PP/03 Bugzilla 994: added openssl_options main configuration option.
1017 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1019 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1021 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1023 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1024 without a peer certificate, leading to a segfault because of an
1025 assumption that peers always have certificates. Be a little more
1026 paranoid. Problem reported by Martin Tscholak.
1028 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1029 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1030 NB: ClamAV planning to remove STREAM in "middle of 2010".
1031 CL also introduces -bmalware, various -d+acl logging additions and
1032 more caution in buffer sizes.
1034 PP/09 Implemented reverse_ip expansion operator.
1036 PP/10 Bugzilla 937: provide a "debug" ACL control.
1038 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1040 PP/12 Bugzilla 973: Implement --version.
1042 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1044 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1046 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1048 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1049 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1050 ignore trailing whitespace.
1052 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1054 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1057 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1058 Notification from Dr Andrew Aitchison.
1060 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1061 ExtendedDetectionInfo response format.
1062 Notification from John Horne.
1064 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1067 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1068 XSL and documented dependency on system catalogs, with examples of how
1071 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1074 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1075 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1076 configuration file which is writeable by the Exim user or group.
1078 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1079 of configuration files to cover files specified with the -C option if
1080 they are going to be used with root privileges, not just the default
1083 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1084 option (effectively making it always true).
1086 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1087 files to be used while preserving root privileges.
1089 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1090 that rogue child processes cannot use them.
1092 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1093 run-time user, instead of root.
1095 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1096 Exim run-time user without dropping privileges.
1098 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1099 result string, instead of calling string_vformat() twice with the same
1102 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1103 for other users. Others should always drop root privileges if they use
1104 -C on the command line, even for a whitelisted configure file.
1106 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1108 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1114 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1115 $data_path, and $header_path variables; fixed documentation bugs and
1118 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1119 exipick to access non-standard spools, including the "frozen" queue
1122 NM/01 Bugzilla 965: Support mysql stored procedures.
1123 Patch from Alain Williams
1125 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1127 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1128 Patch from Andreas Metzler
1130 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1131 Patch from Kirill Miazine
1133 NM/05 Bugzilla 671: Added umask to procmail example.
1135 JJ/03 installed exipick 20100323.0, fixing doc bug
1137 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1138 directory. Notification and patch from Dan Rosenberg.
1140 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1142 TK/02 Improve log output when DKIM signing operation fails.
1144 MH/01 Treat the transport option dkim_domain as a colon separated
1145 list, not as a single string, and sign the message with each element,
1146 omitting multiple occurences of the same signer.
1148 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1149 Bugzilla 985, 986. Patch by Simon Arlott
1151 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1152 Patch by Simon Arlott
1154 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1155 MBX locking. Notification from Dan Rosenberg.
1161 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1163 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1165 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1167 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1169 NM/04 Bugzilla 915: Flag broken perl installation during build.
1175 TK/01 Added patch by Johannes Berg that expands the main option
1176 "spamd_address" if it starts with a dollar sign.
1178 TK/02 Write list of recipients to X-Envelope-Sender header when building
1179 the mbox-format spool file for content scanning (suggested by Jakob
1182 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1183 (http://www.dcc-servers.net/) support via dccifd. Activated by
1184 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1186 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1187 by Mark Daniel Reidel <mr@df.eu>.
1189 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1190 When building exim an external PCRE library is now needed -
1191 PCRE is a system library on the majority of modern systems.
1192 See entry on PCRE_LIBS in EDITME file.
1194 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1195 conversation. Added nologin parameter to request.
1196 Patch contributed by Kirill Miazine.
1198 TF/01 Do not log submission mode rewrites if they do not change the address.
1200 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1202 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1203 log files in place. Contributed by Roberto Lima.
1205 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1207 TF/03 Bugzilla 615: When checking the local_parts router precondition
1208 after a local_part_suffix or local_part_prefix option, Exim now
1209 does not use the address's named list lookup cache, since this
1210 contains cached lookups for the whole local part.
1212 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1213 Robert Millan. Documentation is in experimental-spec.txt.
1215 TF/04 Bugzilla 668: Fix parallel build (make -j).
1217 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1219 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1220 Patch provided by Jan Srzednicki.
1222 TF/05 Leading white space used to be stripped from $spam_report which
1223 wrecked the formatting. Now it is preserved.
1225 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1226 that they are available at delivery time.
1228 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1230 TF/08 TLS error reporting now respects the incoming_interface and
1231 incoming_port log selectors.
1233 TF/09 Produce a more useful error message if an SMTP transport's hosts
1234 setting expands to an empty string.
1236 NM/06 Bugzilla 744: EXPN did not work under TLS.
1237 Patch provided by Phil Pennock.
1239 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1240 Patch provided by Richard Godbee.
1242 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1243 acl_smtp_notquit, added index entry.
1245 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1246 Patch provided by Eugene Bujak.
1248 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1249 accept(). Patch provided by Maxim Dounin.
1251 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1252 Patch provided by Phil Pennock.
1254 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1256 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1257 Patch provided by Brad "anomie" Jorsch.
1259 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1260 Patch provided by Dean Brooks.
1262 TK/05 Add native DKIM support (does not depend on external libraries).
1264 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1265 Patch provided by Graeme Fowler.
1267 NM/16 Bugzilla 851: Documentation example syntax fix.
1269 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1271 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1274 NM/19 Bugzilla 745: TLS version reporting.
1275 Patch provided by Phil Pennock.
1277 NM/20 Bugzilla 167: bool: condition support.
1278 Patch provided by Phil Pennock.
1280 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1281 clients. Patch provided by Phil Pennock.
1283 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1284 Patch provided by Brad "anomie" Jorsch.
1286 NM/23 Bugzilla 687: Fix misparses in eximstats.
1287 Patch provided by Heiko Schlittermann.
1289 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1290 Patch provided by Heiko Schlittermann.
1292 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1293 plus update to original patch.
1295 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1297 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1298 Patch provided by David Brownlee.
1300 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1302 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1304 NM/30 Bugzilla 888: TLS documentation bugfixes.
1306 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1308 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1309 Unlike the original bugzilla I have changed all shell scripts in src tree.
1311 NM/33 Bugzilla 898: Transport filter timeout fix.
1312 Patch by Todd Rinaldo.
1314 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1315 Patch by Serge Demonchaux.
1317 NM/35 Bugzilla 39: Base64 decode bug fixes.
1318 Patch by Jakob Hirsch.
1320 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1322 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1324 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1326 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1332 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1333 ALT-N's libdkim that I have put here:
1334 http://duncanthrax.net/exim-experimental/
1336 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1337 were called 'true' and it seems that C99 defines that as a reserved
1338 keyword to be used with 'bool' variable types. That means you could
1339 not include C99-style headers which use bools without triggering
1340 build errors in sieve.c.
1342 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1343 as mailq or other aliases. Changed the --help handling significantly
1344 to do whats expected. exim_usage() emits usage/help information.
1346 SC/01 Added the -bylocaldomain option to eximstats.
1348 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1350 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1352 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1358 PH/01 Another patch from the Sieve maintainer.
1360 PH/02 When an IPv6 address is converted to a string for single-key lookup
1361 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1362 dots are used instead of colons so that keys in lsearch files need not
1363 contain colons. This was done some time before quoting was made available
1364 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1365 (notated using the quote facility) so as to distinguish them from IPv4
1366 keys. This meant that lookups for IP addresses in host lists did not work
1367 for iplsearch lookups.
1369 This has been fixed by arranging for IPv6 addresses to be expressed with
1370 colons if the lookup type is iplsearch. This is not incompatible, because
1371 previously such lookups could never work.
1373 The situation is now rather anomolous, since one *can* have colons in
1374 ordinary lsearch keys. However, making the change in all cases is
1375 incompatible and would probably break a number of configurations.
1377 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1380 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1381 conversion specification without a maximum field width, thereby enabling
1382 a rogue spamd server to cause a buffer overflow. While nobody in their
1383 right mind would setup Exim to query an untrusted spamd server, an
1384 attacker that gains access to a server running spamd could potentially
1385 exploit this vulnerability to run arbitrary code as the Exim user.
1387 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1388 $primary_hostname instead of what libspf2 thinks the hosts name is.
1390 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1391 a directory entry by the name of the lookup key. Previously, if a
1392 symlink pointed to a non-existing file or a file in a directory that
1393 Exim lacked permissions to read, a lookup for a key matching that
1394 symlink would fail. Now it is enough that a matching directory entry
1395 exists, symlink or not. (Bugzilla 503.)
1397 PH/03 The body_linecount and body_zerocount variables are now exported in the
1400 PH/04 Added the $dnslist_matched variable.
1402 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1403 This means they are set thereafter only if the connection becomes
1406 PH/06 Added the client_condition to authenticators so that some can be skipped
1407 by clients under certain conditions.
1409 PH/07 The error message for a badly-placed control=no_multiline_responses left
1410 "_responses" off the end of the name.
1412 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1414 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1415 (without spaces) instead of just copying the configuration text.
1417 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1419 PH/11 Added $max_received_linelength.
1421 PH/12 Added +ignore_defer and +include_defer to host lists.
1423 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1424 way in which PCRE > 7.0 is built.
1426 PH/14 Implemented queue_only_load_latch.
1428 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1429 MAIL command. The effect was to mangle the value on 64-bit systems.
1431 PH/16 Another patch from the Sieve maintainer.
1433 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1435 PH/18 If a system quota error occurred while trying to create the file for
1436 a maildir delivery, the message "Mailbox is full" was not appended to the
1437 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1438 only to a quota excession during the actual writing of the file.
1440 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1441 characters?) which causes problems in log lines. The DN values are now
1442 passed through string_printing() before being added to log lines.
1444 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1445 and InterBase are left for another time.)
1447 PH/21 Added message_body_newlines option.
1449 PH/22 Guard against possible overflow in moan_check_errorcopy().
1451 PH/23 POSIX allows open() to be a macro; guard against that.
1453 PH/24 If the recipient of an error message contained an @ in the local part
1454 (suitably quoted, of course), incorrect values were put in $domain and
1455 $local_part during the evaluation of errors_copy.
1461 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1462 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1465 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1466 issue a MAIL command.
1468 PH/02 In an ACL statement such as
1470 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1472 if a client was not listed at all, or was listed with a value other than
1473 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1474 the condition was not true (as it should be), so access was not denied.
1475 The bug was that the ! inversion was incorrectly passed on to the second
1476 item. This has been fixed.
1478 PH/03 Added additional dnslists conditions == and =& which are different from
1479 = and & when the dns lookup returns more than one IP address.
1481 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1482 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1484 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1485 FSYNC, which compiles an option called disable_fsync that allows for
1486 bypassing fsync(). The documentation is heavily laced with warnings.
1488 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1490 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1491 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1492 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1493 including adding "make clean"; (3) Added -fPIC when compiling the test
1494 dynamically loaded module, to get rid of a warning.
1496 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1497 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1498 = 0s. The bug is otherwise harmless.
1500 PH/07 There was a bug in the dovecot authenticator such that the value of
1501 $auth1 could be overwritten, and so not correctly preserved, after a
1502 successful authentication. This usually meant that the value preserved by
1503 the server_setid option was incorrect.
1505 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1507 PH/09 Installed PCRE release 7.0.
1509 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1510 run for batched SMTP input. It is now run at the start of every message
1511 in the batch. While fixing this I discovered that the process information
1512 (output by running exiwhat) was not always getting set for -bs and -bS
1513 input. This is fixed, and it now also says "batched" for BSMTP.
1515 PH/11 Added control=no_pipelining.
1517 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1518 patch, slightly modified), and move the expansion of helo_data till after
1519 the connection is made in the smtp transport (so it can use these
1522 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1524 PH/14 Added log_selector = +pid.
1526 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1528 PH/16 Add ${if forany and ${if forall.
1530 PH/17 Added dsn_from option to vary the From: line in DSNs.
1532 PH/18 Flush SMTP output before performing a callout, unless control =
1533 no_callout_flush is set.
1535 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1536 was true (the default) a successful delivery failed to delete the retry
1537 item, thus causing premature timeout of the address. The bug is now
1540 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1542 PH/21 Long custom messages for fakedefer and fakereject are now split up
1543 into multiline reponses in the same way that messages for "deny" and
1544 other ACL rejections are.
1546 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1547 with slight modification.
1549 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1550 draft, changing the syntax and factoring some duplicate code".
1552 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1553 for deliveries of the second and subsequent messages over the same SMTP
1556 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1557 ${reduce, with only minor "tidies".
1559 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1561 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1562 expansion side effects.
1564 PH/27 When a message times out after an over-quota error from an Exim-imposed
1565 quota, the bounce message says "mailbox is full". This message was not
1566 being given when it was a system quota that was exceeded. It now should
1569 MH/03 Made $recipients available in local_scan(). local_scan() already has
1570 better access to the recipient list through recipients_list[], but
1571 $recipients can be useful in postmaster-provided expansion strings.
1573 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1574 in the case of a MAIL command with additional options following the
1575 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1576 were accidentally chopped off.
1578 PH/29 SMTP synchronization checks are implemented when a command is read -
1579 there is a check that no more input is waiting when there shouldn't be
1580 any. However, for some commands, a delay in an ACL can mean that it is
1581 some time before the response is written. In this time, more input might
1582 arrive, invalidly. So now there are extra checks after an ACL has run for
1583 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1584 pipelining has not been advertised.
1586 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1588 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1589 returned in the SMTP response when smtp_return_error_details was false.
1590 This has been fixed.
1592 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1593 instead of the C I/O that was originally supplied, because problems were
1594 reported on Solaris.
1596 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1597 Exim which did not show up earlier: it was assuming that a call to
1598 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1599 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1600 was a macro that became an assignment, so it seemed to work. This has
1601 changed to a proper function call with a void return, hence the compile
1602 error. Exim's code has been fixed.
1604 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1607 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1609 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1611 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1612 because it turns out to be too big or there is a timeout) there is no
1613 "Completed" line in the log. When some messages of this type were
1614 selected by exigrep, they were listed as "not completed". Others were
1615 picked up by some special patterns. I have improved the selection
1616 criteria to be more general.
1618 PH/38 The host_find_failed option in the manualroute router can now be set
1619 to "ignore", to completely ignore a host whose IP address cannot be
1620 found. If all hosts are ignored, the behaviour is controlled by the new
1621 host_all_ignored option.
1623 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1624 homing or because of multiple MX records with /mx) generated more than
1625 one IP address, and the following item turned out to be the local host,
1626 all the secondary addresses of the first item were incorrectly removed
1627 from the list, along with the local host and any following hosts (which
1628 is what is supposed to happen).
1630 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1631 whoever called Exim into the -H file. In the case of the daemon it was
1632 behaving confusingly. When first started, it used values for whoever
1633 started the daemon, but after a SIGHUP it used the Exim user (because it
1634 calls itself on a restart). I have changed the code so that it now always
1637 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1638 message are rejected with the same error (e.g. no authentication or bad
1639 sender address), and a DATA command is nevertheless sent (as can happen
1640 with PIPELINING or a stupid MUA), the error message that was given to the
1641 RCPT commands is included in the rejection of the DATA command. This is
1642 intended to be helpful for MUAs that show only the final error to their
1645 PH/42 Another patch from the Sieve maintainer.
1647 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1648 Eximstats - Fixed some broken HTML links and added missing column headers
1650 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1651 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1653 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1655 PH/43 Yet another patch from the Sieve maintainer.
1657 PH/44 I found a way to check for a TCP/IP connection going away before sending
1658 the response to the final '.' that terminates a message, but only in the
1659 case where the client has not sent further data following the '.'
1660 (unfortunately, this is allowed). However, in many cases there won't be
1661 any further data because there won't be any more messages to send. A call
1662 to select() can be used: if it shows that the input is "ready", there is
1663 either input waiting, or the socket has been closed. An attempt to read
1664 the next input character can distinguish the two cases. Previously, Exim
1665 would have sent an OK response which the client would never have see.
1666 This could lead to message repetition. This fix should cure that, at
1667 least in a lot of common cases.
1669 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1670 advertised in response to EHLO.
1676 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1677 fixed by 4.65/MH/01 (is this a record?) are fixed:
1679 (i) An empty string was always treated as zero by the numeric comparison
1680 operators. This behaviour has been restored.
1682 (ii) It is documented that the numeric comparison operators always treat
1683 their arguments as decimal numbers. This was broken in that numbers
1684 starting with 0 were being interpreted as octal.
1686 While fixing these problems I realized that there was another issue that
1687 hadn't been noticed. Values of message_size_limit (both the global option
1688 and the transport option) were treated as octal if they started with 0.
1689 The documentation was vague. These values are now always treated as
1690 decimal, and I will make that clear in the documentation.
1696 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1697 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1700 MH/01 Don't check that the operands of numeric comparison operators are
1701 integers when their expansion is in "skipping" mode (fixes bug
1702 introduced by 4.64-PH/07).
1704 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1705 child addresses, Exim now panics and dies. Previously, because the count
1706 is held in a short int, deliveries were likely to be lost. As such a
1707 large number of recipients for a single message is ridiculous
1708 (performance will be very, very poor), I have chosen to impose a limit
1709 rather than extend the field.
1715 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1716 leftover -K file (the existence of which was triggered by #402).
1717 While we were at it, introduced process PID as part of the -K
1718 filename. This should rule out race conditions when creating
1721 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1722 processing considerably. Previous code took too long for large mails,
1723 triggering a timeout which in turn triggers #401.
1725 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1726 in the DK code in transports.c. sendfile() is not really portable,
1727 hence the _LINUX specificness.
1729 TF/01 In the add_headers option to the mail command in an Exim filter,
1730 there was a bug that Exim would claim a syntax error in any
1731 header after the first one which had an odd number of characters
1734 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1735 callout verification, Exim cached a "reject" for the entire domain. This
1736 is correct for most verifications, but it is not correct for a recipient
1737 verification with use_sender or use_postmaster set, because in that case
1738 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1739 case of MAIL FROM:<> rejection from other early rejections (e.g.
1740 rejection of HELO). When verifying a recipient using a non-null MAIL
1741 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1742 Whatever the result of the callout, the value of the domain cache is
1743 left unchanged (for any other kind of callout, getting as far as trying
1744 RCPT means that the domain itself is ok).
1746 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1749 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1750 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1751 session, the daemon ignores ECONNECT errors and logs others; it now
1752 ignores EPIPE as well.
1754 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1755 (quoted-printable decoding).
1757 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1758 later the small subsequent patch to fix an introduced bug.
1760 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1762 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1764 PH/08 An error is now given if message_size_limit is specified negative.
1766 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1767 to be given (somewhat) arbitrary names.
1769 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1772 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1773 miscellaneous code fixes
1775 PH/10 Added the log_reject_target ACL modifier to specify where to log
1778 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1779 hostname. This is wrong, because it relates to the incoming message (and
1780 probably the interface on which it is arriving) and not to the outgoing
1781 callout (which could be using a different interface). This has been
1782 changed to use the value of the helo_data option from the smtp transport
1783 instead - this is what is used when a message is actually being sent. If
1784 there is no remote transport (possible with a router that sets up host
1785 addresses), $smtp_active_hostname is used.
1787 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1788 tweaks were necessary in order to get it to work (see also 21 below):
1789 (a) The code assumed that strncpy() returns a negative number on buffer
1790 overflow, which isn't the case. Replaced with Exim's string_format()
1792 (b) There were several signed/unsigned issues. I just did the minimum
1793 hacking in of casts. There is scope for a larger refactoring.
1794 (c) The code used strcasecmp() which is not a standard C function.
1795 Replaced with Exim's strcmpic() function.
1796 (d) The code set only $1; it now sets $auth1 as well.
1797 (e) A simple test gave the error "authentication client didn't specify
1798 service in request". It would seem that Dovecot has changed its
1799 interface. Fortunately there's a specification; I followed it and
1800 changed what the client sends and it appears to be working now.
1802 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1805 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1806 address A is aliased to B and C, where B exists and C does not. Without
1807 -v the output is "A verified" because verification stops after a
1808 successful redirection if more than one address is generated. However,
1809 with -v the child addresses are also verified. Exim was outputting "A
1810 failed to verify" and then showing the successful verification for C,
1811 with its parentage. It now outputs "B failed to verify", showing B's
1812 parentage before showing the successful verification of C.
1814 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1815 look up a TXT record in a specific list after matching in a combined
1818 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1819 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1820 they consult the DNS. I had assumed they would set it the way they
1821 wanted; and indeed my experiments on Linux seem to show that in some
1822 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1823 To be on the safe side, however, I have now made the interface to
1824 host_find_byname() similar to host_find_bydns(), with an argument
1825 containing the DNS resolver options. The host_find_byname() function now
1826 sets these options at its start, just as host_find_bydns() does. The smtp
1827 transport options dns_qualify_single and dns_search_parents are passed to
1828 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1829 of host_find_byname() use the default settings of RES_DEFNAMES
1830 (qualify_single) but not RES_DNSRCH (search_parents).
1832 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1833 spool_read_header() do less string testing, by means of a preliminary
1834 switch on the second character of optional "-foo" lines. (This is
1835 overdue, caused by the large number of possibilities that now exist.
1836 Originally there were few.) While I was there, I also converted the
1837 str(n)cmp tests so they don't re-test the leading "-" and the first
1838 character, in the hope this might squeeze out yet more improvement.
1840 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1841 flag allowing group syntax was set by the header_syntax check but not
1842 turned off, possible causing trouble later; (2) The flag was not being
1843 set at all for the header_verify test, causing "group"-style headers to
1844 be rejected. I have now set it in this case, and also caused header_
1845 verify to ignore an empty address taken from a group. While doing this, I
1846 came across some other cases where the code for allowing group syntax
1847 while scanning a header line wasn't quite right (mostly, not resetting
1848 the flag correctly in the right place). These bugs could have caused
1849 trouble for malformed header lines. I hope it is now all correct.
1851 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1852 with the "reply" argument non-NULL. The code, however (which originally
1853 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1854 but it didn't always do it. This confused somebody who was copying the
1855 code for some other use. I have removed all the tests.
1857 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1858 feature that was used to support insecure browsers during the U.S. crypto
1859 embargo. It requires special client support, and Exim is probably the
1860 only MTA that supported it -- and would never use it because real RSA is
1861 always available. This code has been removed, because it had the bad
1862 effect of slowing Exim down by computing (never used) parameters for the
1863 RSA_EXPORT functionality.
1865 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1866 authenticator to fail if there's a tab character in the incoming data
1867 (there should never be unless someone is messing about, as it's supposed
1868 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1869 if the connection is using TLS or if the remote IP is the same as the
1870 local IP, and the "valid-client-cert option" if a client certificate has
1873 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1874 authenticators. This can be used for authorization after authentication
1875 succeeds. (In the case of plaintext, it servers for both authentication
1878 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1879 if any retry times were supplied.
1881 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1882 connection, where there is no client IP address to check. In this
1883 situation, the verify now always succeeds.
1885 PH/25 Applied John Jetmore's -Mset patch.
1887 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1889 PH/27 In a string expansion for a processed (not raw) header when multiple
1890 headers of the same name were present, leading whitespace was being
1891 removed from all of them, but trailing whitespace was being removed only
1892 from the last one. Now trailing whitespace is removed from each header
1893 before concatenation. Completely empty headers in a concatenation (as
1894 before) are ignored.
1896 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1897 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1899 PH/29 [Removed. This was a change that I later backed out, and forgot to
1900 correct the ChangeLog entry (that I had efficiently created) before
1901 committing the later change.]
1903 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1904 address errors (4xx response to RCPT) over the same connection as other
1905 messages routed to the same hosts. Such deliveries are always "forced",
1906 so retry times are not inspected. This resulted in far too many retries
1907 for the affected addresses. The effect occurred only when there were more
1908 hosts than the hosts_max_try setting in the smtp transport when it had
1909 the 4xx errors. Those hosts that it had tried were not added to the list
1910 of hosts for which the message was waiting, so if all were tried, there
1911 was no problem. Two fixes have been applied:
1913 (i) If there are any address or message errors in an SMTP delivery, none
1914 of the hosts (tried or untried) are now added to the list of hosts
1915 for which the message is waiting, so the message should not be a
1916 candidate for sending over the same connection that was used for a
1917 successful delivery of some other message. This seems entirely
1918 reasonable: after all the message is NOT "waiting for some host".
1919 This is so "obvious" that I'm not sure why it wasn't done
1920 previously. Hope I haven't missed anything, but it can't do any
1921 harm, as the worst effect is to miss an optimization.
1923 (ii) If, despite (i), such a delivery is accidentally attempted, the
1924 routing retry time is respected, so at least it doesn't keep
1925 hammering the server.
1927 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1928 in ${readsocket because some servers need this prod.
1930 PH/32 Added some extra debug output when updating a wait-xxx database.
1932 PH/33 The hint "could be header name not terminated by colon", which has been
1933 given for certain expansion errors for a long time, was not being given
1934 for the ${if def:h_colon_omitted{... case.
1936 PH/34 The spec says: "With one important exception, whenever a domain list is
1937 being scanned, $domain contains the subject domain." There was at least
1938 one case where this was not true.
1940 PH/35 The error "getsockname() failed: connection reset by peer" was being
1941 written to the panic log as well as the main log, but it isn't really
1942 panic-worthy as it just means the connection died rather early on. I have
1943 removed the panic log writing for the ECONNRESET error when getsockname()
1946 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1947 runs only) independently of the message's sender address. This meant
1948 that, if the 4xx error was in fact related to the sender, a different
1949 message to the same recipient with a different sender could confuse
1950 things. In particualar, this can happen when sending to a greylisting
1951 server, but other circumstances could also provoke similar problems.
1952 I have changed the default so that the retry time for these errors is now
1953 based a combination of the sender and recipient addresses. This change
1954 can be overridden by setting address_retry_include_sender=false in the
1957 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1958 remote server are returned as part of bounce messages. This was not
1959 happening for LMTP over a pipe (the lmtp transport), but now it is the
1960 same for both kinds of LMTP.
1962 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1963 in header lines that were in fact CNAMEs. This is no longer the case.
1965 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1966 and queue runs started by the daemon processed all messages. This has
1967 been fixed so that -R and -S can now usefully be given with -q<time>.
1969 PH/40 Import PCRE release 6.7 (fixes some bugs).
1971 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1973 PH/42 Give an error if -q is specified more than once.
1975 PH/43 Renamed the variables $interface_address and $interface_port as
1976 $received_ip_address and $received_port, to make it clear that these
1977 values apply to message reception, and not to the outgoing interface when
1978 a message is delivered. (The old names remain recognized, of course.)
1980 PH/44 There was no timeout on the connect() call when using a Unix domain
1981 socket in the ${readsocket expansion. There now is.
1983 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1984 be meaningful with "accept".
1986 SC/01 Eximstats V1.43
1987 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1989 SC/02 Eximstats V1.44
1990 Use a glob alias rather than an array ref in the generated
1991 parser. This improves both readability and performance.
1993 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1994 Collect SpamAssassin and rejection statistics.
1995 Don't display local sender or destination tables unless
1996 there is data to show.
1997 Added average volumes into the top table text output.
1999 SC/04 Eximstats V1.46
2000 Collect data on the number of addresses (recipients)
2001 as well as the number of messages.
2003 SC/05 Eximstats V1.47
2004 Added 'Message too big' to the list of mail rejection
2005 reasons (thanks to Marco Gaiarin).
2007 SC/06 Eximstats V1.48
2008 Mainlog lines which have GMT offsets and are too short to
2009 have a flag are now skipped.
2011 SC/07 Eximstats V1.49 (Alain Williams)
2012 Added the -emptyok flag.
2014 SC/08 Eximstats V1.50
2015 Fixes for obtaining the IP address from reject messages.
2017 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2018 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2019 whitesspace changes from 4.64-PH/27
2021 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2024 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2025 are found, allow negative numbers in numeric criteria)
2027 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2029 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2030 to match changes made in 4.64-PH/43
2032 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2034 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2035 "domain" is actually an IP address, and force a failure. This locks out
2036 those revolvers/nameservers that support "A-for-A" lookups, in
2037 contravention of the specifications.
2039 PH/48 When a host name was looked up from an IP address, and the subsequent
2040 forward lookup of the name timed out, the host name was left in
2041 $sender_host_name, contrary to the specification.
2043 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2044 restricted to single-key lookups, Exim was not diagnosing an error if
2045 * or *@ was used with a query-style lookup.
2047 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2049 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2050 long ago, but noone interested enough thought of it. Let's just say that
2051 the "1.1" means that there are some new functions that weren't there at
2052 some point in the past.
2054 PH/51 Error processing for expansion failure of helo_data from an smtp
2055 transport during callout processing was broken.
2057 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2058 tested/used via the -bh/-bhc/-bs options.
2060 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2061 bug, fixed in subsequent PCRE releases).
2063 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2064 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2066 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2072 SC/01 Use a glob alias rather than an array ref in eximstats generated
2073 parser. This improves both readability and performance.
2075 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2076 Don't display local sender or destination tables in eximstats unless
2077 there is data to show.
2078 Added average volumes into the eximstats top table text output.
2080 SC/03 Collect data on the number of addresses (recipients) as well
2081 as the number of messages in eximstats.
2083 TF/01 Correct an error in the documentation for the redirect router. Exim
2084 does (usually) call initgroups() when daemonizing.
2086 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2087 with consistent privilege compared to when running as a daemon.
2089 TF/03 Note in the spec that $authenticated_id is not set for local
2090 submissions from trusted users.
2092 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2093 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2095 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2096 by adding some example configuration directives to the default
2097 configuration file. A little bit of work is required to uncomment the
2098 directives and define how usernames and passwords are checked, but
2099 there is now a framework to start from.
2101 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2102 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2103 without this. I don't know how relevant this is to other LDAP libraries.
2105 PH/02 Add the verb name to the "unknown ACL verb" error.
2107 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2109 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2111 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2112 directory not expanded when it should be if an expanded home directory
2113 was set for the address (which is overridden by the transport).
2115 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2118 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2119 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2120 because it is too late at that time, and has no effect.
2122 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2123 security issue with \' (bugzilla #107). I could not use the
2124 PQescapeStringConn() function, because it needs a PGconn value as one of
2127 PH/08 When testing addresses using -bt, indicate those final addresses that
2128 are duplicates that would not cause an additional delivery. At least one
2129 person was confused, thinking that -bt output corresponded to deliveries.
2130 (Suppressing duplicates isn't a good idea as you lose the information
2131 about possibly different redirections that led to the duplicates.)
2133 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2134 systems where poll() doesn't work, in particular OS X.
2136 PH/10 Added more information to debugging output for retry time not reached.
2138 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2139 operations in malware.c.
2141 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2144 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2145 syslog_duplication set false, log lines that would normally be written
2146 both the the main log and to the reject log were not written to syslog at
2149 PH/14 In the default configuration, change the use of "message" in ACL warn
2150 statements to "add_header".
2152 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2153 not followed by a command (e.g. "seen endif").
2155 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2156 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2159 PH/17 Added extra conditions to the default value of delay_warning_condition
2163 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2164 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2165 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2168 The Auto-Submitted: and various List- headers are standardised, whereas I
2169 don't think Precedence: ever was.
2171 PH/18 Refactored debugging code in route_finduser() to show more information,
2172 in particular, the error code if getpwnam() issues one.
2174 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2175 This is apparently needed in addition to the PH/07 change above to avoid
2176 any possible encoding problems.
2178 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2179 but not after initializing Perl.
2181 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2182 output them only if debugging. By default they are written stderr,
2183 apparently, which is not desirable.
2185 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2188 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2191 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2193 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2194 authenticated or an ident call has been made. Suppress the default
2195 values for $authenticated_id and $authenticated_sender (but permit -oMai
2196 and -oMas) when testing with -bh.
2198 PH/24 Re-jigged the order of the tests in the default configuration so that the
2199 tests for valid domains and recipients precede the DNS black list and CSA
2200 tests, on the grounds that those ones are more expensive.
2202 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2203 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2204 This bug exists in every version of Exim that I still have, right back to
2207 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2208 However, an attempt to turn on case-sensitivity in a regex key by
2209 including (?-i) didn't work because the subject string was already
2210 lowercased, and the effects were non-intuitive. It turns out that a
2211 one-line patch can be used to allow (?-i) to work as expected.
2217 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2218 other effects) broke the use of negated acl sub-conditions.
2220 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2223 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2224 "Deny" causes Exim to reject the incoming connection with a 554 error.
2225 Unfortunately, if there is a major crisis, such as a disk failure,
2226 tcp-wrappers gives "deny", whereas what one would like would be some
2227 kind of temporary error. A kludge has been added to help with this.
2228 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2229 554 error is used if errno is still zero or contains ENOENT (which occurs
2230 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2233 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2235 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2236 errors. Otherwise a message that provokes a temporary error (when other
2237 messages do not) can cause a whole host to time out.
2239 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2240 addresses were routed directly to files or pipes from a redirect router.
2241 File deliveries just didn't batch; pipe deliveries might have suffered
2244 PH/06 A failure to get a lock for a hints database would erroneously always say
2245 "Failed to get write lock", even when it was really a read lock.
2247 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2248 of 0600. This has been changed to use the value of the lockfile_mode
2249 option (which defaults to 0600).
2251 PH/08 Applied small patch from the Sieve maintainer.
2253 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2254 folder from quota calculations, a direct delivery into this folder messed
2255 up the contents of the maildirsize file. This was because the regex was
2256 used only to exclude .Trash (or whatever) when the size of the mailbox
2257 was calculated. There was no check that a delivery was happening into an
2258 excluded directory. This bug has been fixed by ignoring all quota
2259 processing for deliveries into excluded directories.
2261 PH/10 Added the maildirfolder_create_regex option to appendfile.
2267 PH/01 The code for finding all the local interface addresses on a FreeBSD
2268 system running IPv6 was broken. This may well have applied to all BSD
2269 systems, as well as to others that have similar system calls. The broken
2270 code found IPv4 interfaces correctly, but gave incorrect values for the
2271 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2272 that it would not match correctly against @[] and not recognize the IPv6
2275 PH/02 The ipliteral router was not recognizing addresses of the form user@
2276 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2278 PH/03 Added disable_ipv6.
2280 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2281 decoded form, because it is most often used to construct To: headers
2282 lines in autoreplies, and the decoded form may well be syntactically
2283 invalid. However, $reply_address has leading white space removed, and all
2284 newlines turned into spaces so that the autoreply transport does not
2287 PH/05 If group was specified without a user on a router, and no group or user
2288 was specified on a transport, the group from the router was ignored.
2290 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2291 for visible compile-time settings that can be used to change these
2292 numbers, for those that want even more. Backwards compatibility with old
2293 spool files has been maintained. However, going back to a previous Exim
2294 release will lost any variables that are in spool files.
2296 PH/07 Two small changes when running in the test harness: increase delay when
2297 passing a TCP/IP connection to a new process, in case the original
2298 process has to generate a bounce, and remove special handling of
2299 127.0.0.2 (sic), which is no longer necessary.
2301 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2302 be the same on different OS.
2304 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2307 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2308 whether --show-vars was specified or not
2310 JJ/02 exipick: Added support for new ACL variable spool format introduced
2313 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2314 syntactically invalid From: or Reply-to: line, and a filter used this to
2315 generate an autoreply, and therefore failed to obtain an address for the
2316 autoreply, Exim could try to deliver to a non-existent relative file
2317 name, causing unrelated and misleading errors. What now happens is that
2318 it logs this as a hard delivery error, but does not attempt to create a
2321 PH/11 The exinext utility has a -C option for testing purposes, but although
2322 the given file was scanned by exinext itself; it wasn't being passed on
2323 when Exim was called.
2325 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2326 an end-of-file indication when reading a command response.
2328 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2329 compiled. In many other places in Exim, IPv6 addresses are always
2330 recognized, so I have changed this. It also means that IPv4 domain
2331 literals of the form [IPV4:n.n.n.n] are now always recognized.
2333 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2334 used if the router is not running as root, for example, when verifying at
2335 ACL time, or when using -bh. The debugging output from this situation was
2336 non-existent - all you got was a failure to exec. I have made two
2339 (a) Failures to set uid/gid, the current directory, or a process leader
2340 in a subprocess such as that created by queryprogram now generate
2341 suitable debugging ouput when -d is set.
2343 (b) The queryprogram router detects when it is not running as root,
2344 outputs suitable debugging information if -d is set, and then runs
2345 the subprocess without attempting to change uid/gid.
2347 PH/15 Minor change to Makefile for building test_host (undocumented testing
2350 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2351 additional section of a DNS packet that returns MX or SRV records.
2352 Instead, it always explicitly searches for A/AAAA records. This avoids
2353 major problems that occur when a DNS server includes only records of one
2354 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2355 fixed another bug: if SRV records were looked up and the corresponding
2356 address records were *not* found in the additional section, the port
2357 values from the SRV records were lost.
2359 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2360 using the correct key (the original address) when searching the retry
2361 rules in order to find which one to use for generating the retry hint.
2363 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2364 adding the default one. Similarly, if it contains a Reply-To: header, the
2365 errors_reply_to option, if set, is not used.
2367 PH/19 When calculating a retry time, Exim used to measure the "time since
2368 failure" by looking at the "first failed" field in the retry record. Now
2369 it does not use this if it is later than than the arrival time of the
2370 message. Instead it uses the arrival time. This makes for better
2371 behaviour in cases where some deliveries succeed, thus re-setting the
2372 "first failed" field. An example is a quota failure for a huge message
2373 when small messages continue to be delivered. Without this change, the
2374 "time since failure" will always be short, possible causing more frequent
2375 delivery attempts for the huge message than are intended.
2376 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2378 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2379 $1, $2, $3) because the numerical variables can be reset during some
2380 expansion items (e.g. "match"), thereby losing the authentication data.
2382 PH/21 Make -bV show the size of off_t variables so that the test suite can
2383 decide whether to run tests for quotas > 2G.
2385 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2386 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2387 filecount value is greater than 2G or if a quota value is greater than 2G
2388 on a system where the size of off_t is not greater than 4, a panic error
2391 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2392 never match. The debug and -bh output now contains an explicit error
2393 message indicating a malformed IPv4 address or mask.
2395 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2396 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2397 PH/23 above applies.
2399 PH/25 Do not write to syslog when running in the test harness. The only
2400 occasion when this arises is a failure to open the main or panic logs
2401 (for which there is an explicit test).
2403 PH/26 Added the /no_tell option to "control=freeze".
2405 PH/27 If a host name lookup failed very early in a connection, for example, if
2406 the IP address matched host_lookup and the reverse lookup yielded a name
2407 that did not have a forward lookup, an error message of the form "no IP
2408 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2409 could be logged. Now it outputs the IP address instead of "NULL".
2411 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2412 allows the sender and the authenticated sender to be set when
2413 submitting a message from within Exim. Since child_open_exim() is
2414 documented for local_scan(), the new function should be too.
2416 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2417 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2418 results in an empty string is now treated as unset.
2420 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2422 PH/31 Added sender_verify_fail logging option.
2424 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2425 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2426 tidied the source and removed it altogether.
2428 PH/33 When a queue run was abandoned because the load average was too high, a
2429 log line was always written; now it is written only if the queue_run log
2430 selector is set. In addition, the log line for abandonment now contains
2431 information about the queue run such as the pid. This is always present
2432 in "start" and "stop" lines but was omitted from the "abandon" line.
2434 PH/34 Omit spaces between a header name and the colon in the error message that
2435 is given when verify = headers_syntax fails (if there are lots of them,
2436 the message gets confusing).
2438 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2439 names, as there are now some PTR records that contain slashes. This check
2440 is only to protect against broken name servers that fall over on strange
2441 characters, so the fact that it applies to all lookups doesn't matter.
2443 PH/36 Now that the new test suite is complete, we can remove some of the
2444 special code in Exim that was needed for the old test suite. For example,
2445 sorting DNS records because real resolvers return them in an arbitrary
2446 order. The new test suite's fake resolver always returns records in the
2449 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2450 bounces) except when queue_only is set, to avoid logging races between
2451 the different processes.
2453 PH/38 Panic-die if .include specifies a non-absolute path.
2455 PH/39 A tweak to the "H" retry rule from its user.
2457 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2458 a label. They prevented compilation on older perls.
2460 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2461 a warning to be raised on newish perls.
2463 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2464 on queue. Changes to match documented behaviour of showing count of
2465 messages matching specified criteria.
2467 PH/40 Changed the default ident timeout from 30s to 5s.
2469 PH/41 Added support for the use of login_cap features, on those BSD systems
2470 that have them, for controlling the resources used by pipe deliveries.
2472 PH/42 The content-scanning code uses fopen() to create files in which to put
2473 message data. Previously it was not paying any attention to the mode of
2474 the files. Exim runs with umask(0) because the rest of the code creates
2475 files with open(), and sets the required mode explicitly. Thus, these
2476 files were ending up world-writeable. This was not a big issue, because,
2477 being within the spool directory, they were not world-accessible. I have
2478 created a function called modefopen, which takes an additional mode
2479 argument. It sets umask(777), creates the file, chmods it to the required
2480 mode, then resets the umask. All the relevant calls to fopen() in the
2481 content scanning code have been changed to use this function.
2483 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2484 to 24 hours. This avoids potential overflow problems when processing G
2485 and H retry rules. I suspect nobody ever tinkers with this value.
2487 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2489 PH/45 When the plaintext authenticator is running as a client, the server's
2490 challenges are checked to ensure they are valid base64 strings. By
2491 default, the authentication attempt is cancelled if an invalid string is
2492 received. Setting client_ignore_invalid_base64 true ignores these errors.
2493 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2494 they are received. Thus, the responses can be made to depend on the
2495 challenges. If an invalid string is ignored, an empty string is placed in
2498 PH/46 Messages that are created by the autoreply transport now contains a
2499 References: header, in accordance with RFCs 2822 and 3834.
2501 PH/47 Added authenticated_sender_force to the smtp transport.
2503 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2505 PH/49 Installed latest patch from the Sieve maintainer.
2507 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2508 was also set, the appendfile transport was unnecessarily scanning a
2509 directory of message files (e.g. for maildir delivery) to find the count
2510 of files (along with the size), even though it did not need this
2511 information. It now does the scan only if it needs to find either the
2512 size of the count of files.
2514 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2516 PH/52 Two bugs concerned with error handling when the smtp transport is
2519 (i) Exim was not creating retry information for temporary errors given
2520 for individual recipients after the DATA command when the smtp transport
2521 was used in LMTP mode. This meant that they could be retried too
2522 frequently, and not timed out correctly.
2524 (ii) Exim was setting the flag that allows error details to be returned
2525 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2526 recipients that were returned after the DATA command.
2528 PH/53 This is related to PH/52, but is more general: for any failing address,
2529 when detailed error information was permitted to be returned to the
2530 sender, but the error was temporary, then after the final timeout, only
2531 "retry timeout exceeded" was returned. Now it returns the full error as
2532 well as "retry timeout exceeded".
2534 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2535 do this, and (what is worse) MTAs that accept it.
2537 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2538 will now be deprecated.
2540 PH/56 New os.c-cygwin from the Cygwin maintainer.
2542 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2543 formats (previously only available in exim formats via -bpr, -bpru,
2544 and -bpra. Now also available in native and exiqgrep formats)
2546 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2547 with very large, slow to parse queues
2549 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2551 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2553 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2554 responses to authentication challenges, though it was showing the
2555 challenges; (ii) I've removed the CR characters from the debug output for
2558 PH/58 Allow for the insertion of a newline as well as a space when a string
2559 is turned into more than one encoded-word during RFC 2047 encoding. The
2560 Sieve code now uses this.
2562 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2563 data_4xx, lost_connection, tls_required.
2565 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2566 message was being sent as an SMTP response.
2568 PH/61 Add -l and -k options to exicyclog.
2570 PH/62 When verifying, if an address was redirected to one new address, so that
2571 verification continued, and the new address failed or deferred after
2572 having set something in $address_data, the value of $address_data was not
2573 passed back to the ACL. This was different to the case when no
2574 redirection occurred. The value is now passed back in both cases.
2576 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2577 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2578 use login_cap.h, so on its own it isn't the distinguishing feature. The
2579 new name refers directly to the setclassresources() function.
2581 PH/65 Added configuration files for NetBSD3.
2583 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2585 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2586 is preferred over IPv4.
2588 PH/68 The bounce_return_message and bounce_return_body options were not being
2589 honoured for bounces generated during the reception of non-SMTP messages.
2590 In particular, this applied to messages rejected by the ACL. This bug has
2591 been fixed. However, if bounce_return_message is true and bounce_return_
2592 body is false, the headers that are returned for a non-SMTP message
2593 include only those that have been read before the error was detected.
2594 (In the case of an ACL rejection, they have all been read.)
2596 PH/69 The HTML version of the specification is now built in a directory called
2597 spec_html instead of spec.html, because the latter looks like a path with
2598 a MIME-type, and this confuses some software.
2600 PH/70 Catch two compiler warnings in sieve.c.
2602 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2603 function verify_get_ident() calls ip_connect() to connect a socket, but
2604 if the "connect()" function timed out, ip_connect() used to close the
2605 socket. However, verify_get_ident() also closes the socket later, and in
2606 between Exim writes to the log, which may get opened at this point. When
2607 the socket was closed in ip_connect(), the log could get the same file
2608 descriptor number as the socket. This naturally causes chaos. The fix is
2609 not to close the socket in ip_connect(); the socket should be closed by
2610 the function that creates it. There was only one place in the code where
2611 this was missing, in the iplookup router, which I don't think anybody now
2612 uses, but I've fixed it anyway.
2614 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2615 well as to direct DNS lookups. Otherwise the handling of names in host
2616 lists is inconsistent and therefore confusing.
2622 PH/01 Two changes to the default runtime configuration:
2624 (1) Move the checks for relay_from_hosts and authenticated clients from
2625 after to before the (commented out) DNS black list checks.
2627 (2) Add control=submission to the relay_from_hosts and authenticated
2628 clients checks, on the grounds that messages accepted by these
2629 statements are most likely to be submissions.
2631 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2633 (1) Generate an error if the third argument for the ${prvs expansion is
2636 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2639 (3) Reset the variables that are obtained from the first argument of
2640 ${prvscheck and used in the second argument before leaving the code,
2641 because their memory is reclaimed, so using them afterwards may do
2644 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2645 one (it's much easier than Tom thought :-).
2647 (5) Because of (4), we can now allow for the use of $prvscheck_result
2648 inside the third argument.
2650 PH/03 For some reason, the default setting of PATH when running a command from
2651 a pipe transport was just "/usr/bin". I have changed it to
2654 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2655 anything to be listed in the output from -bV.
2657 PH/05 When a filter generated an autoreply, the entire To: header line was
2658 quoted in the delivery log line, like this:
2660 => >A.N.Other <ano@some.domain> <original@ddress> ...
2662 This has been changed so that it extracts the operative address. There
2663 may be more than one such address. If so, they are comma-separated, like
2666 => >ano@some.domain,ona@other.domain <original@ddress> ...
2668 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2669 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2670 not being looked up in the rDNS to get a host name, Exim was showing the
2671 IP address twice in Received: lines, even though the IP addresses were
2672 identical. For example:
2674 Received: from [1.2.3.4] (helo=[1.2.3.4])
2676 However, if the real host name was known, it was omitting the HELO data
2677 if it matched the actual IP address. This has been tidied up so that it
2678 doesn't show the same IP address twice.
2680 PH/07 When both +timestamp and +memory debugging was on, the value given by
2681 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2682 called by the debug printing, thereby overwriting the timestamp buffer.
2683 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2685 PH/08 When the original message was included in an autoreply transport, it
2686 always said "this is a copy of the message, including all the headers",
2687 even if body_only or headers_only was set. It now gives an appropriate
2690 PH/09 Applied a patch from the Sieve maintainer which:
2692 o fixes some comments
2693 o adds the (disabled) notify extension core
2694 o adds some debug output for the result of if/elsif tests
2695 o points to the current vacation draft in the documentation
2696 and documents the missing references header update
2700 o fixes a bug in processing the envelope test (when testing
2701 multiple envelope elements, the last element determinted the
2704 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2705 Electronic Mail") by including:
2707 Auto-submitted: auto-generated
2709 in the messages that it generates (bounce messages and others, such as
2710 warnings). In the case of bounce messages for non-SMTP mesages, there was
2711 also a typo: it was using "Auto_submitted" (underscore instead of
2712 hyphen). Since every message generated by Exim is necessarily in response
2713 to another message, thes have all been changed to:
2715 Auto-Submitted: auto-replied
2717 in accordance with these statements in the RFC:
2719 The auto-replied keyword:
2721 - SHOULD be used on messages sent in direct response to another
2722 message by an automatic process,
2724 - MUST NOT be used on manually-generated messages,
2726 - MAY be used on Delivery Status Notifications (DSNs) and Message
2727 Disposition Notifications (MDNs),
2729 - MUST NOT be used on messages generated by automatic or periodic
2730 processes, except for messages which are automatic responses to
2733 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2734 to the default Received: header definition.
2736 PH/12 Added log selector acl_warn_skipped (default on).
2738 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2739 variables because (a) they are in the wrong storage pool and (b) even if
2740 they were copied, it wouldn't work properly because of the caching.
2742 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2743 checking when decoding. Apparently there are clients that generate
2744 overlong encoded strings. Why am I not surprised?
2746 PH/15 If the first argument of "${if match_address" was not empty, but did not
2747 contain an "@" character, Exim crashed. Now it writes a panic log message
2748 and treats the condition as false.
2750 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2752 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2753 extension "envelope-auth". The code is finished and in agreement with
2754 other implementations, but there is no documentation so far and in fact,
2755 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2756 not changing the active code.
2758 Print executed "if" and "elsif" statements when debugging is used. This
2759 helps a great deal to understand what a filter does.
2761 Document more things not specified clearly in RFC3028. I had all this
2762 sorted out, when out of a sudden new issues came to my mind. Oops."
2764 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2767 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2768 canonical form (as documented). However, after a host name lookup from
2769 the IP address, check_host() was doing a simple string comparison with
2770 addresses acquired from the DNS when checking that the found name did
2771 have the original IP as one of its addresses. Since any found IPv6
2772 addresses are likely to be in abbreviated form, the comparison could
2773 fail. Luckily, there already exists a function for doing the comparison
2774 by converting both addresses to binary, so now that is used instead of
2775 the text comparison.
2777 PH/20 There was another similar case to PH/19, when a complete host name was
2778 given in a host list; looking up its IP address could give an abbreviated
2779 form, whereas the current host's name might or might not be abbreviated.
2780 The same fix has been applied.
2786 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2787 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2790 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2791 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2793 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2795 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2796 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2797 Also, for "Auto-Submitted", treat anything other than "no" as
2798 non-personal, in accordance with RFC 3834. (Previously it treated
2799 anything starting "auto-" as non-personal.)
2801 TF/01 The control=submission/name=... option had a problem with syntax
2802 errors if the name included a slash character. The /name= option
2803 now slurps the rest of the string, so it can include any characters
2804 but it must come last in the list of options (after /sender_retain
2807 PH/05 Some modifications to the interface to the fake nameserver for the new
2815 TK/01 Added the "success_on_redirect" address verification option. See
2816 NewStuff for rationale and an example.
2818 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2820 PH/02 Patch to exigrep to allow it to work on syslog lines.
2822 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2823 fread() to skip over the body file's header line, because in Cygwin the
2824 header line is locked and is inaccessible.
2826 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2827 co-exist for some time) to make it clear that it is the Exim ID that is
2828 referenced, not the Message-ID: header line.
2830 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2831 string_format() function, because snprintf() does not exist on all
2834 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2835 ${stat: expansion item.
2837 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2838 protocol synchronization error", to keep the pedants happy.
2840 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2841 well as for IRIX systems, when gcc is being used. See the host.c source
2844 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2846 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2849 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2850 transport and to the smtp transport in LMTP mode.
2852 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2854 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2855 run a filter in a subprocess. This could lead to confusion in subsequent
2856 lookups in the parent process. There should also be a search_tidyup() at
2857 the end of the subprocess.
2859 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2860 only if the host matched helo_try_verify_hosts, which caused the
2861 verification to occur when the EHLO/HELO command was issued. The ACL just
2862 tested the remembered result. Now, if a previous verification attempt has
2863 not happened, "verify = helo" does it there and then.
2865 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2867 TK/03 Fix log output including CR from clamd.
2869 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2870 exist provoked a memory error which could cause a segfault.
2872 PH/15 Installed PCRE 6.2
2874 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2876 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2877 of the problem. Specifically, suggested +O2 rather than +O1 for the
2880 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2882 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2883 record, the port was not correctly logged when the outgoing_port log
2884 selector was set (it logged the transort's default port).
2886 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2887 fallback_hosts, and "hosts" in the smtp transport.
2889 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2890 host errors such as "Connection refused".
2892 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2893 authentication with radiusclient 0.4.9:
2895 - Error returned from rc_read_config was caught wrongly
2896 - Username/password not passed on to radius server due to wrong length.
2898 The presumption is that some radiusclient API changes for 4.51/PH/17
2899 were not taken care of correctly. The code is still untested by me (my
2900 Linux distribution still has 0.3.2 of radiusclient), but it was
2901 contributed by a Radius user.
2903 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2904 expanding the "port" option of the smtp transport.
2906 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2907 while reading a MIME header. Thanks to Tom Hughes for a patch.
2909 PH/24 Include config.h inside local_scan.h so that configuration settings are
2912 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2913 that in an ACL for RCPT (for example), you can examine exactly what was
2916 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2917 commands, but it was not correctly comparing the address with the actual
2918 client host address. Thus, it would show the EHLO address in Received:
2919 header lines when this was not necessary.
2921 PH/27 Added the % operator to ${eval:}.
2923 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2924 it should be ignoring failures (because with -C, for example, it has lost
2925 privilege). It wasn't ignoring creation failures other than "already
2928 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2931 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2932 because an input line was too long, either on its own, or by virtue of
2933 too many continuations, the temporary file was not being removed, and the
2934 return code was incorrect.
2936 PH/31 Missing "BOOL" in function definition in filtertest.c.
2938 PH/32 Applied Sieve patches from the maintainer.
2940 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2942 PH/33 Added "verify = not_blind".
2944 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2945 Local/Makefile (with some defaults set). These are used in built scripts
2946 such as exicyclog, but they have never been used in the exim_install
2947 script (though there are many overriding facilities there). I have
2948 arranged that the exim_install script now takes note of these two
2951 PH/35 Installed configuration files for Dragonfly.
2953 PH/36 When a locally submitted message by a trusted user did not contain a
2954 From: header, and the sender address was obtained from -f or from an SMTP
2955 MAIL command, and the trusted user did not use -F to supply a sender
2956 name, $originator_name was incorrectly used when constructing a From:
2957 header. Furthermore, $originator_name was used for submission mode
2958 messages from external hosts without From: headers in a similar way,
2959 which is clearly wrong.
2961 PH/37 Added control=suppress_local_fixups.
2963 PH/38 When log_selector = +received_sender was set, and the addition of the
2964 sender made the log line's construction buffer exactly full, or one byte
2965 less than full, an overflow happened when the terminating "\n" was
2968 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2969 when the result of a list match is failure because a DNS lookup failed.
2971 PH/40 RM_COMMAND is now used in the building process.
2973 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2974 the "build-* directories that it finds.
2976 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2977 address in a domain literal was a prefix of an interface address.
2979 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2980 when verifying a sender address, unless rewrite_headers is false.
2982 PH/44 Wrote a long comment about why errors_to addresses are verified as
2983 recipients, not senders.
2985 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2986 the ratelimit ACL was added.
2988 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2990 PH/46 Added extra information about PostgreSQL errors to the error string.
2992 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2993 avoiding the need to install special zones in a real server. This is
2994 backwards compatible; if it can't find the fake resolver, it drops back.
2995 Thus, both old and new test suites can be run.
2997 TF/02 Added util/ratelimit.pl
2999 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3000 clock is set back in time.
3002 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3003 Candler <B.Candler@pobox.com>.
3005 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3006 OK for addresses that are the result of redirections.
3008 PH/48 A number of further additions for the benefit of the new test suite,
3009 including a fake gethostbyname() that interfaces to the fake DNS resolver
3012 TF/06 The fix for widen_domains has also been applied to qualify_single and
3013 search_parents which are the other dnslookup options that can cause
3016 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3019 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3021 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3022 service" if it encountered an MX record with an empty target hostname.
3023 The message is now "an MX or SRV record indicated no SMTP service".
3025 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3026 if the DNS of the sending site is misconfigured. This is quite a
3027 common situation. This change restores the behaviour of treating a
3028 helo verification defer as a failure.
3030 PH/51 If self=fail was set on a router, the bounce message did not include the
3031 actual error message.
3037 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3039 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3040 command itself ended in error, the underlying message about the transport
3041 filter timeout was being overwritten with the pipe command error. Now the
3042 underlying error message should be appended to the second error message.
3044 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3046 PH/02 Reduce the amount of output that "make" produces by default. Full output
3047 can still be requested.
3049 PH/03 The warning log line about a condition test deferring for a "warn" verb
3050 was being output only once per connection, rather than after each
3051 occurrence (because it was using the same function as for successful
3052 "warn" verbs). This seems wrong, so I have changed it.
3054 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3055 it should not have, which might have caused a crash in the right
3056 circumstances, but probably never did.
3058 PH/04 Installed a modified version of Tony Finch's patch to make submission
3059 mode fix the return path as well as the Sender: header line, and to
3060 add a /name= option so that you can make the user's friendly name appear
3063 TF/03 Added the control = fakedefer ACL modifier.
3065 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3066 Mark Lowes for thorough testing.
3068 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3070 TK/03 Merged latest SRS patch from Miles Wilton.
3072 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3073 with the definition in sysexits.h (which is #included earlier).
3074 Fortunately, Exim does not actually use EX_OK. The code used to try to
3075 preserve the sysexits.h value, by assumimg that macro definitions were
3076 scanned for macro replacements. I have been disabused of this notion,
3077 so now the code just undefines EX_OK before #including unistd.h.
3079 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3080 in the smtp transport. When a block could not be written in a single
3081 write() function, the timeout was being re-applied to each part-write.
3082 This seems wrong - if the receiver was accepting one byte at a time it
3083 would take for ever. The timeout is now adjusted when this happens. It
3084 doesn't have to be particularly precise.
3086 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3087 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3089 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3090 without a domain if the check to <postmaster@domain> fails.
3092 SC/01 Eximstats: added -xls and the ability to specify output files
3093 (patch written by Frank Heydlauf).
3095 SC/02 Eximstats: use FileHandles for outputing results.
3097 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3099 SC/04 Eximstats: fixed display of large numbers with -nvr option
3101 SC/05 Eximstats: fixed merging of reports with empty tables.
3103 SC/06 Eximstats: added the -include_original_destination flag
3105 SC/07 Eximstats: removed tabs and trailing whitespace.
3107 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3109 TK/06 MBOX spool code: Add real "From " MBOX separator line
3110 so the .eml file is really in mbox format (even though
3111 most programs do not really care). Patch from Alex Miller.
3113 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3114 The latter is generated from $received_to and is only set if the
3115 message has one envelope recipient. SA can use these headers,
3116 obviously out-of-the-box. Patch from Alex Miller.
3118 PH/08 The ${def test on a variable was returning false if the variable's
3119 value was "0", contrary to what the specification has always said!
3120 The result should be true unless the variable is empty.
3122 PH/09 The syntax error of a character other than { following "${if
3123 def:variable_name" (after optional whitespace) was not being diagnosed.
3124 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3125 accidental colon was present, for example, could give incorrect results.
3127 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3128 result is used (not including appendfile, where other changes are about
3131 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3132 This involved changing a lot of size variables from int to off_t. It
3133 should work with maildirs and everything.
3135 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3136 spamd dying while we are connected to it.
3138 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3141 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3142 function for BDB 4.3.
3144 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3146 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3147 the sources for PCRE have been re-arranged and more files are now
3150 PH/15 The code I had for printing potentially long long variables in PH/11
3151 above was not the best (it lost precision). The length of off_t variables
3152 is now inspected at build time, and an appropriate printing format (%ld
3153 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3154 to be "long long int" or "long int". This is needed for the internal
3155 formatting function string_vformat().
3157 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3158 the configuration file to be ":syslog", then the script "guesses" where
3159 the logs files are, rather than using the compiled in default. In our
3160 case the guess is not the same as the compiled default, so the script
3161 suddenly stopped working when I started to use syslog. The patch checks
3162 to see if log_file_path is "". If so, it attempts to read it from exim
3163 with no configuration file to get the compiled in version, before it
3164 falls back to the previous guessing code."
3166 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3167 implementing BATV in an Exim configuration. See NewStuff for the gory
3170 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3171 Makefile that are specific to HP-UX.
3173 PH/18 If the "use_postmaster" option was set for a recipient callout together
3174 with the "random" option, the postmaster address was used as the MAIL
3175 FROM address for the random test, but not for the subsequent recipient
3176 test. It is now used for both.
3178 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3179 patch removes a few documentation additions to RFC 3028, because the
3180 latest draft now contains them. It adds the new en;ascii-case comparator
3181 and a new error check for 8bit text in MIME parts. Comparator and
3182 require names are now matched exactly. I enabled the subaddress
3183 extension, but it is not well tested yet (read: it works for me)."
3185 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3186 rework some of the code of TK/09 above to avoid the hardwired use of
3187 "%lld" and "long long". Replaced the call to snprintf() with a call to
3190 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3191 records point to non-existent hosts", "retry timeout exceeded", and
3192 "retry time not reached for any host after a long failure period".
3194 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3195 experimental DomainKeys support:
3197 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3198 (2) On an error such as an illegally used "control", the wrong name for
3199 the control was given.
3201 These problems did NOT occur unless DomainKeys support was compiled.
3203 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3205 PH/24 Added ${if match_ip condition.
3207 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3208 databases so that it will be absolutely obvious if a crash occurs in the
3209 DB library. This is a regular occurrence (often caused by mis-matched
3212 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3213 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3214 picked up on a user's system that detects such things. There doesn't seem
3215 to be a gcc warning option for this - only an attribute that has to be
3216 put on the function's prototype. It seems that in Fedora Core 4 they have
3217 set this on a number of new functions. No doubt there will be more in due
3220 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3221 specify a transport. However, if an address that was verified by such a
3222 router was the subject of a callout, Exim crashed because it tried to
3223 read the rcpt_include_affixes from the non-existent transport. Now it
3224 just assumes that the setting of that option is false. This bug was
3225 introduced by 4.51/PH/31.
3227 PH/28 Changed -d+all to exclude +memory, because that information is very
3228 rarely of interest, but it makes the output a lot bigger. People tend to
3229 do -d+all out of habit.
3231 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3232 code in os-type was giving problems when libc.so lives in lib64, like on
3235 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3236 aren't the modern standard, and it seems that some systems' include files
3237 don't always have them. Exim was already checking for some of the newer
3238 ones like T_AAAA, and defining it itself. I've added checks for all the
3239 record types that Exim uses.
3241 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3242 not automatically generating a new one, as it is supposed to. This
3243 prevented TLS from working. If the file did exist, but contained invalid
3244 data, a new version was generated, as expected. It was only the case of a
3245 non-existent file that was broken.
3247 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3248 with a change in libdomainkeys > 0.64.
3250 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3251 from DNS. If the selector record carries the flag, it now has
3252 precedence over the domain-wide flag.
3254 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3256 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3257 the use of a port name that isn't defined in /etc/services) occurred, the
3258 message was deferred as in a normal delivery, and thus remained on the
3259 spool, instead of being failed because of the mua_wrapper setting. This
3260 is now fixed, and I tidied up some of the mua_wrapper messages at the
3263 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3264 the messages in a hash of arrays rather than using individual hashes.
3265 This is a bit cleaner and results in dramatic memory savings, albeit
3266 at a slight CPU cost.
3268 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3269 as requested by Marc Sherman.
3271 SC/10 Eximstats: added histograms for user specified patterns as requested
3274 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3276 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3277 fopen() in the content-scanning modules that did not already have it.
3283 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3284 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3286 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3288 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3290 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3291 to test Sieve filters that use "vacation".
3293 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3294 that changes the way the GnuTLS parameters are stored in the cache file.
3295 The new format can be generated externally. For backward compatibility,
3296 if the data in the cache doesn't make sense, Exim assumes it has read an
3297 old-format file, and it generates new data and writes a new file. This
3298 means that you can't go back to an older release without removing the
3301 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3302 work if there are any delivery delays because "one_time" forces the
3303 parent to be marked "delivered", so its unseen clone is never tried
3304 again. For this reason, Exim now forbids the simultaneous setting of
3307 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3308 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3309 entry if you want to know the details. The fix, however, neglected to
3310 consider the case when local delivery batching is involved. The test for
3311 "previously delivered" was not happening when checking to see if an
3312 address could be batched with a previous (undelivered) one; under
3313 certain circumstances this could lead to multiple deliveries to the same
3316 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3317 in its include files, and this causes problems building Exim.
3319 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3320 header_syntax) but Exim was just ignoring anything given after a slash.
3321 In particular, this caused confusion with an attempt to use "verify =
3322 reverse_host_lookup/defer_ok". An error is now given when options are
3323 supplied for verify items that do not have them. (Maybe reverse_host_
3324 lookup should have a defer_ok option, but that's a different point.)
3326 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3327 defined by RFC 821) to 2048, because there were problems with some AUTH
3328 commands, and RFC 1869 says the size should be increased for extended
3329 SMTP commands that take arguments.
3331 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3334 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3335 "unknown" error; now it says that the functionality isn't in the binary.
3337 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3338 an address' error message when a string expansion fails (syntax or
3339 whatever). Otherwise the password may appear in the log. Following change
3340 PH/42 below, there is no longer a chance of it appearing in a bounce
3343 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3345 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3346 of its addresses was ever tried. (Bugzilla bug #2.)
3348 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3349 the result incorrectly in the debug output. (It correctly added a newline
3350 to what was transported.)
3352 TF/01 Added $received_time.
3354 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3355 commented out examples of how to interface to a virus scanner and to
3356 SpamAssassin. Also added commented examples of av_scanner and
3357 spamd_address settings.
3359 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3360 and controls are allowed in which ACLs. There were a couple of minor
3361 errors. Some of the entries in the conditions table (which is a table of
3362 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3363 negation of where the condition IS allowed.
3365 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3367 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3368 header file does not have a version number, so I've had to invent a new
3369 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3370 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3371 radiusclient), but it was contributed by a Radius user.
3373 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3374 files or directories, for OpenSSL.
3376 PH/19 When an Exim process that is running as root has to create an Exim log
3377 file, it does so in a subprocess that runs as exim:exim so as to get the
3378 ownership right at creation (otherwise, other Exim processes might see
3379 the file with the wrong ownership). There was no test for failure of this
3380 fork() call, which would lead to the process getting stuck as it waited
3381 for a non-existent subprocess. Forks do occasionally fail when resources
3382 run out. I reviewed all the other calls to fork(); they all seem to check
3385 PH/20 When checking for unexpected SMTP input at connect time (before writing
3386 the banner), Exim was not dealing correctly with a non-positive return
3387 from the read() function. If the client had disconnected by this time,
3388 the result was a log entry for a synchronization error with an empty
3389 string after "input=" when read() returned zero. If read() returned -1
3390 (an event I could not check), uninitialized data bytes were printed.
3391 There were reports of junk text (parts of files, etc) appearing after
3394 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3396 PH/22 Added support for macro redefinition, and (re)definition in between
3397 driver and ACL definitions.
3399 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3400 forgetting to use the resulting value; it was using the unexpanded value.
3402 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3403 hadn't been configured. The fix is from Juergen Kreileder, who
3404 understands it better than I do:
3406 "Here's what I see happening with three configured cyrus_sasl
3407 authenticators configured (plain, login, cram-md5):
3409 On startup auth_cyrus_sasl_init() gets called for each of these.
3410 This means three calls to sasl_listmech() without a specified mech_list.
3411 => SASL tests which mechs of all available mechs actually work
3412 => three warnings about OTP not working
3413 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3415 With the patch, sasl_listmech() also gets called three times. But now
3416 SASL's mech_list option is set to the server_mech specified in the the
3417 authenticator. Or in other words, the answer from sasl_listmech()
3418 gets limited to just the mech you're testing for (which is different
3420 => the return list contains just 'plain' or 'login', 'cram-md5' or
3421 nothing depending on the value of ob->server_mech.
3423 I've just tested the patch: Authentication still works fine,
3424 unavailable mechs specified in the exim configuration are still
3425 caught, and the auth.log warnings about OTP are gone."
3427 PH/25 When debugging is enabled, the contents of the command line are added
3428 to the debugging output, even when log_selector=+arguments is not
3431 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3432 answer is "GNU", and only if the return is "GNU/something" is the answer
3435 PH/27 $acl_verify_message is now set immediately after the failure of a
3436 verification in an ACL, and so is available in subsequent modifiers. In
3437 particular, the message can be preserved by coding like this:
3439 warn !verify = sender
3440 set acl_m0 = $acl_verify_message
3442 Previously, $acl_verify_message was set only while expanding "message"
3443 and "log_message" when a very denied access.
3445 PH/28 Modified OS/os.c-Linux with
3447 -#ifndef OS_LOAD_AVERAGE
3448 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3450 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3451 nomenclature these days.)
3453 PH/29 Installed patch from the Sieve maintainer that adds the options
3454 sieve_useraddress and sieve_subaddress to the redirect router.
3456 PH/30 In these circumstances:
3457 . Two addresses routed to the same list of hosts;
3458 . First host does not offer TLS;
3459 . First host accepts first address;
3460 . First host gives temporary error to second address;
3461 . Second host offers TLS and a TLS session is established;
3462 . Second host accepts second address.
3463 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3464 and peerdn, if requested) that were in fact used only for the second
3467 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3468 attention to any local part prefix or suffix that was matched by the
3469 router that accepted the address. It now behaves in the same way as it
3470 does for delivery: the affixes are removed from the local part unless
3471 rcpt_include_affixes is set on the transport.
3473 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3474 timeout during the DATA phase of an incoming message.
3476 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3477 applied a patch sanctioned by the Sieve maintainer.
3479 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3480 the uid or gid is negative. A case of a negative gid caused this to be
3481 noticed. The fix allows for either to be negative.
3483 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3484 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3487 PH/36 The change PH/12 above was broken. Fixed it.
3489 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3490 the grounds that routing the same address twice would always produce the
3491 same answer. This might have been true once, but it is certainly no
3492 longer true now. Routing a child address may depend on the previous
3493 routing that produced that child. Some complicated redirection strategies
3494 went wrong when messages had multiple recipients, and made Exim's
3495 behaviour dependent on the order in which the addresses were given.
3497 I have moved the duplicate checking until after the routing is complete.
3498 Exim scans the addresses that are assigned to local and remote
3499 transports, and removes any duplicates. This means that more work will be
3500 done, as duplicates will always all be routed, but duplicates are
3501 presumably rare, so I don't expect this is of any significance.
3503 For deliveries to pipes, files, and autoreplies, the duplicate checking
3504 still happens during the routing process, since they are not going to be
3507 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3508 It corrects a timeout issue with spamd. This is Ian's comment: "The
3509 background is that sometimes spamd either never reads data from a
3510 connection it has accepted, or it never writes response data. The exiscan
3511 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3512 blindly assumes that writes won't block so it may never time out."
3514 PH/39 Allow G after quota size as well as K and M.
3516 PH/40 The value set for $authenticated_id in an authenticator may not contain
3517 binary zeroes or newlines because the value is written to log lines and
3518 to spool files. There was no check on this. Now the value is run through
3519 the string_printing() function so that such characters are converted to
3520 printable escape sequences.
3522 PH/41 $message_linecount is a new variable that contains the total number of
3523 lines in the message. Compare $body_linecount, which is the count for the
3526 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3527 bounce and delay warning messages, except in certain special cases, which
3530 (a) An SMTP error message from a remote host;
3531 (b) A message specified in a :fail: redirection;
3532 (c) A message specified in a "fail" command in a system filter;
3533 (d) A message specified in a FAIL return from the queryprogram router;
3534 (e) A message specified by the cannot_route_message router option.
3536 In these cases only, Exim does include the error details in bounce and
3537 warning messages. There are also a few cases where bland messages such
3538 as "unrouteable address" or "local delivery error" are given.
3540 PH/43 $value is now also set for the "else" part of a ${run expansion.
3542 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3543 being worked on, but at least Exim now implements the latest version to
3546 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3547 process to complete was treated as a delivery failure, a timeout while
3548 writing the message to the pipe was logged, but erroneously treated as a
3549 successful delivery. Such timeouts include transport filter timeouts. For
3550 consistency with the overall process timeout, these timeouts are now
3551 treated as errors, giving rise to delivery failures by default. However,
3552 there is now a new Boolean option for the pipe transport called
3553 timeout_defer, which, if set TRUE, converts the failures into defers for
3554 both kinds of timeout. A transport filter timeout is now identified in
3557 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3558 systems where "make" and "gmake" are different, calling "gmake" at top
3559 level broke things. I've arranged for the value of $(MAKE) to be passed
3560 from the Makefile to this script so that it can call the same version of
3564 A note about Exim versions 4.44 and 4.50
3565 ----------------------------------------
3567 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3568 changes of various kinds. As a consequence, a big documentation update was
3569 needed. This delayed the release for rather longer than seemed good, especially
3570 in the light of a couple of (minor) security issues. Therefore, the changes
3571 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3572 release. So 4.44 and 4.50 are in effect two different branches that both start
3575 I have left the 4.50 change log unchanged; it contains all the changes since
3576 4.43. The change log for 4.44 is below; many of its items are identical to
3577 those for 4.50. This seems to be the most sensible way to preserve the
3578 historical information.
3584 1. Minor wording change to the doc/README.SIEVE file.
3586 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3587 computation of the current number of files was incorrect.
3589 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3590 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3593 4. Give more explanation in the error message when the command for a transport
3594 filter fails to execute.
3596 5. There are several places where Exim runs a non-Exim command in a
3597 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3598 was being done only for the command run by the queryprogram router. It is
3599 now done for all such subprocesses. The other cases are: ${run, transport
3600 filters, and the commands run by the lmtp and pipe transports.
3602 6. Added CONFIGURE_GROUP build-time option.
3604 7. Some older OS have a limit of 256 on the maximum number of file
3605 descriptors. Exim was using setrlimit() to set 1000 as a large value
3606 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3607 systems. I've change it so that if it can't get 1000, it tries for 256.
3609 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3610 was an oversight, and furthermore, ever since the addition of extra
3611 controls (e.g. 4.43/32), the checks on when to allow different forms of
3612 "control" were broken. There should now be diagnostics for all cases when a
3613 control that does not make sense is encountered.
3615 9. Added the /retain_sender option to "control=submission".
3617 10. $recipients is now available in the predata ACL (oversight).
3619 11. Tidy the search cache before the fork to do a delivery from a message
3620 received from the command line. Otherwise the child will trigger a lookup
3621 failure and thereby defer the delivery if it tries to use (for example) a
3622 cached ldap connection that the parent has called unbind on.
3624 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3625 of $address_data from the recipient verification was clobbered by the
3626 sender verification.
3628 13. The value of address_data from a sender verification is now available in
3629 $sender_address_data in subsequent conditions in the ACL statement.
3631 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3633 15. Added a new option "connect=<time>" to callout options, to set a different
3636 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3637 was its contents. (It was OK if the option was not defined at all.)
3639 17. A "Completed" log line is now written for messages that are removed from
3640 the spool by the -Mrm option.
3642 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3643 information about exactly what failed.
3645 19. Added -dd to debug only the daemon process.
3647 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3648 handles timeouts, both on the server side and network timeouts. Renamed the
3649 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3651 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3652 It is now set to "smtps".
3654 22. $host_address is now set to the target address during the checking of
3655 ignore_target_hosts.
3657 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3658 being passed; this would have caused $sender_host_name to have been used if
3659 matching the list had actually called for a host name (not very likely,
3660 since this list is usually IP addresses). A host name is now passed as
3663 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3664 code that specifies a non-restarting handler (typically sigaction() in
3665 modern systems) in an attempt to fix a rare and obscure crash bug.
3667 25. Narrowed the window for a race in the daemon that could cause it to ignore
3668 SIGCHLD signals. This is not a major problem, because they are used only to
3669 wake it up if nothing else does.
3671 26. A malformed maildirsize file could cause Exim to calculate negative values
3672 for the mailbox size or file count. Odd effects could occur as a result.
3673 The maildirsize information is now recalculated if the size or filecount
3676 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3677 support for a long time. Removed HAVE_SYS_VFS_H.
3679 28. Installed the latest version of exipick from John Jetmore.
3681 29. In an address list, if the pattern was not a regular expression, an empty
3682 subject address (from a bounce message) matched only if the pattern was an
3683 empty string. Non-empty patterns were not even tested. This was the wrong
3684 because it is perfectly reasonable to use an empty address as part of a
3685 database query. An empty address is now tested by patterns that are
3686 lookups. However, all the other forms of pattern expect the subject to
3687 contain a local part and a domain, and therefore, for them, an empty
3688 address still always fails if the pattern is not itself empty.
3690 30. Exim went into a mad DNS loop when attempting to do a callout where the
3691 host was specified on an smtp transport, and looking it up yielded more
3692 than one IP address.
3694 31. Re-factored the code for checking spool and log partition space into a
3695 function that finds that data and another that does the check. The former
3696 is then used to implement four new variables: $spool_space, $log_space,
3697 $spool_inodes, and $log_inodes.
3699 32. The RFC2047 encoding function was originally intended for short strings
3700 such as real names; it was not keeping to the 75-character limit for
3701 encoded words that the RFC imposes. It now respects the limit, and
3702 generates multiple encoded words if necessary. To be on the safe side, I
3703 have increased the buffer size for the ${rfc2047: expansion operator from
3706 33. It is now permitted to omit both strings after an "if" condition; if the
3707 condition is true, the result is "true". As before, when the second string
3708 is omitted, a false condition yields an empty string. This makes it less
3709 cumbersome to write custom ACL and router conditions.
3711 34. Failure to deliver a bounce message always caused it to be frozen, even if
3712 there was an errors_to setting on the router. The errors_to setting is now
3715 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3716 canonical form (fully expanded) before being placed in
3717 $sender_host_address.
3719 36. The table in the code that translates DNS record types into text (T_A to
3720 "A" for instance) was missing entries for NS and CNAME. It is just possible
3721 that this could have caused confusion if both these types were looked up
3722 for the same domain, because the text type is used as part of Exim's
3723 per-process caching. But the chance of anyone hitting this buglet seems
3726 37. The dnsdb lookup has been extended in a number of ways.
3728 (1) There is a new type, "zns", which walks up the domain tree until it
3729 finds some nameserver records. It should be used with care.
3731 (2) There is a new type, "mxh", which is like "mx" except that it returns
3732 just the host names, not the priorities.
3734 (3) It is now possible to give a list of domains (or IP addresses) to be
3735 looked up. The behaviour when one of the lookups defers can be
3736 controlled by a keyword.
3738 (4) It is now possible to specify the separator character for use when
3739 multiple records are returned.
3741 38. The dnslists ACL condition has been extended: it is now possible to supply
3742 a list of IP addresses and/or domains to be looked up in a particular DNS
3745 39. Added log_selector=+queue_time_overall.
3747 40. When running the queue in the test harness, wait just a tad after forking a
3748 delivery process, to get repeatability of debugging output.
3750 41. Include certificate and key file names in error message when GnuTLS fails
3751 to set them up, because the GnuTLS error message doesn't include the name
3752 of the failing file when there is a problem reading it.
3754 42. Allow both -bf and -bF in the same test run.
3756 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3758 44. The "Exiscan patch" is now merged into the mainline Exim source.
3760 45. Sometimes the final signoff response after QUIT could fail to get
3761 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3762 < 0 before doing a fflush(). This bug looks as though it goes back to the
3763 introduction of TLS in release 3.20, but "sometimes" must have been rare
3764 because the tests only now provoked it.
3766 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3767 (this can affect the format of dates).
3769 47. exim_tidydb, when checking for the continued existence of a message for
3770 which it has found a message-specific retry record, was not finding
3771 messages that were in split spool directories. Consequently, it was
3772 deleting retry records that should have stayed in existence.
3774 48. Steve fixed some bugs in eximstats.
3776 49. The SPA authentication driver was not abandoning authentication and moving
3777 on to the next authenticator when an expansion was forced to fail,
3778 contradicting the general specification for all authenticators. Instead it
3779 was generating a temporary error. It now behaves as specified.
3781 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3782 (the order specifies the preference for clients). The order is now AES256,
3783 AES128, 3DES, ARCFOUR128.
3785 51. Small patch to Sieve code - explicitly set From: when generating an
3788 52. Exim crashed if a remote delivery caused a very long error message to be
3789 recorded - for instance if somebody sent an entire SpamAssassin report back
3790 as a large number of 550 error lines. This bug was coincidentally fixed by
3791 increasing the size of one of Exim's internal buffers (big_buffer) that
3792 happened as part of the Exiscan merge. However, to be on the safe side, I
3793 have made the code more robust (and fixed the comments that describe what
3796 53. Now that there can be additional text after "Completed" in log lines (if
3797 the queue_time_overall log selector is set), a one-byte patch to exigrep
3798 was needed to allow it to recognize "Completed" as not the last thing in
3801 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3802 patch that reportedly fixes this has been added. I am not expert enough to
3803 create a test for it. This is what the patch creator wrote:
3805 "I found a little strange behaviour of ldap code when working with
3806 Windows 2003 AD Domain, where users was placed in more than one
3807 Organization Units. When I tried to give exim partial DN, the exit code
3808 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3809 But simultaneously result of request was absolutely normal ldap result,
3810 so I produce this patch..."
3812 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3813 REFERENCE, so I have modified the code to exclude the patch when that macro
3816 55. Some experimental protocols are using DNS PTR records for new purposes. The
3817 keys for these records are domain names, not reversed IP addresses. The
3818 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3819 leaves it alone. Component reversal etc. now happens only for IP addresses.
3822 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3824 57. Double the size of the debug message buffer (to 2048) so that more of very
3825 long debug lines gets shown.
3827 58. The exicyclog utility now does better if the number of log files to keep
3828 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3830 59. Two changes related to the smtp_active_hostname option:
3832 (1) $smtp_active_hostname is now available as a variable.
3833 (2) The default for smtp_banner uses $smtp_active_hostname instead
3834 of $primary_hostname.
3836 60. The host_aton() function is supposed to be passed a string that is known
3837 to be a valid IP address. However, in the case of IPv6 addresses, it was
3838 not checking this. This is a hostage to fortune. Exim now panics and dies
3839 if the condition is not met. A case was found where this could be provoked
3840 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3841 components; fortuitously, this particular loophole had already been fixed
3842 by change 4.50/55 above.
3844 If there are any other similar loopholes, the new check in host_aton()
3845 itself should stop them being exploited. The report I received stated that
3846 data on the command line could provoke the exploit when Exim was running as
3847 exim, but did not say which command line option was involved. All I could
3848 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3849 running as the user.
3852 61. There was a buffer overflow vulnerability in the SPA authentication code
3853 (which came originally from the Samba project). I have added a test to the
3854 spa_base64_to_bits() function which I hope fixes it.
3857 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3858 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3860 63. The daemon start-up calls getloadavg() while still root for those OS that
3861 need the first call to be done as root, but it missed one case: when
3862 deliver_queue_load_max is set with deliver_drop_privilege. This is
3863 necessary for the benefit of the queue runner, because there is no re-exec
3864 when deliver_drop_privilege is set.
3866 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3867 This has been fixed.
3869 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3870 was in use, was not putting the data itself into the right store pool;
3871 consequently, it could be overwritten for a subsequent message in the same
3872 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3875 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3877 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3878 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3879 the function were treating the return as a boolean value, which happened to
3880 work because 0=false and not-0=true, but is not correct code.
3882 68. The host_aton() function was not handling scoped IPv6 addresses (those
3883 with, for example, "%eth0" on the end) correctly.
3885 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3886 negated items (that is, ~something) in unsigned ints. Some compilers
3887 apparently mutter when there is no cast.
3889 70. If an address verification called from an ACL failed, and did not produce a
3890 user-specific message (i.e. there was only a "system" message), nothing was
3891 put in $acl_verify_message. In this situation, it now puts the system
3894 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3895 session; change 4.31/43 added the unwanted input to the log line - except
3896 that it did not do this in the start of session case. It now does.
3898 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3899 This is wrong and can cause the other end to generate a synchronization
3900 error if it is another Exim or anything else that does the synchronization
3901 check. A QUIT command is no longer sent after a timeout.
3903 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3904 during host lookups.
3906 74. The defer_ok option of callout verification was not working if it was used
3907 when verifying addresses in header lines, that is, for this case:
3909 verify = header_sender/callout=defer_ok
3911 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3912 those file descriptors could be used for SMTP connections. If anything
3913 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3914 could be sent to the SMTP client, causing chaos. The daemon now opens
3915 stdin, stdout, and stderr to /dev/null when it puts itself into the
3918 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3919 log by default. The user can override this with suitable Perl magic.
3921 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3922 the log message when discard triggers, was not working for the DATA ACL or
3923 for the non-SMTP ACL.
3925 78. Error message wording change in sieve.c.
3927 79. If smtp_accept_max_per_host was set, the number of connections could be
3928 restricted to fewer than expected, because the daemon was trying to set up
3929 a new connection before checking whether the processes handling previous
3930 connections had finished. The check for completed processes is now done
3931 earlier. On busy systems, this bug wouldn't be noticed because something
3932 else would have woken the daemon, and it would have reaped the completed
3935 80. If a message was submitted locally by a user whose login name contained one
3936 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3937 It caused a spool format error. I have fixed the spool reading code. A
3938 related problem was that the "from" clause in the Received: line became
3939 illegal because of the space(s). It is now covered by ${quote_local_part.
3941 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3944 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3946 83. Patch from Sieve maintainer to fix unterminated string problem in
3947 "vacation" handling.
3949 84. Some minor changes to the Linux configuration files to help with other
3950 OS variants using glibc.
3952 85. One more patch for Sieve to update vacation handling to latest spec.
3955 ----------------------------------------------------
3956 See the note above about the 4.44 and 4.50 releases.
3957 ----------------------------------------------------
3963 1. Change 4.43/35 introduced a bug that caused file counts to be
3964 incorrectly computed when quota_filecount was set in an appendfile
3967 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3968 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3971 3. Give more explanation in the error message when the command for a transport
3972 filter fails to execute.
3974 4. There are several places where Exim runs a non-Exim command in a
3975 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3976 was being done only for the command run by the queryprogram router. It is
3977 now done for all such subprocesses. The other cases are: ${run, transport
3978 filters, and the commands run by the lmtp and pipe transports.
3980 5. Some older OS have a limit of 256 on the maximum number of file
3981 descriptors. Exim was using setrlimit() to set 1000 as a large value
3982 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3983 systems. I've change it so that if it can't get 1000, it tries for 256.
3985 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3986 was an oversight, and furthermore, ever since the addition of extra
3987 controls (e.g. 4.43/32), the checks on when to allow different forms of
3988 "control" were broken. There should now be diagnostics for all cases when a
3989 control that does not make sense is encountered.
3991 7. $recipients is now available in the predata ACL (oversight).
3993 8. Tidy the search cache before the fork to do a delivery from a message
3994 received from the command line. Otherwise the child will trigger a lookup
3995 failure and thereby defer the delivery if it tries to use (for example) a
3996 cached ldap connection that the parent has called unbind on.
3998 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3999 of $address_data from the recipient verification was clobbered by the
4000 sender verification.
4002 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4003 was its contents. (It was OK if the option was not defined at all.)
4005 11. A "Completed" log line is now written for messages that are removed from
4006 the spool by the -Mrm option.
4008 12. $host_address is now set to the target address during the checking of
4009 ignore_target_hosts.
4011 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4012 being passed; this would have caused $sender_host_name to have been used if
4013 matching the list had actually called for a host name (not very likely,
4014 since this list is usually IP addresses). A host name is now passed as
4017 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4018 code that specifies a non-restarting handler (typically sigaction() in
4019 modern systems) in an attempt to fix a rare and obscure crash bug.
4021 15. Narrowed the window for a race in the daemon that could cause it to ignore
4022 SIGCHLD signals. This is not a major problem, because they are used only to
4023 wake it up if nothing else does.
4025 16. A malformed maildirsize file could cause Exim to calculate negative values
4026 for the mailbox size or file count. Odd effects could occur as a result.
4027 The maildirsize information is now recalculated if the size or filecount
4030 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4031 support for a long time. Removed HAVE_SYS_VFS_H.
4033 18. Updated exipick to current release from John Jetmore.
4035 19. Allow an empty sender to be matched against a lookup in an address list.
4036 Previously the only cases considered were a regular expression, or an
4039 20. Exim went into a mad DNS lookup loop when doing a callout where the
4040 host was specified on the transport, if the DNS lookup yielded more than
4043 21. The RFC2047 encoding function was originally intended for short strings
4044 such as real names; it was not keeping to the 75-character limit for
4045 encoded words that the RFC imposes. It now respects the limit, and
4046 generates multiple encoded words if necessary. To be on the safe side, I
4047 have increased the buffer size for the ${rfc2047: expansion operator from
4050 22. Failure to deliver a bounce message always caused it to be frozen, even if
4051 there was an errors_to setting on the router. The errors_to setting is now
4054 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4055 canonical form (fully expanded) before being placed in
4056 $sender_host_address.
4058 24. Updated eximstats to version 1.33
4060 25. Include certificate and key file names in error message when GnuTLS fails
4061 to set them up, because the GnuTLS error message doesn't include the name
4062 of the failing file when there is a problem reading it.
4064 26. Expand error message when OpenSSL has problems setting up cert/key files.
4067 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4068 (this can affect the format of dates).
4070 28. exim_tidydb, when checking for the continued existence of a message for
4071 which it has found a message-specific retry record, was not finding
4072 messages that were in split spool directories. Consequently, it was
4073 deleting retry records that should have stayed in existence.
4075 29. eximstats updated to version 1.35
4076 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4077 1.35 - bugfix such that pie charts by volume are generated correctly
4079 30. The SPA authentication driver was not abandoning authentication and moving
4080 on to the next authenticator when an expansion was forced to fail,
4081 contradicting the general specification for all authenticators. Instead it
4082 was generating a temporary error. It now behaves as specified.
4084 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4085 (the order specifies the preference for clients). The order is now AES256,
4086 AES128, 3DES, ARCFOUR128.
4088 31. Small patch to Sieve code - explicitly set From: when generating an
4091 32. Exim crashed if a remote delivery caused a very long error message to be
4092 recorded - for instance if somebody sent an entire SpamAssassin report back
4093 as a large number of 550 error lines. This bug was coincidentally fixed by
4094 increasing the size of one of Exim's internal buffers (big_buffer) that
4095 happened as part of the Exiscan merge. However, to be on the safe side, I
4096 have made the code more robust (and fixed the comments that describe what
4099 33. Some experimental protocols are using DNS PTR records for new purposes. The
4100 keys for these records are domain names, not reversed IP addresses. The
4101 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4102 leaves it alone. Component reversal etc. now happens only for IP addresses.
4105 34. The host_aton() function is supposed to be passed a string that is known
4106 to be a valid IP address. However, in the case of IPv6 addresses, it was
4107 not checking this. This is a hostage to fortune. Exim now panics and dies
4108 if the condition is not met. A case was found where this could be provoked
4109 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4110 components; fortuitously, this particular loophole had already been fixed
4111 by change 4.50/55 or 4.44/33 above.
4113 If there are any other similar loopholes, the new check in host_aton()
4114 itself should stop them being exploited. The report I received stated that
4115 data on the command line could provoke the exploit when Exim was running as
4116 exim, but did not say which command line option was involved. All I could
4117 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4118 running as the user.
4121 35. There was a buffer overflow vulnerability in the SPA authentication code
4122 (which came originally from the Samba project). I have added a test to the
4123 spa_base64_to_bits() function which I hope fixes it.
4126 36. The daemon start-up calls getloadavg() while still root for those OS that
4127 need the first call to be done as root, but it missed one case: when
4128 deliver_queue_load_max is set with deliver_drop_privilege. This is
4129 necessary for the benefit of the queue runner, because there is no re-exec
4130 when deliver_drop_privilege is set.
4132 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4133 was in use, was not putting the data itself into the right store pool;
4134 consequently, it could be overwritten for a subsequent message in the same
4135 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4138 38. Sometimes the final signoff response after QUIT could fail to get
4139 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4140 < 0 before doing a fflush(). This bug looks as though it goes back to the
4141 introduction of TLS in release 3.20, but "sometimes" must have been rare
4142 because the tests only now provoked it.
4148 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4149 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4150 Now it returns an integer. A number of calls were still expecting a T/F
4151 return. Fortuitously, in all cases, the tests worked in OK situations,
4152 which is the norm. However, things would have gone wrong on any write
4153 failures on the smtp file descriptor. This function is used when sending
4154 messages over SMTP and also when doing verify callouts.
4156 2. When Exim is called to do synchronous delivery of a locally submitted
4157 message (the -odf or -odi options), it no longer closes stderr before doing
4160 3. Implemented the mua_wrapper option.
4162 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4164 5. Implemented the functions header_remove(), header_testname(),
4165 header_add_at_position(), and receive_remove_recipient(), and exported them
4168 6. If an ACL "warn" statement specified the addition of headers, Exim already
4169 inserted X-ACL-Warn: at the start if there was no header name. However, it
4170 was not making this test for the second and subsequent header lines if
4171 there were newlines in the string. This meant that an invalid header could
4172 be inserted if Exim was badly configured.
4174 7. Allow an ACL "warn" statement to add header lines at the start or after all
4175 the Received: headers, as well as at the end.
4177 8. Added the rcpt_4xx retry error code.
4179 9. Added postmaster_mailfrom=xxx to callout verification option.
4181 10. Added mailfrom=xxxx to the callout verification option, for verify=
4184 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4185 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4187 12. Inserted some casts to stop certain compilers warning when using pointer
4188 differences as field lengths or precisions in printf-type calls (mostly
4189 affecting debugging statements).
4191 13. Added optional readline() support for -be (dynamically loaded).
4193 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4194 same clock tick as a message's arrival, so that its received time was the
4195 same as the "first fail" time on the retry record, and that message
4196 remained on the queue past the ultimate address timeout, every queue runner
4197 would try a delivery (because it was past the ultimate address timeout) but
4198 after another failure, the ultimate address timeout, which should have then
4199 bounced the address, did not kick in. This was a "< instead of <=" error;
4200 in most cases the first failure would have been in the next clock tick
4201 after the received time, and all would be well.
4203 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4204 being recognized when the domain list was tested by the match_domain
4205 condition in an expansion string.
4207 16. Added the ${str2b64: operator.
4209 17. Exim was always calling setrlimit() to set a large limit for the number of
4210 processes, without checking whether the existing limit was already
4211 adequate. (It did check for the limit on file descriptors.) Furthermore,
4212 errors from getrlimit() and setrlimit() were being ignored. Now they are
4213 logged to the main and panic logs, but Exim does carry on, to try to do its
4214 job under whatever limits there are.
4216 18. Imported PCRE 5.0.
4218 19. Trivial typo in log message " temporarily refused connection" (the leading
4221 20. If the log selector return_path_on_delivery was set and an address was
4222 redirected to /dev/null, the delivery process crashed because it assumed
4223 that a return path would always be set for a "successful" delivery. In this
4224 case, the whole delivery is bypassed as an optimization, and therefore no
4227 21. Internal re-arrangement: the function for sending a challenge and reading
4228 a response while authentication was assuming a zero-terminated challenge
4229 string. It's now changed to take a pointer and a length, to allow for
4230 binary data in such strings.
4232 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4234 23. Exim was not respecting finduser_retries when seeking the login of the
4235 uid under which it was called; it was always trying 10 times. (The default
4236 setting of finduser_retries is zero.) Also, it was sleeping after the final
4237 failure, which is pointless.
4239 24. Implemented tls_on_connect_ports.
4241 25. Implemented acl_smtp_predata.
4243 26. If the domain in control=submission is set empty, Exim assumes that the
4244 authenticated id is a complete email address when it generates From: or
4245 Sender: header lines.
4247 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4248 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4249 chown and chgrp in /bin and hostname in /usr/bin.
4251 28. Exim was keeping the "process log" file open after each use, just as it
4252 does for the main log. This opens the possibility of it remaining open for
4253 long periods when the USR1 signal hits a daemon. Occasional processlog
4254 errors were reported, that could have been caused by this. Anyway, it seems
4255 much more sensible not to leave this file open at all, so that is what now
4258 29. The long-running daemon process does not normally write to the log once it
4259 has entered its main loop, and it closes the log before doing so. This is
4260 so that log files can straightforwardly be renamed and moved. However,
4261 there are a couple of unusual error situations where the daemon does write
4262 log entries, and I had neglected to close the log afterwards.
4264 30. The text of an SMTP error response that was received during a remote
4265 delivery was being truncated at 512 bytes. This is too short for some of
4266 the long messages that one sometimes sees. I've increased the limit to
4269 31. It is now possible to make retry rules that apply only when a message has a
4270 specific sender, in particular, an empty sender.
4272 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4273 it possible to be selective about when SMTP synchronization is enforced.
4275 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4277 32. Implemented hosts_connection_nolog.
4279 33. Added an ACL for QUIT.
4281 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4284 35. Added mailbox_size and mailbox_filecount to appendfile.
4286 36. Added control = no_multiline_responses to ACLs.
4288 37. There was a bug in the logic of the code that waits for the clock to tick
4289 in the case where the clock went backwards by a substantial amount such
4290 that the microsecond fraction of "now" was more than the microsecond
4291 fraction of "then" (but the whole seconds number was less).
4293 38. Added support for the libradius Radius client library this is found on
4294 FreeBSD (previously only the radiusclient library was supported).
4300 1. When certain lookups returned multiple values in the form name=value, the
4301 quoting of the values was not always being done properly. Specifically:
4302 (a) If the value started with a double quote, but contained no whitespace,
4304 (b) If the value contained whitespace other than a space character (i.e.
4305 tabs or newlines or carriage returns) it was not quoted.
4306 This fix has been applied to the mysql and pgsql lookups by writing a
4307 separate quoting function and calling it from the lookup code. The fix
4308 should probably also be applied to nisplus, ibase and oracle lookups, but
4309 since I cannot test any of those, I have not disturbed their existing code.
4311 2. A hit in the callout cache for a specific address caused a log line with no
4312 reason for rejecting RCPT. Now it says "Previous (cached) callout
4313 verification failure".
4315 3. There was an off-by-one bug in the queryprogram router. An over-long
4316 return line was truncated at 256 instead of 255 characters, thereby
4317 overflowing its buffer with the terminating zero. As well as fixing this, I
4318 have increased the buffer size to 1024 (and made a note to document this).
4320 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4321 when Exim is waiting for an SMTP response from a remote server, Exim
4322 restarts its select() call on the socket, thereby resetting its timeout.
4323 This is not a problem when such interrupts are rare. Somebody set up a cron
4324 job to run exiwhat every 2 minutes, which is less than the normal select()
4325 timeout (5 or 10 minutes). This meant that the select() timeout never
4326 kicked in because it was always reset. I have fixed this by comparing the
4327 time when an interrupt arrives with the time at the start of the first call
4328 to select(). If more time than the timeout has elapsed, the interrupt is
4329 treated as a timeout.
4331 5. Some internal re-factoring in preparation for the addition of Sieve
4332 extensions (by MH). In particular, the "personal" test is moved to a
4333 separate function, and given an option for scanning Cc: and Bcc: (which is
4334 not set for Exim filters).
4336 6. When Exim created an email address using the login of the caller as the
4337 local part (e.g. when creating a From: or Sender: header line), it was not
4338 quoting the local part when it contained special characters such as @.
4340 7. Installed new OpenBSD configuration files.
4342 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4343 try to make them clearer.
4345 9. Callout options, other than the timeout value, were being ignored when
4346 verifying sender addresses in header lines. For example, when using
4348 verify = header_sender/callout=no_cache
4350 the cache was (incorrectly) being used.
4352 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4353 only the Cygwin environment.
4355 11. When return_path_on_delivery was set as a log selector, if different remote
4356 addresses in the same message used different return paths and parallel
4357 remote delivery occurred, the wrong values would sometimes be logged.
4358 (Whenever a remote delivery process finished, the return path value from
4359 the most recently started remote delivery process was logged.)
4361 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4362 lines when AUTH and/or TLS are in use. This is the "received protocol"
4363 field. Exim used to use "asmtp" for authenticated SMTP, without any
4364 indication (in the protocol name) for TLS use. Now it follows the RFC and
4365 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4366 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4367 names appear in log lines as well as in Received: header lines.
4369 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4370 extensions, and comparison tests, and to fix some bugs.
4372 14. Changes to the "personal" filter test:
4374 (1) The test was buggy in that it was just doing the equivalent of
4375 "contains" tests on header lines. For example, if a user's address was
4376 anne@some.where, the "personal" test would incorrectly be true for
4378 To: susanne@some.where
4380 This test is now done by extracting each address from the header in turn,
4381 and checking the entire address. Other tests that are part of "personal"
4382 are now done using regular expressions (for example, to check local parts
4383 of addresses in From: header lines).
4385 (2) The list of non-personal local parts in From: addresses has been
4386 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4387 taken from the Sieve specification recommendations.
4389 (3) If the message contains any header line starting with "List-" it is
4390 treated as non-personal.
4392 (4) The test for "circular" in the Subject: header line has been removed
4393 because it now seems ill-conceived.
4395 15. Minor typos in src/EDITME comments corrected.
4397 16. Installed latest exipick from John Jetmore.
4399 17. If headers_add on a router specified a text string that was too long for
4400 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4401 of string_sprintf() is now avoided.
4403 18. $message_body_size was not set (it was always zero) when running the DATA
4404 ACL and the local_scan() function.
4406 19. For the "mail" command in an Exim filter, no default was being set for
4407 the once_repeat time, causing a random time value to be used if "once" was
4408 specified. (If the value happened to be <= 0, no repeat happened.) The
4409 default is now 0s, meaning "never repeat". The "vacation" command was OK
4410 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4411 (I found it when inspecting the code).
4413 20. There is now an overall timeout for performing a callout verification. It
4414 defaults to 4 times the callout timeout, which applies to individual SMTP
4415 commands during the callout. The overall timeout applies when there is more
4416 than one host that can be tried. The timeout is checked before trying the
4417 next host. This prevents very long delays if there are a large number of
4418 hosts and all are timing out (e.g. when the network connections are timing
4419 out). The value of the overall timeout can be changed by specifying an
4420 additional sub-option for "callout", called "maxwait". For example:
4422 verify = sender/callout=5s,maxwait=20s
4424 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4425 to the end before writing, but this should make it even safer).
4427 22. Exim was forgetting that it had advertised PIPELINING for the second and
4428 subsequent messages on an SMTP connection. It was also not resetting its
4429 memory on STARTTLS and an internal HELO.
4431 23. When Exim logs an SMTP synchronization error within a session, it now
4432 records whether PIPELINING has been advertised or not.
4434 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4435 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4436 rather than long int.
4438 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4440 26. Added the never_mail option to autoreply.
4446 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4447 crash if the getsockname() call failed; this can happen if a connection is
4448 closed very soon after it is established. The problem was simply in the
4449 order in which certain operations were done, causing Exim to try to write
4450 to the SMTP stream before it had set up the file descriptor. The bug has
4451 been fixed by making things happen in the correct order.
4457 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4458 before the connection was closed, thus losing the rejection response.
4460 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4461 some early Solaris releases, but causes trouble in current releases where
4462 socklen_t is defined.
4464 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4467 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4470 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4471 that releasing the top of it at the end releases what was used for sub-
4472 expansions (unless the block got too big). However, discard this block if
4473 the first thing is a variable or header, so that we can use its block when
4474 it is dynamic (useful for very large $message_headers, for example).
4476 6. Lookups now cache *every* query, not just the most recent. A new, separate
4477 store pool is used for this. It can be recovered when all lookup caches are
4478 flushed. Lookups now release memory at the end of their result strings.
4479 This has involved some general refactoring of the lookup sources.
4481 7. Some code has been added to the store_xxx() functions to reduce the amount
4482 of flapping under certain conditions.
4484 8. log_incoming_interface used to affect only the <= reception log lines. Now
4485 it causes the local interface and port to be added to several more SMTP log
4486 lines, for example "SMTP connection from", and rejection lines.
4488 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4490 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4492 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4493 could be overwritten at the end of the current message (or the start of a
4494 new message if it was set in a HELO ACL). The value is now preserved for
4495 the duration of the SMTP connection.
4497 12. If a transport had a headers_rewrite setting, and a matching header line
4498 contained an unqualified address, that address was qualified, even if it
4499 did not match any rewriting rules. The underlying bug was that the values
4500 of the flags that permit the existence of unqualified sender and recipient
4501 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4502 non-local messages, and by -bnq for local messages) were not being
4503 preserved with the message after it was received.
4505 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4506 "next input=" as part of the text comprising the host identity instead of
4507 the correct text. The code was using the same buffer for two different
4508 strings. However, depending on which order the printing function evaluated
4509 its arguments, the bug did not always show up. Under Linux, for example, my
4510 test suite worked just fine.
4512 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4513 doesn't work with some older versions of Perl. It has been changed to "my",
4514 which in any case is probably the better facility to use.
4516 15. A really picky compiler found some instances of statements for creating
4517 error messages that either had too many or two few arguments for the format
4520 16. The size of the buffer for calls to the DNS resolver has been increased
4521 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4522 for addresses that have a lot of PTR records. This alleviates a problem; it
4523 does not fully solve it.
4525 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4526 buffer now truncates the list and logs the incident, which is the same
4527 action as happens when Exim is looking up a host name and its aliases.
4528 Previously in this situation something unpredictable would happen;
4529 sometimes it was "internal error: store_reset failed".
4531 18. If a server dropped the connection unexpectedly when an Exim client was
4532 using GnuTLS and trying to read a response, the client delivery process
4533 crashed while trying to generate an error log message.
4535 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4536 string, for example:
4538 warn message = H1: something\nH2: something
4540 the text was added as a single header line from Exim's point of view
4541 though it ended up OK in the delivered message. However, searching for the
4542 second and subsequent header lines using $h_h2: did not work. This has been
4543 fixed. Similarly, if a system filter added multiple headers in this way,
4544 the routers could not see them.
4546 20. Expanded the error message when iplsearch is called with an invalid key to
4547 suggest using net-iplsearch in a host list.
4549 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4550 ACLs are no longer actually imposed (and a message to that effect is
4553 22. If a "gecos" field in a passwd entry contained escaped characters, in
4554 particular, if it contained a \" sequence, Exim got it wrong when building
4555 a From: or a Sender: header from that name. A second bug also caused
4556 incorrect handling when an unquoted " was present following a character
4557 that needed quoting.
4559 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4560 was not being matched caselessly.
4562 24. Arranged for all hyphens in the exim.8 source to be escaped with
4565 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4566 properly. Recipient callout cache records were still being keyed to include
4567 the sender, even when use_sender was set false. This led to far more
4568 callouts that were necessary. The sender is no longer included in the key
4569 when use_sender is false.
4571 26. Added "control = submission" modifier to ACLs.
4573 27. Added the ${base62d: operator to decode base 62 numbers.
4575 28. dnsdb lookups can now access SRV records.
4577 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4578 the configuration file.
4580 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4581 (-v) mode. This makes the output for a verbose queue run more intelligible.
4583 31. Added a use_postmaster feature to recipient callouts.
4585 32. Added the $body_zerocount variable, containing the number of binary zero
4586 bytes in the message body.
4588 33. The time of last modification of the "new" subdirectory is now used as the
4589 "mailbox time last read" when there is a quota error for a maildir
4592 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4594 35. Added +ignore_unknown as a special item in host lists.
4596 36. Code for decoding IPv6 addresses in host lists is now included, even if
4597 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4598 address was recognized as an IP address, but was then not correctly decoded
4599 into binary, causing unexpected and incorrect effects when compared with
4606 1. Very minor rewording of debugging text in manualroute to say "list of
4607 hosts" instead of "hostlist".
4609 2. If verify=header_syntax was set, and a header line with an unqualified
4610 address (no domain) and a large number of spaces between the end of the
4611 name and the colon was received, the reception process suffered a buffer
4612 overflow, and (when I tested it) crashed. This was caused by some obsolete
4613 code that should have been removed. The fix is to remove it!
4615 3. When running in the test harness, delay a bit after writing a bounce
4616 message to get a bit more predictability in the log output.
4618 4. Added a call to search_tidyup() just before forking a reception process. In
4619 theory, someone could use a lookup in the expansion of smtp_accept_max_
4620 per_host which, without the tidyup, could leave open a database connection.
4622 5. Added the variables $recipient_data and $sender_data which get set from a
4623 lookup success in an ACL "recipients" or "senders" condition, or a router
4624 "senders" option, similar to $domain_data and $local_part_data.
4626 6. Moved the writing of debug_print from before to after the "senders" test
4629 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4630 problems for message scanning, either using a data ACL, or using
4631 local_scan() because the Received: header was not generated till after they
4632 were called (in order to set the time as the time of reception completion).
4633 I have revised the way this works. The header is now generated after the
4634 body is received, but before the ACL or local_scan() are called. After they
4635 are run, the timestamp in the header is updated.
4641 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4642 before starting a queue runner without re-exec. This happened only when
4643 deliver_drop_privilege was set or when the Exim user was set to root. The
4644 effect of the bug was that timeouts during subsequent deliveries caused
4645 crashes instead of being properly handled. The handler is now left at its
4646 default (and expected) setting.
4648 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4649 message, again when deliver_drop_privilege is set or Exim is run as root.
4650 The bug described in (1) was not present in this case, but the tidying up
4651 of the other signals was missing. I have made the two cases consistent.
4653 3. The ignore_target_hosts setting on a manualroute router was being ignored
4654 for hosts that were looked up using the /MX notation.
4656 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4659 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4660 operated on the sender address. After changing the $sender_address to <>
4661 for the sender address verify, Exim was re-instated it as the original
4662 (before rewriting) address, but remembering that it had rewritten it, so it
4663 wasn't rewriting it again. This bug also had the effect of breaking the
4664 sender address verification caching when the sender address was rewritten.
4666 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4667 This has been changed so that if the ip literal address matches
4668 ignore_target_hosts, the router declines.
4670 7. Added expansion conditions match_domain, match_address, and match_local_
4671 part (NOT match_host).
4673 8. The placeholder for the Received: header didn't have a length field set.
4675 9. Added code to Exim itself and to exim_lock to test for a specific race
4676 condition that could lead to file corruption when using MBX delivery. The
4677 issue is with the lockfile that is created in /tmp. If this file is removed
4678 after a process has opened it but before that process has acquired a lock,
4679 there is the potential for a second process to recreate the file and also
4680 acquire a lock. This could lead to two Exim processes writing to the file
4681 at the same time. The added code performs the same test as UW imapd; it
4682 checks after acquiring the lock that its file descriptor still refers to
4683 the same named file.
4685 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4686 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4689 11. Added the smtp_active_hostname option. If used, this will typically be made
4690 to depend on the incoming interface address. Because $interface_address is
4691 not set up until the daemon has forked a reception process, error responses
4692 that can happen earlier (such as "too many connections") no longer contain
4695 12. If an expansion in a condition on a "warn" statement fails because a lookup
4696 defers, the "warn" statement is abandoned, and the next ACL statement is
4697 processed. Previously this caused the whole ACL to be aborted.
4699 13. Added the iplsearch lookup type.
4701 14. Added ident_timeout as a log selector.
4703 15. Added tls_certificate_verified as a log selector.
4705 16. Added a global option tls_require_ciphers (compare the smtp transport
4706 option of the same name). This controls incoming TLS connections.
4708 17. I finally figured out how to make tls_require_ciphers do a similar thing
4709 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4710 before starting the TLS session.
4712 18. Tabs are now shown as \t in -bP output.
4714 19. If the log selector return_path_on_delivery was set, Exim crashed when
4715 bouncing a message because it had too many Received: header lines.
4717 20. If two routers both had headers_remove settings, and the first one included
4718 a superfluous trailing colon, the final name in the first list and the
4719 first name in the second list were incorrectly joined into one item (with a
4720 colon in the middle).
4726 1. Added -C and -D options to the exinext utility, mainly to make it easier
4727 to include in the automated testing, but these could be helpful when
4728 multiple configurations are in use.
4730 2. The exinext utility was not formatting the output nicely when there was
4731 an alternate port involved in the retry record key, nor when there was a
4732 message id as well (for retries that were specific to a specific message
4733 and a specific host). It was also confused by IPv6 addresses, because of
4734 the additional colons they contain. I have fixed the IPv4 problem, and
4735 patched it up to do a reasonable job for IPv6.
4737 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4738 delivery, the log line now contains "pipelined" if PIPELINING was used.
4740 4. An SMTP transport process used to panic and die if the bind() call to set
4741 an explicit outgoing interface failed. This has been changed; it is now
4742 treated in the same way as a connect() failure.
4744 5. A reference to $sender_host_name in the part of a conditional expansion
4745 that was being skipped was still causing a DNS lookup. This no longer
4748 6. The def: expansion condition was not recognizing references to header lines
4749 that used bh_ and bheader_.
4751 7. Added the _cache feature to named lists.
4753 8. The code for checking quota_filecount in the appendfile transport was
4754 allowing one more file than it should have been.
4756 9. For compatibility with Sendmail, the command line option
4764 and sets the incoming protocol and host name (for trusted callers). The
4765 host name and its colon can be omitted when only the protocol is to be set.
4766 Note the Exim already has two private options, -pd and -ps, that refer to
4767 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4768 "s", but I don't think that's a major issue.
4770 10. A number of refactoring changes to the code, none of which should affect
4773 (a) The number of logging options was getting close to filling up the
4774 32-bit word that was used as a bit map. I have split them into two classes:
4775 those that are passed in the argument to log_write(), and those that are
4776 only ever tested independently outside of that function. These are now in
4777 separate 32-bit words, so there is plenty of room for expansion again.
4778 There is no change in the user interface or the logging behaviour.
4780 (b) When building, for example, log lines, the code previously used a
4781 macro that called string_cat() twice, in order to add two strings. This is
4782 not really sufficiently general. Furthermore, there was one instance where
4783 it was actually wrong because one of the argument was used twice, and in
4784 one call a function was used. (As it happened, calling the function twice
4785 did not affect the overall behaviour.) The macro has been replaced by a
4786 function that can join an arbitrary number of extra strings onto a growing
4789 (c) The code for expansion conditions now uses a table and a binary chop
4790 instead of a serial search (which was left over from when there were very
4791 few conditions). Also, it now recognizes conditions like "pam" even when
4792 the relevant support is not compiled in: a suitably worded error message is
4793 given if an attempt is made to use such a condition.
4795 11. Added ${time_interval:xxxxx}.
4797 12. A bug was causing one of the ddress fields not to be passed back correctly
4798 from remote delivery subprocesses. The field in question was not being
4799 subsequently used, so this caused to problems in practice.
4801 13. Added new log selectors queue_time and deliver_time.
4803 14. Might have fixed a bug in maildirsizefile handling that threw up
4804 "unexpected character" debug warnings, and recalculated the data
4805 unnecessarily. In any case, I expanded the warning message to give more
4808 15. Added the message "Restricted characters in address" to the statements in
4809 the default ACL that block characters like @ and % in local parts.
4811 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4812 Three changes have been made:
4814 (a) There was a serious bug; a negative response to MAIL caused the whole
4815 recipient domain to be cached as invalid, thereby blocking all messages
4816 to all local parts at the same domain, from all senders. This bug has
4817 been fixed. The domain is no longer cached after a negative response to
4818 MAIL if the sender used is not empty.
4820 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4823 (c) A new callout option, "use_sender" has been added for people who want
4824 the modified behaviour.
4830 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4833 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4834 indeed breaks things for older releases.
4836 3. Added additional logging to the case where there is a problem reading data
4837 from a filter that is running in a subprocess using a pipe, in order to
4838 try to track down a specific problem.
4840 4. Testing facility fudge: when running in the test harness and attempting
4841 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4842 getting "No route to host". Convert this to a timeout.
4844 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4847 6. Some OS don't have socklen_t but use size_t instead. This affects the
4848 fifth argument of getsockopt() amongst other things. This is now
4849 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4850 can be set for individual OS. I have set it for SunOS5, OSF1, and
4851 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4852 some earlier ones do not.
4854 7. Change 4.30/15 was not doing the test caselessly.
4856 8. The standard form for an IPv6 address literal was being rejected by address
4857 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4858 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4859 this, as well as the form without the "IPv6" on the front (but only when
4860 address literals are enabled, of course).
4862 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4864 10. Exim crashed if a message with an empty sender address specified by -f
4865 encountered a router with an errors_to setting. This could be provoked only
4866 by a command such as
4870 where an empty string was supplied; "<>" did not hit this bug.
4872 11. Installed PCRE release 4.5.
4874 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4875 remained set. It is now erased.
4877 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4878 times from message ids (which are base 36 rather than the normal 62).
4880 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4881 were being counted as actual protocol errors, and logged if the log
4882 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4883 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4884 and DATA following a set of rejected RCPTs do not count as protocol errors.
4885 In other words, Exim assumes they were pipelined, though this may not
4886 actually be the case. Of course, in all cases the client gets an
4887 appropriate error code.
4889 15. If a lookup fails in an ACL condition, a message about the failure may
4890 be available; it is used if testing the ACL cannot continue, because most
4891 such messages specify what the cause of the deferral is. However, some
4892 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4893 that caused an old message to be retained and used if a later statement
4894 caused a defer, replacing the real cause of the deferral.
4896 16. If an IP address had so many PTR records that the DNS lookup buffer
4897 was not large enough to hold them, Exim could crash while trying to process
4898 the truncated data. It now detects and logs this case.
4900 17. Further to 4.21/58, another change has been made: if (and only if) the
4901 first line of a message (the first header line) ends with CRLF, a bare LF
4902 in a subsequent header line has a space inserted after it, so as not to
4903 terminate the header.
4905 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4906 unnecessarily, used atoi() instead of strtol(), and didn't check the
4907 termination when getting file sizes from file names by regex.
4909 19. Completely re-implemented the support for maildirsize files, in the light
4910 of a number of problems with the previous contributed implementation
4911 (4.30/29). In particular:
4913 . If the quota is zero, the maildirsize file is maintained, but no quota is
4916 . If the maildir directory does not exist, it is created before any attempt
4917 to write a maildirsize file.
4919 . The quota value in the file is just a cache; if the quota is changed in
4920 the transport, the new value overrides.
4922 . A regular expression is available for excluding directories from the
4925 20. The autoreply transport checks the characters in options that define the
4926 message's headers; it allows continued headers, but it was checking with
4927 isspace() after an embedded newline instead of explicitly looking for a
4930 21. If all the "regular" hosts to which an address was routed had passed their
4931 expiry times, and had not reached their retry times, the address was
4932 bounced, even if fallback hosts were defined. Now Exim should go on to try
4935 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4936 equivalent code in the SMTP transport. Some hosts send humungous responses
4937 to HELO/EHLO, more than 1024 it seems.
4939 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4940 gives compiler warnings in some environments. I've now done it "properly",
4943 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4944 (because of problems with the built-in one) was declared to return uschar *
4945 instead of char *, causing compiler failure.
4947 25. Fixed a file descriptor leak when processing alias/forward files.
4949 26. Fixed a minor format string issue in dbfn.c.
4951 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4953 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4954 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4957 29. When a sender address is verified, it is cached, to save repeating the test
4958 when there is more than one recipient in a message. However, when the
4959 verification involves a callout, it is possible for different callout
4960 options to be set for different recipients. It is too complicated to keep
4961 track of this in the cache, so now Exim always runs a verification when a
4962 callout is required, relying on the callout cache for the optimization.
4963 The overhead is duplication of the address routing, but this should not be
4966 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4967 to be verified with callout=postmaster, and the main callout worked but the
4968 postmaster check failed, the verification correctly failed. However, if a
4969 subsequent RCPT command asked for sender verification *without* the
4970 postmaster check, incorrect caching caused this verification also to fail,
4973 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4974 it was not caching the DNS options (qualify_single, search_parents) that
4975 were used when the lookup failed. A subsequent lookup with different
4976 options therefore always gave the same answer, though there were cases
4977 where it should not have. (Example: a "domains = !$mx_any" option on a
4978 dnslookup router: the "domains" option is always processed without any
4979 widening, but the router might have qualify_single set.) Now Exim uses the
4980 cached value only when the same options are set.
4982 32. Added John Jetmore's "exipick" utility to the distribution.
4984 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4985 than a timeout (e.g. a certificate is required, and is not provided), an
4986 Exim server now closes the connection immediately. Previously it waited for
4987 the client to close - but if the client is SSL, it seems that they each
4988 wait for each other, leading to a delay before one of them times out.
4990 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4991 maintained 0.8.x compatibility because I don't think many are using it, and
4992 it is clearly obsolete.
4994 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4997 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4998 client certificate was expired. A simple patch fixes this, though I don't
4999 understand the full logic of why the verify callback is called multiple
5002 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5003 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5004 which causes problems with some clients (such as the Certicom SSL Plus
5005 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5006 disables the coutermeasure allowing Eudora to connect."
5008 38. Exim was not checking that a write() to a log file succeeded. This could
5009 lead to Bad Things if a log got too big, in particular if it hit a file
5010 size limit. Exim now panics and dies if it cannot write to a log file, just
5011 as it does if it cannot open a log file.
5013 39. Modified OS/Makefile-Linux so that it now contains
5015 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5017 The two -D definitions ensure that Exim is compiled with large file
5018 support, which makes it possible to handle log files that are bigger than
5021 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5022 instance) a domain was checked against a named list that involved a lookup,
5023 causing $domain_data to be set, then another domain was checked against the
5024 same list, then the first domain was re-checked, the value of $domain_data
5025 after the final check could be wrong. In particular, if the second check
5026 failed, it could be set empty. This bug probably also applied to
5029 41. The strip_trailing_dot option was not being applied to the address given
5030 with the -f command-line option.
5032 42. The code for reading a message's header from the spool was incrementing
5033 $received_count, but never initializing it. This meant that the value was
5034 incorrect (doubled) while delivering a message in the same process in which
5035 it was received. In the most common configuration of Exim, this never
5036 happens - a fresh exec is done - but it can happen when
5037 deliver_drop_privilege is set.
5039 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5040 it now includes up to 150 characters of the unexpected data in the log
5043 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5044 and building data strings. The size of both of these buffers was 10 000
5045 bytes - far larger than anybody would *ever* want, thought I. Needless to
5046 say, somebody hit the limit. I have increased the maximum line length to
5047 20 000 and the maximum data length of concatenated lines to 100 000. I have
5048 also fixed two bugs, because there was no checking on these buffers. Tsk,
5049 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5050 buffer is too small.
5052 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5053 lsearch lookups. Now it does.
5055 46. When parsing a route_list item in a manualroute router, a fixed-length
5056 buffer was used for the list of hosts. I made this 1024 bytes long,
5057 thinking that nobody would ever have a list of hosts that long. Wrong.
5058 Somebody had a whole pile of complicated expansion conditions, and the
5059 string was silently truncated, leading to an expansion error. It turns out
5060 that it is easier to change to an unlimited length (owing to other changes
5061 that have happened since this code was originally written) than to build
5062 structure for giving a limitation error. The length of the item that
5063 expands into the list of hosts is now unlimited.
5065 47. The lsearch lookup could not handle data where the length of text line was
5066 more than 4095 characters. Such lines were truncated, leading to shortened
5067 data being returned. It should now handle lines of any length.
5069 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5070 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5073 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5074 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5075 no longer happens in comment lines. A list of replacements is now placed
5076 at the head of all of the source files, except those whose only change is
5077 to replace PERL_COMMAND in the very first #! line.
5079 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5080 messages on the queue, with a bottom-up merge sort, using code contributed
5081 by Michael Haardt. This should make operations like -bp somewhat faster on
5082 large queues. It won't affect queue runners, except when queue_run_in_order
5085 51. Installed eximstats 1.31 in the distribution.
5087 52. Added support for SRV lookups to the dnslookup router.
5089 53. If an ACL referred to $message_body or $message_body_end, the value was not
5090 reset for any messages that followed in the same SMTP session.
5092 54. The store-handling optimization for building very long strings was not
5093 differentiating between the different store pools. I don't think this
5094 actually made any difference in practice, but I've tidied it.
5096 55. While running the routers to verify a sender address, $sender_address
5097 was still set to the sender address. This is wrong, because when routing to
5098 send a bounce to the sender, it would be empty. Therefore, I have changed
5099 it so that, while verifying a sender address, $sender_address is set to <>.
5100 (There is no change to what happens when verifying a recipient address.)
5102 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5103 target A or AAAA records (if not already returned) without resetting the
5104 qualify_single or search_parents options of the DNS resolver. These are
5105 inappropriate in this case because the targets of MX and SRV records must
5106 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5107 target that, when qualified, matched something in the local domain. These
5108 two options are now turned off when doing these lookups.
5110 57. It seems that at least some releases of Reiserfs (which does not have the
5111 concept of a fixed number of inodes) returns zero and not -1 for the
5112 number of available inodes. This interacted badly with check_spool_inodes,
5113 which assumed that -1 was the "no such thing" setting. What I have done is
5114 to check that the total number of inodes is greater than zero before doing
5115 the test of how many are available.
5117 58. When a "warn" ACL statement has a log_message modifier, the message is
5118 remembered, and not repeated. This is to avoid a lot of repetition when a
5119 message has many recipients that cause the same warning to be written.
5120 Howewer, Exim was preserving the list of already written lines for an
5121 entire SMTP session, which doesn't seem right. The memory is now reset if a
5122 new message is started.
5124 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5125 debugging output unless log_rewrite was also set.
5127 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5128 of (int)(handle) when we know that handle contains (void *)(-1).
5130 61. The Exim daemon panic-logs an error return when it closes the incoming
5131 connection. However "connection reset by peer" seems to be common, and
5132 isn't really an error worthy of noting specially, so that particular error
5135 62. When Exim is trying to find all the local interfaces, it used to panic and
5136 die if the ioctl to get the interface flags failed. However, it seems that
5137 on at least one OS (Solaris 9) it is possible to have an interface that is
5138 included in the list of interfaces, but for which you get a failure error
5139 for this call. This happens when the interface is not "plumbed" into a
5140 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5141 failure of the "get flags" call assumes that the interface is down.
5143 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5144 makes life easier for people who are doing arithmetic on fields extracted
5145 from dates, where you often get leading zeros that should not be
5146 interpreted as octal.
5148 64. Added qualify_domain to the redirect router, to override the global
5151 65. If a pathologically long header line contained very many addresses (the
5152 report of this problem mentioned 10 000) and each of them was rewritten,
5153 Exim could use up a very large amount of memory. (It kept on making new
5154 copies of the header line as it rewrote, and never released the old ones.)
5155 At the expense of a bit more processing, the header rewriting function has
5156 been changed so that it no longer eats memory in this way.
5158 66. The generation of the Received: header has been moved from the time that a
5159 message starts to be received, to the time that it finishes. The timestamp
5160 in the Received: header should now be very close to that of the <= log
5161 line. There are two side-effects of this change:
5163 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5164 logged header lines no longer include the local Received: line, because
5165 it has not yet been created. The same applies to a copy of the message
5166 that is returned to a non-SMTP sender when a message is rejected.
5168 (b) When a filter file is tested using -bf, no additional Received: header
5169 is added to the test message. After some thought, I decided that this
5172 This change does not affect the value of $received_for. It is still set
5173 after address rewriting, but before local_scan() is called.
5175 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5177 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5178 gave an unhelpful panic error message, and a defer error. I have managed to
5179 change this behaviour so that it now rejects any supplied certificate,
5180 which seems right, as the list of acceptable certificates is empty.
5182 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5183 gave an unhelpful defer error. I have not managed to make this reject any
5184 supplied certificates, but the error message it gives is "no certificate
5185 supplied", which is not helpful.
5187 70. exigrep's output now also includes lines that are not associated with any
5188 message, but which match the given pattern. Implemented by a patch from
5189 Martin Sluka, which also tidied up the Perl a bit.
5191 71. Recipient callout verification, like sender verification, was using <> in
5192 the MAIL FROM command. This isn't really the right thing, since the actual
5193 sender may affect whether the remote host accepts the recipient or not. I
5194 have changed it to use the actual sender in the callout; this means that
5195 the cache record is now keyed on a recipient/sender pair, not just the
5196 recipient address. There doesn't seem to be a real danger of callout loops,
5197 since a callout by the remote host to check the sender would use <>.
5198 [SEE ABOVE: changed after hitting problems.]
5200 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5201 temporary errors. However, in the case of such a code being given after
5202 the end of a data transmission (i.e. after ".") Exim was failing to write
5203 a retry record for the message. (Yes, there was some broken host that was
5204 actually sending 8xx at this point.)
5206 73. An unknown lookup type in a host list could cause Exim to panic-die when
5207 the list was checked. (An example that provoked this was putting <; in the
5208 middle of a list instead of at the start.) If this happened during a DATA
5209 ACL check, a -D file could be left lying around. This kind of configuration
5210 error no longer causes Exim to die; instead it causes a defer errror. The
5211 incident is still logged to the main and panic logs.
5213 74. Buglet left over from Exim 3 conversion. The message "too many messages
5214 in one connection" was written to the rejectlog but not the mainlog, except
5215 when address rewriting (yes!) was being logged.
5217 75. Added write_rejectlog option.
5219 76. When a system filter was run not as root (that is, when system_filter_user
5220 was set), the values of the $n variables were not being returned to the
5221 main process; thus, they were not subsequently available in the $sn
5224 77. Added +return_path_on_delivery log selector.
5226 78. A connection timeout was being treated differently from recipients deferred
5227 when testing hosts_max_try with a message that was older than the host's
5228 retry timeout. (The host should not be counted, thus allowing all hosts to
5229 be tried at least once before bouncing.) This may have been the cause of an
5230 occasionally reported bug whereby a message would remain on the queue
5231 longer than the retry timeout, but would be bounced if a delivery was
5232 forced. I say "may" because I never totally pinned down the problem;
5233 setting up timeout/retry tests is difficult. See also the next item.
5235 79. The ultimate address timeout was not being applied to errors that involved
5236 a combination of host plus message (for example, a timeout on a MAIL
5237 command). When an address resolved to a number of possible hosts, and they
5238 were not all tried for each delivery (e.g. because of hosts_max_try), a
5239 message could remain on the queue longer than the retry timeout.
5241 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5244 81. Fixed an obscure SMTP outgoing bug which required at least the following
5245 conditions: (a) there was another message waiting for the same server;
5246 (b) the server returned 5xx to all RCPT commands in the first message so
5247 that the message was not completed; (c) the server dropped the connection
5248 or gave a negative response to the RSET that Exim sends to abort the
5249 transaction. The observed case was a dropped connection after DATA that had
5250 been sent in pipelining mode. That is, the server had advertised PIPELINING
5251 but was not implementing it correctly. The effect of the bug was incorrect
5252 behaviour, such as trying another host, and this could lead to a crash.
5258 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5259 and daemon.c were passed as pointers to ints; they should have been
5260 pointers to socklen_t variables (which are typically unsigned ints).
5262 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5265 3. Fixed a really odd bug that affected only the testing scheme; patching a
5266 certain fixed string in the binary changed the value of another string that
5267 happened to be identical to the end of the original first string.
5269 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5270 name", it returns that address as the IP address. On some operating
5271 systems (e.g. Solaris), it also passes back the IP address string as the
5272 "host name". However, on others (e.g. Linux), it passes back an empty
5273 string. Exim wasn't checking for this, and was changing the host name to an
5274 empty string, assuming it had been canonicized.
5276 5. Although rare, it is permitted to have more than one PTR record for a given
5277 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5278 all the names associated with an address, because they do in Solaris.
5279 However, it seems that they do not in Linux for data that comes from the
5280 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5281 I found this out when I moved to a new Linux workstation and tried to run
5282 the Exim test suite.
5284 To get round this problem I have changed the code so that it now does its
5285 own call to the DNS to look up PTR records when searching for a host name.
5286 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5287 addresses that are only in /etc/hosts are still found.
5289 This behaviour is, however, controlled by an option called host_lookup_
5290 order, which defaults to "bydns:byaddr". If people want to use the other
5291 order, or indeed, just use one or the other means of lookup, they can
5292 specify it in this variable.
5294 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5295 some operating systems, this comes back from gethostbyaddr() as an empty
5296 string, and this is what Exim used to test for. However, it seems that in
5297 other systems, "." is yielded. Exim now tests for this case too.
5299 7. The values of check_spool_space and check_log_space are now held internally
5300 as a number of kilobytes instead of an absolute number of bytes. If a
5301 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5302 kilobyte. This means that much larger values can be stored.
5304 8. Exim monitor: an attempt to get the action menu when not actually pointing
5305 at a message produces an empty menu entitled "No message selected". This
5306 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5307 no entries in it ("Shell widget menu has zero width and/or height"). So I
5308 have added a single, blank menu entry in this case.
5310 9. Added ${quote_local_part.
5312 10. MIME decoding is now applied to the contents of Subject: header lines when
5315 11. Now that a reference to $sender_host_address automatically causes a reverse
5316 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5317 host lookup before query-style lookups in lists that might use this
5318 variable. This has therefore been abolished, and the "net-" prefix is no
5319 longer necessary for query-style lookups.
5321 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5322 have been a typo for LFLAGS, so it has been changed.
5324 13. The install script calls Exim with "-C /dev/null" in order to find the
5325 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5326 to be output. Howeve, since Exim outputs its version number before the
5327 error, it didn't break the script. It just looked ugly. I fixed this by
5328 always allowing "-C /dev/null" if the caller is root.
5330 14. Ignore overlarge ACL variable number when reading spool file - insurance
5331 against a later release with more variables having written the file.
5333 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5334 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5335 this, as well as the form without the "IPv6" on the front.
5337 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5338 OS/Makefile-Darwin file.
5340 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5341 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5343 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5344 variable. However, some OS use ints for inodes. I've added cast to long int
5345 to get rid of the compiler warning.
5347 19. I had forgotten to lock out "/../" in configuration file names when
5348 ALT_CONFIG_PREFIX was set.
5350 20. Routers used for verification do not need to specify transports. However,
5351 if such a router generated a host list, and callout was configured, Exim
5352 crashed, because it could not find a port number from the (non-existent)
5353 transport. It now assumes port 25 in this circumstance.
5355 21. Added the -t option to exigrep.
5357 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5358 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5359 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5360 them is set, LOOKUP_LSEARCH is forced.
5362 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5364 24. Added sender and host information to the "rejected by local_scan()" log
5365 line; previously there was no indication of these.
5367 25. Added .include_if_exists.
5369 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5370 Linux. It turns out that not all file systems support this. Apparently some
5371 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5372 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5373 not supported on the file descriptor, is now ignored when Exim is trying to
5374 sync a directory. This applies only to Linux.
5376 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5378 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5379 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5382 29. Added support for maildirsize files from supplied patch (modified a bit).
5384 30. The use of :fail: followed by an empty string could lead Exim to respond to
5385 sender verification failures with (e.g.):
5387 550 Verification failed for <xxx>
5388 550 Sender verify failed
5390 where the first response line was missing the '-' that indicates it is not
5391 the final line of the response.
5393 31. The loop for finding the name of the user that called Exim had a hardwired
5394 limit of 10; it now uses the value of finduser_retries, which is used for
5395 all other user lookups.
5397 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5400 33. Exim was neglecting to zero errno before one call of strtol() when
5401 expanding a string and expecting an integer value. On some systems this
5402 resulted in spurious "integer overflow" errors. Also, it was casting the
5403 result into an int without checking.
5405 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5406 did not work. The code looks as if it has *never* worked, though it appears
5407 to have been documented since at least releast 1.62. I have made it work.
5409 35. The "timeout_DNS" error in retry rules, also documented since at least
5410 1.62, also never worked. As it isn't clear exactly what this means, and
5411 clearly it isn't a major issue, I have abolished the feature by treating it
5412 as "timeout", and writing a warning to the main and panic logs.
5414 36. The display of retry rules for -brt wasn't always showing the error code
5417 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5418 timeout_connect_A, timeout_connect_MX.
5420 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5421 to the empty sender.
5423 39. The daemon was not analysing the content of -oX till after it had closed
5424 stderr and disconnected from the controlling terminal. This meant that any
5425 syntax errors were only noted on the panic log, and the return code from
5426 the command was 0. By re-arranging the code a little, I've made the
5427 decoding happen first, so such errors now appear on stderr, and the return
5428 code is 1. However, the actual setting up of the sockets still happens in
5429 the disconnected process, so errors there are still only recorded on the
5432 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5433 connections (as happens on some IP stacks) was logged at start up time as
5434 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5435 it from "IPv6 and IPv4", which means that two separate sockets are being
5438 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5439 says whether AF_INET or AF_INET6 was passed as an argument.
5441 42. Exiwhat output was messed up when time zones were included in log
5444 43. Exiwhat now gives more information about the daemon's listening ports,
5445 and whether -tls-on-connect was used.
5447 44. The "port" option of the smtp transport is now expanded.
5449 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5450 silently ignored. Now an error message is written to the main and panic
5453 46. There's a new ACL modifier called "logwrite" which writes to a log file
5454 as soon as it is encountered.
5456 47. Added $local_user_uid and $local_user_gid at routing time.
5458 48. Exim crashed when trying to verify a sender address that was being
5461 49. Exim was recognizing only a space character after ".include". It now also
5462 recognizes a tab character.
5464 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5465 extracting the relevant information from the specification. The man page no
5466 longer contains scrambled data for the -d option, and I've added a section
5467 at the front about calling Exim under different names.
5469 51. Added "extra_headers" argument to the "mail" command in filter files.
5471 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5474 53. Installed eximstats 1.29.
5476 54. Added transport_filter_timeout as a generic transport option.
5478 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5479 Cc: header lines. This was required by RFC 822, but it not required by RFC
5482 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5483 incoming messages that did not have them. Now it does so only if the
5484 message originates locally, that is, if there is no associated remote host
5485 address. When Resent- header lines are present, this applies to the Resent-
5486 lines rather than the non-Resent- lines.
5488 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5489 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5491 58. Messages for configuration errors now include the name of the main
5492 configuration file - useful now that there may be more than one file in a
5493 list (.included file names were always shown).
5495 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5496 for those rare installations that do not start the daemon as root or run it
5497 setuid root. I've cut out the call to initgroups() if the daemon is not
5500 60. The Exim user and group can now be bound into the binary as text strings
5501 that are looked up at the start of Exim's processing.
5503 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5505 62. Added $mailstore_basename variable.
5507 63. Installed patch to sieve.c from Michael Haardt.
5509 64. When Exim failed to open the panic log after failing to open the main log,
5510 the original message it was trying to log was written to stderr and debug
5511 output, but if they were not available (the usual case in production), it
5512 was lost. Now it is written to syslog before the two lines that record the
5513 failures to open the logs.
5515 65. Users' Exim filters run in subprocesses under the user's uid. It is
5516 possible for a "deliver" command or an alias in a "personal" command to
5517 provoke an address rewrite. If logging of address rewriting is configured,
5518 this fails because the process is not running as root or exim. There may be
5519 a better way of dealing with this, but for the moment (because 4.30 needs
5520 to be released), I have disabled address rewrite logging when running a
5521 filter in a non-root, non-exim process.
5527 1. The buildconfig auxiliary program wasn't quoting the value set for
5528 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5529 not defined. This bug was masked in 4.22 by the effect that was fixed in
5532 2. Some messages that were rejected after a message id was allocated were
5533 shown as "incomplete" by exigrep. It no longer does this for messages that
5534 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5536 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5537 have allow_domain_literals set, the ID did not get logged in the <= line.
5538 Domain literals are now always recognized in Message-ID: header lines.
5540 4. The first argument for a ${extract expansion item is the key name or field
5541 number. Leading and trailing spaces in this item were not being ignored,
5542 causing some misleading effects.
5544 5. When deliver_drop_privilege was set, single queue runner processes started
5545 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5546 same command in the process it spins off) were not dropping privilege.
5548 6. When the daemon running as "exim" started a queue runner, it always
5549 re-executed Exim in the spun-off process. This is a waste of effort when
5550 deliver_drop_privilege is set. The new process now just calls the
5551 queue-runner function directly.
5557 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5560 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5561 ignored. Though the use of -oP was forcing the writing of a pid file, it
5562 was always written to the default place.
5564 3. If the message "no IP address found for host xxxx" is generated during
5565 incoming verification, it is now followed by identification of the incoming
5566 connection (so you can more easily find what provoked it).
5568 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5570 5. Added some features to "harden" Exim a bit more against certain attacks:
5572 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5573 be put in Local/Makefile. This is like the never_users runtime option,
5574 but it cannot be overridden. The default setting is "root".
5576 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5577 prefix string with which any file named in a -C command line option
5580 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5581 is retained for -C and -D only if the caller of Exim is root. Without
5582 it, the exim user may also use -C and -D and retain privilege.
5584 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5585 command line option is disabled.
5587 6. Macro names set by the -D option must start with an upper case letter, just
5588 like macro names defined in the configuration file.
5590 7. Added "dereference=" facility to LDAP.
5592 8. Two instances of the typo "uknown" in the source files are fixed.
5594 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5595 the Configure-Makefile script screwed up while processing it.
5597 10. Incorporated PCRE 4.4.
5599 11. The SMTP synchronization check was not operating right at the start of an
5600 SMTP session. For example, it could not catch a HELO sent before the client
5601 waited for the greeting. There is now a check for outstanding input at the
5602 point when the greeting is written. Because of the duplex, asynchronous
5603 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5604 way, but not yet received, when the check is performed.
5606 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5607 on TCP/IP sockets, because this apparently causes some broken clients to
5610 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5611 unchanged) from the Cygwin maintainer.
5613 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5614 was supported instead of testing for mbx. Effectively a typo.
5616 15. The spa authenticator server code was not checking that the input it
5617 received was valid base64.
5619 16. The debug output line for the "set" modifier in ACLs was not showing the
5620 name of the variable that was being set.
5622 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5624 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5625 lookup on its own. Something else was needed to trigger the lookup. For
5626 example, a match in host_lookup or the need for a host name in a host list.
5627 Now, if $sender_host_name is referenced and the host name has not yet been
5628 looked up, a lookup is performed. If the lookup fails, the variable remains
5629 empty, and $host_lookup_failed is set to "1".
5631 19. Added "eqi" as a case-independent comparison operator.
5633 20. The saslauthd authentication condition could segfault if neither service
5634 nor realm was specified.
5636 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5637 error message that was logged was misleading, and incoming SMTP
5638 connections were dropped. The message is now more accurate, and temporary
5639 errors are given to SMTP connections.
5641 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5642 (except RSET) with a 421 error, until QUIT is received. However, it was
5643 failing to send a response to QUIT.
5645 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5646 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5647 ACL is now run after the helo_try_verify_hosts code.
5649 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5650 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5651 case-independent, so other case variants are also recognized). Apparently
5652 some systems use these upper case variants.
5654 25. If more than two messages were waiting for the same host, and a transport
5655 filter was specified for the transport, Exim sent two messages over the
5656 same TCP/IP connection, and then failed with "socket operation on non-
5657 socket" when it tried to send the third.
5659 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5661 27. The extern definition of crypt16() in expand.c was not being excluded when
5662 the OS had its own crypt16() function.
5664 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5665 as a preferred synonym for return_size_limit, both as an option and as an
5668 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5670 30. Changed the default configuration ACL to relax the local part checking rule
5671 for addresses that are not in any local domains. For these addresses,
5672 slashes and pipe symbols are allowed within local parts, but the sequence
5673 /../ is explicitly forbidden.
5675 31. SPA server authentication was not clearing the challenge buffer before
5678 32. log_message in a "warn" ACL statement was writing to the reject log as
5679 well as to the main log, which contradicts the documentation and doesn't
5680 seem right (because no rejection is happening). So I have stopped it.
5682 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5683 However, I am unable to do any testing of this.
5685 34. Fixed an infelicity in the appendfile transport. When checking directories
5686 for a mailbox, to see if any needed to be created, it was accidentally
5687 using path names with one or more superfluous leading slashes; tracing
5688 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5690 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5691 contents are added to the log line that is written for every discarded
5692 recipient. (Previously a log_message setting was ignored.)
5694 36. The ${quote: operator now quotes the string if it is empty.
5696 37. The install script runs exim in order to find its version number. If for
5697 some reason other than non-existence or emptiness, which it checks, it
5698 could not run './exim', it was installing it with an empty version number,
5699 i.e. as "exim-". This error state is now caught, and the installation is
5702 38. An argument was missing from the function that creates an error message
5703 when Exim fails to connect to the socket for saslauthd authentication.
5704 This could cause Exim to crash, or give a corrupted message.
5706 39. Added isip, isip4, and isip6 to ${if conditions.
5708 40. The ACL variables $acl_xx are now saved with the message, and can be
5709 accessed later in routers, transports, and filters.
5711 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5712 strings in the file are not string-expanded.
5714 42. If a MAIL command specified a SIZE value that was too large to fit into an
5715 int variable, the check against message_size_limit failed. Such values are
5716 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5717 day this will have to be increased, but I don't think I want to be around
5718 when emails are that large.
5725 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5726 iconv() is not standard in FreeBSD.
5728 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5729 IPv6 enabled. The observed symptom was a segmentation fault on return from
5730 the function os_common_find_running_interfaces() in src/os.c.
5732 3. In the check_special_case() function in daemon.c I had used "errno" as an
5733 argument name, which causes warnings on some systems. This was basically a
5734 typo, since it was named "eno" in the comments!
5736 4. The code that waits for the clock to tick (at a resolution of some fraction
5737 of a second) so as to ensure message-id uniqueness was always waiting for
5738 at least one whole tick, when it could have waited for less. [This is
5739 almost certainly not relevant at current processor speeds, where it is
5740 unlikely to ever wait at all. But we try to future-proof.]
5742 5. The function that sleeps for a time interval that includes fractions of a
5743 second contained a race. It did not block SIGALRM between setting the
5744 timer, and suspending (a couple of lines later). If the interval was short
5745 and the sigsuspend() was delayed until after it had expired, the suspension
5746 never ended. On busy systems this could lead to processes getting stuck for
5749 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5750 process, before it forks any delivery processes. The open lookup caching
5751 mechanism meant that the open file or database connection was passed into
5752 the delivery process. The problem was that delivery processes always tidy
5753 up cached lookup data. This could cause a problem for the next delivery
5754 process started by the queue runner, because the external queue runner
5755 process does not know about the closure. So the next delivery process
5756 still has data in the lookup cache. In the case of a file lookup, there was
5757 no problem because closing a file descriptor in a subprocess doesn't affect
5758 the parent. However, if the lookup was caching a connection to a database,
5759 the connection was closed, and the second delivery process was likely to
5760 see errors such as "PGSQL: query failed: server closed the connection
5761 unexpectedly". The problem has been fixed by closing all cached lookups
5762 in a queue runner before running a delivery process.
5764 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5765 seem to have the "const" qualifier which it has on other OS. I've
5768 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5769 *of the same type* (client or server) with the same public name that an
5770 error should be diagnosed.
5772 9. When Exim looked up a host name for an IP address, but failed to find the
5773 original IP address when looking up the host name (a safety check), it
5774 output the message "<ip address> does not match any IP for NULL", which was
5775 confusing, to say the least. The bug was that the host name should have
5776 appeared instead of "NULL".
5778 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5779 user that is built into the binary, and the -C or -D options is used, root
5780 privilege is dropped before the configuration file is read. In addition,
5781 logging is switched to stderr instead of the normal log files. If the
5782 configuration then re-defines the Exim user, the unprivileged environment
5783 is probably not what is expected, so Exim logs a panic warning message (but
5786 However, if deliver_drop_privilege is set, the unprivileged state may well
5787 be exactly what is intended, so the warning has been cut out in that case,
5788 and Exim is allowed to try to write to its normal log files.
5794 1. smtp_return_error_details was not giving details for temporary sender
5795 or receiver verification errors.
5797 2. Diagnose a configuration error if two authenticators have the same public
5800 3. Exim used not to create the message log file for a message until the first
5801 delivery attempt. This could be confusing when incoming messages were held
5802 for policy or load reasons. The message log file is now created at the time
5803 the message is received, and an initial "Received" line is written to it.
5805 4. The automatically generated man page for command line options had a minor
5806 bug that caused no ill effects; however, a more serious problem was that
5807 the procedure for building the man page automatically didn't always
5808 operate. Consequently, release 4.20 contains an out-of-date version. This
5809 shouldn't happen again.
5811 5. When building Exim with embedded Perl support, the script that builds the
5812 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5813 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5815 6. The freeze_tell option was not being used for messages that were frozen on
5816 arrival, either by an ACL or by local_scan().
5818 7. Added the smtp_incomplete_transaction log selector.
5820 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5821 was accepting AUTH without a new EHLO.
5823 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5824 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5827 10. Logging of TCP/IP connections (when configured) now happens in the main
5828 daemon process instead of the child process, so that the TCP/IP connection
5829 count is more accurate (but it can never be perfect).
5831 11. The use of "drop" in a nested ACL was not being handled correctly in the
5832 outer ACL. Now, if condition failure induced by the nested "drop" causes
5833 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5834 or "require"), the connection is dropped.
5836 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5837 that yield "discard" can now be used with an "accept" or a "discard" verb,
5838 but an error is generated for any others (because I can't see a useful way
5839 to define what should happen).
5841 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5842 are now processed in the same way as lines in the Exim configuration file.
5843 In particular, continuation lines are supported.
5845 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5847 15. Added -ti meaning -t -i.
5849 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5850 domains, and warn by logging if others are found.
5852 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5853 returned from GIFCONF when Exim is trying to find the list of interfaces on
5854 a host. The code in os.c has been modified to copy each ifreq to an aligned
5855 structure in all cases.
5857 Also, in some cases, the returned ifreq's were being copied to a 'struct
5858 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5859 means the last couple of bytes of an IPv6 address could be chopped if the
5860 ifreq contained only a normal sockaddr (14 bytes storage).
5862 18. Named domain lists were not supported in the hosts_treat_as_local option.
5863 An entry such as +xxxx was not recognized, and was treated as a literal
5866 19. Ensure that header lines added by a DATA ACL are included in the reject log
5867 if the ACL subsequently rejects the message.
5869 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5870 MD5 (which is deprecated).
5872 21. When testing a filter file using -bf, Exim was writing a message when it
5873 took the sender from a "From " line in the message, but it was not doing so
5874 when it took $return_path from a Return-Path: header line. It now does.
5876 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5877 with a valid header line field name (a series of printing characters
5878 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5880 23. Changed "disc" in the source to "disk" to conform to the documentation and
5881 the book and for uniformity.
5883 24. Ignore Sendmail's -Ooption=value command line item.
5885 25. When execve() failed while trying to run a command in a pipe transport,
5886 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5887 could be confused with a return value of 69 from the command itself. This
5888 has been changed to 127, the value the shell returns if it is asked to run
5889 a non-existent command. The wording for the related log line suggests a
5890 non-existent command as the problem.
5892 26. If received_header_text expands to an empty string, do not add a Received:
5893 header line to the message. (Well, it adds a token one on the spool, but
5894 marks it "old" so that it doesn't get used or transmitted.)
5896 27. Installed eximstats 1.28 (addition of -nt option).
5898 28. There was no check for failure on the call to getsockname() in the daemon
5899 code. This can fail if there is a shortage of resources on the system, with
5900 ENOMEM, for example. A temporary error is now given on failure.
5902 29. Contrary to the C standard, it seems that in some environments, the
5903 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5904 program. Exim now does this explicitly; it affects the formatting of
5905 timestamps using strftime().
5907 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5908 complaints. I've now initialized all the variables, to avoid this.
5910 32. Header lines added by a system filter were not being "seen" during
5911 transport-time rewrites.
5913 33. The info_callback() function passed to OpenSSL is set up with type void
5914 (*)(SSL *, int, int), as described somewhere. However, when calling the
5915 function (actually a macro) that sets it up, the type void(*)() is
5916 expected. I've put in a cast to prevent warnings from picky compilers.
5918 34. If a DNS black list lookup found a CNAME record, but there were no A
5919 records associated with the domain it pointed at, Exim crashed.
5921 35. If a DNS black list lookup returned more than one A record, Exim ignored
5922 all but the first. It now scans all returned addresses if a particular IP
5923 value is being sought. In this situation, the contents of the
5924 $dnslist_value variable are a list of all the addresses, separated by a
5927 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5928 to accept a host name and all its aliases if the forward lookup for any of
5929 them yielded the IP address of the incoming connection. Now it accepts only
5930 those names whose forward lookup yields the correct IP address. Any other
5931 names are discarded. This closes a loophole whereby a rogue DNS
5932 administrator could create reverse DNS records to break through a
5933 wildcarded host restriction in an ACL.
5935 37. If a user filter or a system filter that ran in a subprocess used any of
5936 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5937 the wrong values were passed to the pipe command ($thisaddress had the
5938 value of $0, $0 had the value of $1, etc). This bug was introduced by
5939 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5941 38. Improved the line breaking for long SMTP error messages from ACLs.
5942 Previously, if there was no break point between 40 and 75 characters, Exim
5943 left the rest of the message alone. Two changes have been made: (a) I've
5944 reduced the minimum length to 35 characters; (b) if it can't find a break
5945 point between 35 and 75 characters, it looks ahead and uses the first one
5946 that it finds. This may give the occasional overlong line, but at least the
5947 remaining text gets split now.
5949 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5950 file descriptors might be low, and that setting 1000 would always raise it.
5951 It turns out that in some environments, the limit is already over 1000 and
5952 that lowering it causes trouble. So now Exim takes care not to decrease it.
5954 40. When delivering a message, the value of $return_path is set to $sender_
5955 address at the start of routing (routers may change the value). By an
5956 oversight, this default was not being set up when an address was tested by
5957 -bt or -bv, which affected the outcome if any router or filter referred to
5960 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5961 log only when "message" or "log_message" are set. However, if one of the
5962 conditions was an address verification, or a call to a nested ACL, the
5963 messages generated by the underlying test were being passed through. This
5964 no longer happens. The underlying message is available in $acl_verify_
5965 message for both "message" and "log_message" expansions, so it can be
5966 passed through if needed.
5968 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5969 new expansion $bh_ to give the encoded byte string without charset
5970 translation. Translation happens only if iconv() is available; HAVE_ICONV
5971 indicates this at build time. HEADERS_CHARSET gives the charset to
5972 translate to; headers_charset can change it in the configuration, and
5973 "headers charset" can change it in an individual filter file.
5975 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5976 that creates RFC 2047 encoded "words" labels them as that charset instead
5977 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5978 expansion operator; (ii) when Exim creates a From: line for a local
5979 message; (iii) when a header line is rewritten to include a "phrase" part.
5981 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5982 buggy, causing it to skip the first lines of messages whose message ID
5983 ended in 'D'. This would not have bitten before Exim release 4.14, because
5984 message IDs were unlikely to end in 'D' before then. The effect was to have
5985 incorrect size information for certain domains.
5987 45. #include "config.h" was missing at the start of the crypt16.c module. This
5988 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5991 46. If there was a timeout during a "random" callout check, Exim treated it as
5992 a failure of the random address, and carried on sending RSET and the real
5993 address. If the delay was just some slowness somewhere, the response to the
5994 original RCPT would be taken as a response to RSET and so on, causing
5995 mayhem of various kinds.
5997 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5998 when I implemented it. It didn't allow for the fact that some option values
5999 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6000 the right test for positive values.
6002 48. Domain names in DNS records are case-independent. Exim always looks them up
6003 in lower case. Some resolvers return domain names in exactly the case they
6004 appear in the zone file, that is, they may contain uppercase letters. Not
6005 all resolvers do this - some return always lower case. Exim was treating a
6006 change of case by a resolver as a change of domain, similar to a widening
6007 of a domain abbreviation. This triggered its re-routing code and so it was
6008 trying to route what was effectively the same domain again. This normally
6009 caused routing to fail (because the router wouldn't handle the domain
6010 twice). Now Exim checks for this case specially, and just changes the
6011 casing of the domain that it ultimately uses when it transmits the message
6014 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6017 50. If a filter generated a file delivery with a non-absolute name (possible if
6018 no home directory exists for the router), the forbid_file option was not
6021 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6022 the existing equality matching.
6024 52. Exim was using ints instead of ino_t variables in some places where it was
6025 dealing with inode numbers.
6027 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6028 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6029 called TMPDIR, and if it finds it is different, it changes its value.
6031 54. The smtp_printf() function is now made available to local_scan() so
6032 additional output lines can be written before returning. There is also an
6033 smtp_fflush() function to enable the detection of a dropped connection.
6034 The variables smtp_input and smtp_batched_input are exported to
6037 55. Changed the default runtime configuration: the message "Unknown user"
6038 has been removed from the ACL, and instead placed on the localuser router,
6039 using the cannot_route_message feature. This means that any verification
6040 failures that generate their own messages won't get overridden. Similarly,
6041 the "Unrouteable address" message that was in the ACL for unverifiable
6042 relay addresses has also been removed.
6044 56. Added hosts_avoid_esmtp to the smtp transport.
6046 57. The exicyclog script was not checking for the esoteric option
6047 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6048 will work only if exicyclog is run under the appropriate euid.
6050 58. Following a discussion on the list, the rules by which Exim recognises line
6051 endings on incoming messages have been changed. The -dropcr and drop_cr
6052 options are now no-ops, retained only for backwards compatibility. The
6053 following line terminators are recognized: LF CRLF CR. However, special
6054 processing applies to CR:
6056 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6057 nor a local message in the state where . is a terminator.
6059 (ii) If a bare CR is encountered in a header line, an extra space is added
6060 after the line terminator so as not to end the header. The reasoning
6061 behind this is that bare CRs in header lines are most likely either
6062 to be mistakes, or people trying to play silly games.
6064 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6065 was being incorrectly given as 18 bytes larger than it should have been.
6066 This is a VOB (very old bug).
6068 60. This may never have affected anything current, but just in case it has:
6069 When the local host is found other than at the start of a list of hosts,
6070 the local host, those with the same MX, and any that follow, are discarded.
6071 When the list in question was part of a longer list of hosts, the following
6072 hosts (not currently being processed) were also being discarded. This no
6073 longer happens. I'm not sure if this situation could ever has previously
6076 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6079 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6080 header. This is something that is recommended in a new Internet Draft, and
6081 is something that is documented as being done by Sendmail. There are two
6082 possible values. For messages generated by the autoreply transport, Exim
6085 Auto-Submitted: auto-replied
6087 whereas for all other generated messages (e.g. bounces) it adds
6089 Auto-Submitted: auto-generated
6091 63. The "personal" condition in filters now includes a test for the
6092 Auto-Submitted: header. If it contains the string "auto-" the message it
6093 not considered personal.
6095 64. Added rcpt_include_affixes as a generic transport option.
6097 65. Added queue_only_override (default true).
6099 66. Added the syslog_duplication option.
6101 67. If what should have been the first header line of a message consisted of
6102 a space followed by a colon, Exim was mis-interpreting it as a header line.
6103 It isn't of course - it is syntactically invalid and should therefore be
6104 treated as the start of the message body. The misbehaviour could have
6105 caused a number of strange effects, including loss of data in subsequent
6106 header lines, and spool format errors.
6108 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6109 client host had authenticated. This control can now be exercised by an ACL
6110 for more flexibility.
6112 69. By default, callouts do not happen when testing with -bh. There is now a
6113 variant, -bhc, which does actually run the callout code, including
6114 consulting and updating the callout cache.
6116 70. Added support for saslauthd authentication, courtesy of Alexander
6119 71. If statvfs() failed on the spool or log directories while checking their
6120 size for availability, Exim confusingly gave the error "space shortage".
6121 Furthermore, in debugging mode it crashed with a floating point exception.
6122 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6123 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6124 serious problem, Exim now writes to the main and panic logs when this
6125 happens, with details of the failure. It then refuses to accept the
6126 incoming message, giving the message "spool directory problem" or "log
6127 directory problem" with a 421 code for SMTP messages.
6129 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6130 0, 1, and 2 exist, because some OS complain for execs without them (see
6131 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6132 descriptors. However, the code omitted to check that the open succeeded,
6133 causing mysterious errors if for some reason the permissions on /dev/null
6134 got screwed. Now Exim writes a message to the main and panic logs, and
6135 bombs out if it can't open /dev/null.
6137 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6138 interact so that it is all more flexible. It is supposed to remain
6139 backwards compatible. Also added extra_local_interfaces.
6141 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6142 to bomb out with an assertion failure - to the client this appears as a
6143 connection drop. This problem occurs in the part of the code that was taken
6144 from the Samba project. Fortunately, the assertion is in a very simple
6145 function, so I have fixed this by reproducing the function inline in the
6146 one place where it is called, and arranging for authentication to fail
6147 instead of killing the process with assert().
6149 75. The SPA client code was not working when the server requested OEM rather
6150 than Unicode encoding.
6152 76. Added code to make require_files with a specific uid setting more usable in
6153 the case where statting the file as root fails - usually a non-root-mounted
6154 NFS file system. When this happens and the failure is EACCES, Exim now
6155 forks a subprocess and does the per-uid checking as the relevant uid.
6157 77. Added process_log_path.
6159 78. If log_file_path was not explicitly set, a setting of check_log_space or
6160 check_log_inodes was ignored.
6162 79. If a space check for the spool or log partitions fails, the incident is now
6163 logged. Of course, in the latter case the data may get lost...
6165 80. Added the %p formatting code to string_format() so that it can be used to
6166 print addresses in debug_print(). Adjusted all the address printing in the
6167 debugging in store.c to use %p rather than %d.
6169 81. There was a concern that a line of code in smtp_in.c could overflow a
6170 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6171 initially expressed, the concern was not well-founded, because trailing
6172 spaces are removed early. However, if the trailing spaces were followed by
6173 a NULL, they did not get removed, so the overflow was possible. Two fixes
6176 (a) I re-wrote the offending code in a cleaner fashion.
6177 (b) If an incoming SMTP command contains a NULL character, it is rejected
6180 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6181 runs initgroups(), so that if the Exim user is in any additional groups,
6182 they will be used during message reception.
6188 The change log for 4.20 and earlier releases has been archived.
projects / user / henk / code / exim.git / blob