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
13 TL/01 When running the test suite, the README says that variables such as
14 no_msglog_check are global and can be placed anywhere in a specific
15 test's script, however it was observed that placement needed to be near
16 the beginning for it to behave that way. Changed the runtest perl
17 script to read through the entire script once to detect and set these
18 variables, reset to the beginning of the script, and then run through
19 the script parsing/test process like normal.
21 TL/02 The BSD's have an arc4random API. One of the functions to induce
22 adding randomness was arc4random_stir(), but it has been removed in
23 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
24 function when detected.
26 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
27 cause callback expansion.
29 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
30 syntax errors in an expansion can be treated as a string instead of
31 logging or causing an error, due to the internal use of bool_lax
32 instead of bool when processing it.
34 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
35 server certificates when making smtp deliveries.
37 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
39 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
41 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
43 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
44 Merged patch from Sebastian Wiedenroth.
46 JH/05 Fix results-pipe from transport process. Several recipients, combined
47 with certificate use, exposed issues where response data items split
48 over buffer boundaries were not parsed properly. This eventually
49 resulted in duplicates being sent. This issue only became common enough
50 to notice due to the introduction of conection certificate information,
51 the item size being so much larger. Found and fixed by Wolfgang Breyha.
53 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
54 size buffer was used, resulting in syntax errors when an expansion
57 JH/07 Add support for directories of certificates when compiled with a GnuTLS
58 version 3.3.6 or later.
60 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
61 is EXPERIMENTAL_EVENT, the main-configuration and transport options
62 both become "event_action", the variables become $event_name, $event_data
63 and $event_defer_errno. There is a new variable $verify_mode, usable in
64 routers, transports and related events. The tls:cert event is now also
65 raised for inbound connections, if the main configuration event_action
68 TL/06 In test suite, disable OCSP for old versions of openssl which contained
69 early OCSP support, but no stapling (appears to be less than 1.0.0).
71 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
72 server certificate names available under the smtp transport option
73 "tls_verify_cert_hostname" now do not permit multi-component wildcard
76 JH/10 Time-related extraction expansions from certificates now use the main
77 option "timezone" setting for output formatting, and are consistent
78 between OpenSSL and GnuTLS compilations. Bug 1541.
80 JH/11 Fix a crash in mime ACL when meeting a zero-length parameter in the
83 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
84 include certificate info, eximon was claiming there were spoolfile
87 JH/13 Buf 1521: Fix ldap lookup for single-attr request, multiple-attr return.
89 JH/14 Log delivery-related information more consistently, using the sequence
90 "H=<name> [<ip>]" wherever possible.
95 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
96 checkers that were complaining about end of non-void function with no
99 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
100 This was a regression intruduced in 4.83 by another bugfix.
102 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
104 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
105 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
111 TF/01 Correctly close the server side of TLS when forking for delivery.
113 When a message was received over SMTP with TLS, Exim failed to clear up
114 the incoming connection properly after forking off the child process to
115 deliver the message. In some situations the subsequent outgoing
116 delivery connection happened to have the same fd number as the incoming
117 connection previously had. Exim would try to use TLS and fail, logging
118 a "Bad file descriptor" error.
120 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
121 utilities have not been installed.
123 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
124 temporary space as the ACL may create new global variables.
126 TL/01 LDAP support uses per connection or global context settings, depending
127 upon the detected version of the libraries at build time.
129 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
130 to extract and use the src ip:port in logging and expansions as if it
131 were a direct connection from the outside internet. PPv2 support was
132 updated based on HAProxy spec change in May 2014.
134 JH/02 Add ${listextract {number}{list}{success}{fail}}.
136 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
137 Properly escape header and check for NULL return.
139 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
142 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
144 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
145 characters in header names, implemented as a verify condition.
146 Contributed by Michael Fischer v. Mollard.
148 TL/05 Rename SPF condition results err_perm and err_temp to standardized
149 results permerror and temperror. Previous values are deprecated but
150 still accepted. In a future release, err_perm and err_temp will be
151 completely removed, which will be a backward incompatibility if the
152 ACL tests for either of these two old results. Patch contributed by
153 user bes-internal on the mailing list.
155 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
157 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
158 selectors, in both main and reject logs.
160 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
163 JH/07 Add malware type "sock" for talking to simple daemon.
165 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
167 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
168 routers/transports under cutthrough routing.
170 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
171 numbers. Touch up "bool" conditional to keep the same definition.
173 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
175 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
177 JH/12 Expand items in router/transport headers_add or headers_remove lists
178 individually rather than the list as a whole. Bug 1452.
180 Required for reasonable handling of multiple headers_ options when
181 they may be empty; requires that headers_remove items with embedded
182 colons must have them doubled (or the list-separator changed).
184 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
185 view the policy declared in the DMARC record. Currently, $dmarc_status
186 is a combined value of both the record presence and the result of the
189 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
191 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
192 dnslookup router and the smtp transport (applying to the forward
195 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
196 of ldap servers used for a specific lookup. Patch provided by Heiko
199 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
200 New variable $lookup_dnssec_authenticated for observability.
202 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
203 Patch submitted by Lars Timman.
205 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
207 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
208 Requires trusted mode and valid format message id, aborts otherwise.
209 Patch contributed by Heiko Schlichting.
211 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
212 certextract with support for various fields. Bug 1358.
214 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
215 is requested by default, modifiable by smtp transport option
218 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
219 operate on certificate variables to give certificate fingerprints
220 Also new ${sha256:cert_variable}.
222 JH/23 The PRDR feature is moved from being Experimental into the mainline.
224 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
225 Christian Aistleitner.
227 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
229 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
230 file. Patch from Wolfgang Breyha.
232 JH/25 Expand the coverage of the delivery $host and $host_address to
233 client authenticators run in verify callout. Bug 1476.
235 JH/26 Port service names are now accepted for tls_on_connect_ports, to
236 align with daemon_smtp_ports. Bug 72.
238 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
239 support and error reporting did not work properly.
241 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
242 and is readable. Patch from Andrew Colin Kissa.
244 TL/14 Enhance documentation of ${run expansion and how it parses the
245 commandline after expansion, particularly in the case when an
246 unquoted variable expansion results in an empty value.
248 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
250 PP/02 Fix internal collision of T_APL on systems which support RFC3123
251 by renaming away from it. Addresses GH issue 15, reported by
254 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
256 TL/15 SECURITY: prevent double expansion in math comparison functions
257 (can expand unsanitized data). Not remotely exploitable.
264 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
266 PP/02 Make -n do something, by making it not do something.
267 When combined with -bP, the name of an option is not output.
269 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
272 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
273 $sender_host_name and config options to manage this, and basic check
276 PP/05 DSCP support for outbound connections and control modifier for inbound.
278 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
279 (Only plugin which currently uses this is kerberos4, which nobody should
280 be using, but we should make it available and other future plugins might
281 conceivably use it, even though it would break NAT; stuff *should* be
282 using channel bindings instead).
284 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
285 name; added for Sendmail compatibility; requires admin caller.
286 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
287 just ignore it); requires trusted caller.
288 Also parse but ignore: -Ac -Am -X<logfile>
291 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
293 TL/02 Add +smtp_confirmation as a default logging option.
295 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
296 Patch by Magnus Holmgren from 2007-02-20.
298 TL/04 Bugzilla 1281 - Spec typo.
299 Bugzilla 1283 - Spec typo.
300 Bugzilla 1290 - Spec grammar fixes.
302 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
304 TL/06 Add Experimental DMARC support using libopendmarc libraries.
306 TL/07 Fix an out of order global option causing a segfault. Reported to dev
307 mailing list by by Dmitry Isaikin.
309 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
311 JH/02 Support "G" suffix to numbers in ${if comparisons.
313 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
315 NM/01 Bugzilla 1197 - Spec typo
316 Bugzilla 1196 - Spec examples corrections
318 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
320 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
321 gnutls_enable_pkcs11, but renamed to more accurately indicate its
324 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
325 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
327 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
328 "acl {{name}{arg}...}", and optional args on acl condition
331 JH/05 Permit multiple router/transport headers_add/remove lines.
333 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
335 JH/07 Avoid using a waiting database for a single-message-only transport.
336 Performance patch from Paul Fisher. Bugzilla 1262.
338 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
341 JH/09 Add $headers_added variable, with content from use of ACL modifier
342 add_header (but not yet added to the message). Bugzilla 199.
344 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
345 Pulled from Bugzilla 817 by Wolfgang Breyha.
347 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
349 (nb: this is the same fix as in Exim 4.80.1)
351 JH/11 Add A= logging on delivery lines, and a client_set_id option on
354 JH/12 Add optional authenticated_sender logging to A= and a log_selector
357 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
359 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
360 advertise SMTP AUTH mechanism to us, instead of a generic
361 protocol violation error. Also, make Exim more robust to bad
362 data from the Dovecot auth socket.
364 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
366 When a queue runner is handling a message, Exim first routes the
367 recipient addresses, during which it prunes them based on the retry
368 hints database. After that it attempts to deliver the message to
369 any remaining recipients. It then updates the hints database using
372 So if a recipient address works intermittently, it can get repeatedly
373 deferred at routing time. The retry hints record remains fresh so the
374 address never reaches the final cutoff time.
376 This is a fairly common occurrence when a user is bumping up against
377 their storage quota. Exim had some logic in its local delivery code
378 to deal with this. However it did not apply to per-recipient defers
379 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
381 This change adds a proper retry rule check during routing so that the
382 final cutoff time is checked against the message's age. We only do
383 this check if there is an address retry record and there is not a
384 domain retry record; this implies that previous attempts to handle
385 the address had the retry_use_local_parts option turned on. We use
386 this as an approximation for the destination being like a local
387 delivery, as in LMTP.
389 I suspect this new check makes the old local delivery cutoff check
390 redundant, but I have not verified this so I left the code in place.
392 TF/02 Correct gecos expansion when From: is a prefix of the username.
394 Test 0254 submits a message to Exim with the header
398 When I ran the test suite under the user fanf2, Exim expanded
399 the header to contain my full name, whereas it should have added
400 a Resent-Sender: header. It erroneously treats any prefix of the
401 username as equal to the username.
403 This change corrects that bug.
405 GF/01 DCC debug and logging tidyup
406 Error conditions log to paniclog rather than rejectlog.
407 Debug lines prefixed by "DCC: " to remove any ambiguity.
409 TF/03 Avoid unnecessary rebuilds of lookup-related code.
411 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
412 Bug spotted by Jeremy Harris; was flawed since initial commit.
413 Would have resulted in OCSP responses post-SNI triggering an Exim
414 NULL dereference and crash.
416 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
418 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
419 Bug detection, analysis and fix by Samuel Thibault.
420 Bugzilla 1331, Debian bug #698092.
422 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
424 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
425 Server implementation by Todd Lyons, client by JH.
426 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
427 config variable "prdr_enable" controls whether the server
428 advertises the facility. If the client requests PRDR a new
429 acl_data_smtp_prdr ACL is called once for each recipient, after
430 the body content is received and before the acl_smtp_data ACL.
431 The client is controlled by bolth of: a hosts_try_prdr option
432 on the smtp transport, and the server advertisement.
433 Default client logging of deliveries and rejections involving
434 PRDR are flagged with the string "PRDR".
436 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
437 fclose(). Diagnosis by Todd Lyons.
439 PP/17 Update configure.default to handle IPv6 localhost better.
440 Patch by Alain Williams (plus minor tweaks).
443 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
444 This is now consistent with GnuTLS, and is now documented: the
445 previous undocumented portable approach to treating the option as
446 unset was to force an expansion failure. That still works, and
447 an empty string is now equivalent.
449 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
450 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
451 not performing validation itself.
453 PP/20 Added force_command boolean option to pipe transport.
454 Patch from Nick Koston, of cPanel Inc.
456 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
459 TF/04 Added udpsend ACL modifer and hexquote expansion operator
461 PP/21 Fix eximon continuous updating with timestamped log-files.
462 Broken in a format-string cleanup in 4.80, missed when I repaired the
463 other false fix of the same issue.
464 Report and fix from Heiko Schlichting.
467 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
468 Report from Prashanth Katuri.
470 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
471 It's SecureTransport, so affects any MacOS clients which use the
472 system-integrated TLS libraries, including email clients.
474 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
475 using a MIME ACL for non-SMTP local injection.
476 Report and assistance in diagnosis by Warren Baker.
478 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
480 JH/16 Fix comparisons for 64b. Bugzilla 1385.
482 TL/09 Add expansion variable $authenticated_fail_id to keep track of
483 last id that failed so it may be referenced in subsequent ACL's.
485 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
488 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
489 ldap library initialization, allowing self-signed CA's to be
490 used. Also properly sets require_cert option later in code by
491 using NULL (global ldap config) instead of ldap handle (per
492 session). Bug diagnosis and testing by alxgomz.
494 TL/12 Enhanced documentation in the ratelimit.pl script provided in
495 the src/util/ subdirectory.
497 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
498 renamed to Transport Post Delivery Action by Jeremy Harris, as
501 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
502 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
503 redis_servers = needs to be configured which will be used by the redis
504 lookup. Patch from Warren Baker, of The Packet Hub.
506 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
508 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
509 hostname or reverse DNS when processing a host list. Used suggestions
510 from multiple comments on this bug.
512 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
514 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
515 Missed a few lines, added it to make the runtest require no keyboard
518 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
519 contains upper case chars. Make router use caseful_local_part.
521 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
522 support when GnuTLS has been built with p11-kit.
528 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
530 This, or similar/improved, will also be change PP/11 of 4.82.
536 PP/01 Handle short writes when writing local log-files.
537 In practice, only affects FreeBSD (8 onwards).
538 Bugzilla 1053, with thanks to Dmitry Isaikin.
540 NM/01 Bugzilla 949 - Documentation tweak
542 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
545 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
547 PP/02 Implemented gsasl authenticator.
549 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
551 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
552 `pkg-config foo` for cflags/libs.
554 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
555 with rest of GSASL and with heimdal_gssapi.
557 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
558 `pkg-config foo` for cflags/libs for the TLS implementation.
560 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
561 properties get this fed in as external SSF. A number of robustness
562 and debugging improvements to the cyrus_sasl authenticator.
564 PP/08 cyrus_sasl server now expands the server_realm option.
566 PP/09 Bugzilla 1214 - Log authentication information in reject log.
567 Patch by Jeremy Harris.
569 PP/10 Added dbmjz lookup type.
571 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
573 PP/12 MAIL args handles TAB as well as SP, for better interop with
574 non-compliant senders.
575 Analysis and variant patch by Todd Lyons.
577 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
578 Bug report from Lars Müller <lars@samba.org> (via SUSE),
579 Patch from Dirk Mueller <dmueller@suse.com>
581 PP/13 tls_peerdn now print-escaped for spool files.
582 Observed some $tls_peerdn in wild which contained \n, which resulted
583 in spool file corruption.
585 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
586 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
587 or write after TLS renegotiation, which otherwise led to messages
590 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
591 as a tracking header (ie: a signed header comes before the signature).
592 Patch from Wolfgang Breyha.
594 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
595 comma-sep list; embedded commas doubled.
597 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
599 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
601 Report and patch from Dmitry Banschikov.
603 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
604 Removed SSL_clear() after SSL_new() which led to protocol negotiation
605 failures. We appear to now support TLS1.1+ with Exim.
607 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
608 lets Exim select keys and certificates based upon TLS SNI from client.
609 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
610 before an outbound SMTP session. New log_selector, +tls_sni.
612 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
613 NULL dereference. Report and patch from Alun Jones.
615 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
616 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
617 Not seeing resolver debug output on NetBSD, but suspect this is a
618 resolver implementation change.
620 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
621 Left warnings. Added "eximon gdb" invocation mode.
623 PP/21 Defaulting "accept_8bitmime" to true, not false.
625 PP/22 Added -bw for inetd wait mode support.
627 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
628 locate the relevant includes and libraries. Made this the default.
630 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
631 Bugzilla 1246, report and most of solution from Tomasz Kusy.
633 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
634 This may cause build issues on older platforms.
636 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
637 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
638 gnutls_require_mac & gnutls_require_protocols (no longer supported).
639 Added SNI support via GnuTLS too.
640 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
642 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
644 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
645 Applied second patch from Janne, implementing suggestion to default
646 multiple-strings-in-record handling to match SPF spec.
648 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
650 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
651 read-only, out of scope).
652 Patch from Wolfgang Breyha, report from Stuart Northfield.
654 PP/29 Fix three issues highlighted by clang analyser static analysis.
655 Only crash-plausible issue would require the Cambridge-specific
656 iplookup router and a misconfiguration.
657 Report from Marcin Mirosław.
659 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
661 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
662 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
663 As part of this, removing so much warning spew let me fix some minor
664 real issues in debug logging.
666 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
667 assignment on my part. Fixed.
669 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
670 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
671 Janne Snabb (who went above and beyond: thank you).
673 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
674 string otherwise requires a connection and a bunch more work and it's
675 relatively easy to get wrong. Should also expose TLS library linkage
678 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
679 64-bit ${eval} (JH/03).
681 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
682 GNU libc to support some of the 64-bit stuff, should not lead to
683 conflicts. Defined before os.h is pulled in, so if a given platform
684 needs to override this, it can.
686 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
687 protection layer was required, which is not implemented.
688 Bugzilla 1254, patch from Wolfgang Breyha.
690 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
691 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
692 tls_dhparam take prime identifiers. Also unbreak combination of
693 OpenSSL+DH_params+TLSSNI.
695 PP/39 Disable SSLv2 by default in OpenSSL support.
701 PP/01 Solaris build fix for Oracle's LDAP libraries.
702 Bugzilla 1109, patch from Stephen Usher.
704 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
706 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
709 TF/02 Fix a couple more cases where we did not log the error message
710 when unlink() failed. See also change 4.74-TF/03.
712 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
713 lock up or crash if it happened to be inside a call to libc when it
714 got a SIGUSR1 from exiwhat.
716 The SIGUSR1 handler appends the current process status to the process
717 log which is later printed by exiwhat. It used to use the general
718 purpose logging code to do this, but several functions it calls are
719 not safe for signals.
721 The new output code in the SIGUSR1 handler is specific to the process
722 log, and simple enough that it's easy to inspect for signal safety.
723 Removing some special cases also simplifies the general logging code.
724 Removing the spurious timestamps from the process log simplifies
727 TF/04 Improved ratelimit ACL condition.
729 The /noupdate option has been deprecated in favour of /readonly which
730 has clearer semantics. The /leaky, /strict, and /readonly update modes
731 are mutually exclusive. The update mode is no longer included in the
732 database key; it just determines when the database is updated. (This
733 means that when you upgrde Exim will forget old rate measurements.)
735 Exim now checks that the per_* options are used with an update mode that
736 makes sense for the current ACL. For example, when Exim is processing a
737 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
738 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
739 must specify per_mail/readonly. If you omit the update mode it defaults to
740 /leaky where that makes sense (as before) or /readonly where required.
742 The /noupdate option is now undocumented but still supported for
743 backwards compatibility. It is equivalent to /readonly except that in
744 ACLs where /readonly is required you may specify /leaky/noupdate or
745 /strict/noupdate which are treated the same as /readonly.
747 A useful new feature is the /count= option. This is a generalization
748 of the per_byte option, so that you can measure the throughput of other
749 aggregate values. For example, the per_byte option is now equivalent
750 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
752 The per_rcpt option has been generalized using the /count= mechanism
753 (though it's more complicated than the per_byte equivalence). When it is
754 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
755 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
756 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
757 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
758 Note that using per_rcpt with a non-readonly update mode in more than
759 one ACL will cause the recipients to be double-counted. (The per_mail
760 and per_byte options don't have this problem.)
762 The handling of very low rates has changed slightly. If the computed rate
763 is less than the event's count (usually one) then this event is the first
764 after a long gap. In this case the rate is set to the same as this event's
765 count, so that the first message of a spam run is counted properly.
767 The major new feature is a mechanism for counting the rate of unique
768 events. The new per_addr option counts the number of different
769 recipients that someone has sent messages to in the last time period. It
770 behaves like per_rcpt if all the recipient addresses are different, but
771 duplicate recipient addresses do not increase the measured rate. Like
772 the /count= option this is a general mechanism, so the per_addr option
773 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
774 example, measure the rate that a client uses different sender addresses
775 with the options per_mail/unique=$sender_address. There are further
776 details in the main documentation.
778 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
780 TF/06 Removed a few PCRE remnants.
782 TF/07 Automatically extract Exim's version number from tags in the git
783 repository when doing development or release builds.
785 PP/02 Raise smtp_cmd_buffer_size to 16kB.
786 Bugzilla 879. Patch from Paul Fisher.
788 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
789 Heavily based on revision 40f9a89a from Simon Arlott's tree.
792 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
794 PP/05 Variable $av_failed, true if the AV scanner deferred.
795 Bugzilla 1078. Patch from John Horne.
797 PP/06 Stop make process more reliably on build failure.
798 Bugzilla 1087. Patch from Heiko Schlittermann.
800 PP/07 Make maildir_use_size_file an _expandable_ boolean.
801 Bugzilla 1089. Patch from Heiko Schlittermann.
803 PP/08 Handle ${run} returning more data than OS pipe buffer size.
804 Bugzilla 1131. Patch from Holger Weiß.
806 PP/09 Handle IPv6 addresses with SPF.
807 Bugzilla 860. Patch from Wolfgang Breyha.
809 PP/10 GnuTLS: support TLS 1.2 & 1.1.
811 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
814 PP/11 match_* no longer expand right-hand-side by default.
815 New compile-time build option, EXPAND_LISTMATCH_RHS.
816 New expansion conditions, "inlist", "inlisti".
818 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
820 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
822 PP/14 fix log_write() format string regression from TF/03.
823 Bugzilla 1152. Patch from Dmitry Isaikin.
829 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
831 PP/02 Harmonised TLS library version reporting; only show if debugging.
832 Layout now matches that introduced for other libraries in 4.74 PP/03.
834 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
836 PP/04 New "dns_use_edns0" global option.
838 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
841 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
842 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
844 TK/01 Updated PolarSSL code to 0.14.2.
845 Bugzilla 1097. Patch from Andreas Metzler.
847 PP/07 Catch divide-by-zero in ${eval:...}.
850 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
853 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
854 format-string attack -- SECURITY: remote arbitrary code execution.
856 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
857 time unintentionally subject to list matching rules, letting the header
858 cause arbitrary Exim lookups (of items which can occur in lists, *not*
859 arbitrary string expansion). This allowed for information disclosure.
861 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
862 INT_MIN/-1 -- value coerced to INT_MAX.
868 NM/01 Workround for PCRE version dependency in version reporting
871 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
872 This fixes portability to compilers other than gcc, notably
873 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
875 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
876 makefiles for portability to HP-UX and POSIX correctness.
878 PP/01 Permit LOOKUP_foo enabling on the make command-line.
879 Also via indented variable definition in the Makefile.
880 (Debugging by Oliver Heesakkers).
882 PP/02 Restore caching of spamd results with expanded spamd_address.
883 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
885 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
886 Improves build reliability. Fix from: Frank Elsner
888 NM/02 Fix wide character breakage in the rfc2047 coding
889 Fixes bug 1064. Patch from Andrey N. Oktyabrski
891 NM/03 Allow underscore in dnslist lookups
892 Fixes bug 1026. Patch from Graeme Fowler
894 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
895 Code patches from Adam Ciarcinski of NetBSD.
897 NM/04 Fixed exiqgrep to cope with mailq missing size issue
900 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
901 is logged, to avoid truncation. Patch from John Horne.
903 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
904 Patch from Jakob Hirsch.
906 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
907 SQL string expansion failure details.
908 Patch from Andrey Oktyabrski.
910 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
911 Patch from Simon Arlott.
913 PP/09 New lookups functionality failed to compile on old gcc which rejects
914 extern declarations in function scope.
915 Patch from Oliver Fleischmann
917 PP/10 Use sig_atomic_t for flags set from signal handlers.
918 Check getgroups() return and improve debugging.
919 Fixed developed for diagnosis in bug 927 (which turned out to be
922 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
923 Patch from Mark Zealey.
925 PP/12 Bugzilla 1056: Improved spamd server selection.
926 Patch from Mark Zealey.
928 PP/13 Bugzilla 1086: Deal with maildir quota file races.
929 Based on patch from Heiko Schlittermann.
931 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
932 Patch from Uwe Doering, sign-off by Michael Haardt.
934 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
935 variable declaration deep within a block. Bug and patch from
938 PP/15 lookups-Makefile IRIX compatibilty coercion.
940 PP/16 Make DISABLE_DKIM build knob functional.
942 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
943 Patch by Simon Arlott
945 TF/03 Fix valgrind.h portability to C89 compilers that do not support
946 variable argument macros. Our copy now differs from upstream.
952 TF/01 Failure to get a lock on a hints database can have serious
953 consequences so log it to the panic log.
955 TF/02 Log LMTP confirmation messages in the same way as SMTP,
956 controlled using the smtp_confirmation log selector.
958 TF/03 Include the error message when we fail to unlink a spool file.
960 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
961 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
962 for maintaining out-of-tree patches for some time.
964 PP/01 Bugzilla 139: Documentation and portability issues.
965 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
966 Handle per-OS dynamic-module compilation flags.
968 PP/02 Let /dev/null have normal permissions.
969 The 4.73 fixes were a little too stringent and complained about the
970 permissions on /dev/null. Exempt it from some checks.
971 Reported by Andreas M. Kirchwitz.
973 PP/03 Report version information for many libraries, including
974 Exim version information for dynamically loaded libraries. Created
975 version.h, now support a version extension string for distributors
976 who patch heavily. Dynamic module ABI change.
978 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
979 privilege escalation vulnerability whereby the Exim run-time user
980 can cause root to append content of the attacker's choosing to
983 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
986 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
987 If dropping privileges for untrusted macros, we disabled normal logging
988 on the basis that it would fail; for the Exim run-time user, this is not
989 the case, and it resulted in successful deliveries going unlogged.
990 Fixed. Reported by Andreas Metzler.
996 PP/01 Date: & Message-Id: revert to normally being appended to a message,
997 only prepend for the Resent-* case. Fixes regression introduced in
998 Exim 4.70 by NM/22 for Bugzilla 607.
1000 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1001 increasing numbers of administrators be bitten by this.
1003 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1005 PP/03 Bugzilla 994: added openssl_options main configuration option.
1007 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1009 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1011 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1013 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1014 without a peer certificate, leading to a segfault because of an
1015 assumption that peers always have certificates. Be a little more
1016 paranoid. Problem reported by Martin Tscholak.
1018 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1019 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1020 NB: ClamAV planning to remove STREAM in "middle of 2010".
1021 CL also introduces -bmalware, various -d+acl logging additions and
1022 more caution in buffer sizes.
1024 PP/09 Implemented reverse_ip expansion operator.
1026 PP/10 Bugzilla 937: provide a "debug" ACL control.
1028 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1030 PP/12 Bugzilla 973: Implement --version.
1032 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1034 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1036 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1038 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1039 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1040 ignore trailing whitespace.
1042 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1044 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1047 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1048 Notification from Dr Andrew Aitchison.
1050 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1051 ExtendedDetectionInfo response format.
1052 Notification from John Horne.
1054 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1057 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1058 XSL and documented dependency on system catalogs, with examples of how
1061 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1064 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1065 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1066 configuration file which is writeable by the Exim user or group.
1068 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1069 of configuration files to cover files specified with the -C option if
1070 they are going to be used with root privileges, not just the default
1073 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1074 option (effectively making it always true).
1076 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1077 files to be used while preserving root privileges.
1079 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1080 that rogue child processes cannot use them.
1082 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1083 run-time user, instead of root.
1085 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1086 Exim run-time user without dropping privileges.
1088 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1089 result string, instead of calling string_vformat() twice with the same
1092 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1093 for other users. Others should always drop root privileges if they use
1094 -C on the command line, even for a whitelisted configure file.
1096 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1098 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1104 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1105 $data_path, and $header_path variables; fixed documentation bugs and
1108 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1109 exipick to access non-standard spools, including the "frozen" queue
1112 NM/01 Bugzilla 965: Support mysql stored procedures.
1113 Patch from Alain Williams
1115 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1117 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1118 Patch from Andreas Metzler
1120 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1121 Patch from Kirill Miazine
1123 NM/05 Bugzilla 671: Added umask to procmail example.
1125 JJ/03 installed exipick 20100323.0, fixing doc bug
1127 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1128 directory. Notification and patch from Dan Rosenberg.
1130 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1132 TK/02 Improve log output when DKIM signing operation fails.
1134 MH/01 Treat the transport option dkim_domain as a colon separated
1135 list, not as a single string, and sign the message with each element,
1136 omitting multiple occurences of the same signer.
1138 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1139 Bugzilla 985, 986. Patch by Simon Arlott
1141 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1142 Patch by Simon Arlott
1144 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1145 MBX locking. Notification from Dan Rosenberg.
1151 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1153 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1155 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1157 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1159 NM/04 Bugzilla 915: Flag broken perl installation during build.
1165 TK/01 Added patch by Johannes Berg that expands the main option
1166 "spamd_address" if it starts with a dollar sign.
1168 TK/02 Write list of recipients to X-Envelope-Sender header when building
1169 the mbox-format spool file for content scanning (suggested by Jakob
1172 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1173 (http://www.dcc-servers.net/) support via dccifd. Activated by
1174 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1176 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1177 by Mark Daniel Reidel <mr@df.eu>.
1179 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1180 When building exim an external PCRE library is now needed -
1181 PCRE is a system library on the majority of modern systems.
1182 See entry on PCRE_LIBS in EDITME file.
1184 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1185 conversation. Added nologin parameter to request.
1186 Patch contributed by Kirill Miazine.
1188 TF/01 Do not log submission mode rewrites if they do not change the address.
1190 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1192 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1193 log files in place. Contributed by Roberto Lima.
1195 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1197 TF/03 Bugzilla 615: When checking the local_parts router precondition
1198 after a local_part_suffix or local_part_prefix option, Exim now
1199 does not use the address's named list lookup cache, since this
1200 contains cached lookups for the whole local part.
1202 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1203 Robert Millan. Documentation is in experimental-spec.txt.
1205 TF/04 Bugzilla 668: Fix parallel build (make -j).
1207 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1209 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1210 Patch provided by Jan Srzednicki.
1212 TF/05 Leading white space used to be stripped from $spam_report which
1213 wrecked the formatting. Now it is preserved.
1215 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1216 that they are available at delivery time.
1218 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1220 TF/08 TLS error reporting now respects the incoming_interface and
1221 incoming_port log selectors.
1223 TF/09 Produce a more useful error message if an SMTP transport's hosts
1224 setting expands to an empty string.
1226 NM/06 Bugzilla 744: EXPN did not work under TLS.
1227 Patch provided by Phil Pennock.
1229 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1230 Patch provided by Richard Godbee.
1232 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1233 acl_smtp_notquit, added index entry.
1235 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1236 Patch provided by Eugene Bujak.
1238 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1239 accept(). Patch provided by Maxim Dounin.
1241 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1242 Patch provided by Phil Pennock.
1244 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1246 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1247 Patch provided by Brad "anomie" Jorsch.
1249 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1250 Patch provided by Dean Brooks.
1252 TK/05 Add native DKIM support (does not depend on external libraries).
1254 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1255 Patch provided by Graeme Fowler.
1257 NM/16 Bugzilla 851: Documentation example syntax fix.
1259 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1261 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1264 NM/19 Bugzilla 745: TLS version reporting.
1265 Patch provided by Phil Pennock.
1267 NM/20 Bugzilla 167: bool: condition support.
1268 Patch provided by Phil Pennock.
1270 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1271 clients. Patch provided by Phil Pennock.
1273 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1274 Patch provided by Brad "anomie" Jorsch.
1276 NM/23 Bugzilla 687: Fix misparses in eximstats.
1277 Patch provided by Heiko Schlittermann.
1279 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1280 Patch provided by Heiko Schlittermann.
1282 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1283 plus update to original patch.
1285 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1287 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1288 Patch provided by David Brownlee.
1290 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1292 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1294 NM/30 Bugzilla 888: TLS documentation bugfixes.
1296 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1298 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1299 Unlike the original bugzilla I have changed all shell scripts in src tree.
1301 NM/33 Bugzilla 898: Transport filter timeout fix.
1302 Patch by Todd Rinaldo.
1304 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1305 Patch by Serge Demonchaux.
1307 NM/35 Bugzilla 39: Base64 decode bug fixes.
1308 Patch by Jakob Hirsch.
1310 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1312 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1314 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1316 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1322 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1323 ALT-N's libdkim that I have put here:
1324 http://duncanthrax.net/exim-experimental/
1326 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1327 were called 'true' and it seems that C99 defines that as a reserved
1328 keyword to be used with 'bool' variable types. That means you could
1329 not include C99-style headers which use bools without triggering
1330 build errors in sieve.c.
1332 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1333 as mailq or other aliases. Changed the --help handling significantly
1334 to do whats expected. exim_usage() emits usage/help information.
1336 SC/01 Added the -bylocaldomain option to eximstats.
1338 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1340 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1342 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1348 PH/01 Another patch from the Sieve maintainer.
1350 PH/02 When an IPv6 address is converted to a string for single-key lookup
1351 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1352 dots are used instead of colons so that keys in lsearch files need not
1353 contain colons. This was done some time before quoting was made available
1354 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1355 (notated using the quote facility) so as to distinguish them from IPv4
1356 keys. This meant that lookups for IP addresses in host lists did not work
1357 for iplsearch lookups.
1359 This has been fixed by arranging for IPv6 addresses to be expressed with
1360 colons if the lookup type is iplsearch. This is not incompatible, because
1361 previously such lookups could never work.
1363 The situation is now rather anomolous, since one *can* have colons in
1364 ordinary lsearch keys. However, making the change in all cases is
1365 incompatible and would probably break a number of configurations.
1367 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1370 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1371 conversion specification without a maximum field width, thereby enabling
1372 a rogue spamd server to cause a buffer overflow. While nobody in their
1373 right mind would setup Exim to query an untrusted spamd server, an
1374 attacker that gains access to a server running spamd could potentially
1375 exploit this vulnerability to run arbitrary code as the Exim user.
1377 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1378 $primary_hostname instead of what libspf2 thinks the hosts name is.
1380 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1381 a directory entry by the name of the lookup key. Previously, if a
1382 symlink pointed to a non-existing file or a file in a directory that
1383 Exim lacked permissions to read, a lookup for a key matching that
1384 symlink would fail. Now it is enough that a matching directory entry
1385 exists, symlink or not. (Bugzilla 503.)
1387 PH/03 The body_linecount and body_zerocount variables are now exported in the
1390 PH/04 Added the $dnslist_matched variable.
1392 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1393 This means they are set thereafter only if the connection becomes
1396 PH/06 Added the client_condition to authenticators so that some can be skipped
1397 by clients under certain conditions.
1399 PH/07 The error message for a badly-placed control=no_multiline_responses left
1400 "_responses" off the end of the name.
1402 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1404 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1405 (without spaces) instead of just copying the configuration text.
1407 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1409 PH/11 Added $max_received_linelength.
1411 PH/12 Added +ignore_defer and +include_defer to host lists.
1413 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1414 way in which PCRE > 7.0 is built.
1416 PH/14 Implemented queue_only_load_latch.
1418 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1419 MAIL command. The effect was to mangle the value on 64-bit systems.
1421 PH/16 Another patch from the Sieve maintainer.
1423 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1425 PH/18 If a system quota error occurred while trying to create the file for
1426 a maildir delivery, the message "Mailbox is full" was not appended to the
1427 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1428 only to a quota excession during the actual writing of the file.
1430 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1431 characters?) which causes problems in log lines. The DN values are now
1432 passed through string_printing() before being added to log lines.
1434 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1435 and InterBase are left for another time.)
1437 PH/21 Added message_body_newlines option.
1439 PH/22 Guard against possible overflow in moan_check_errorcopy().
1441 PH/23 POSIX allows open() to be a macro; guard against that.
1443 PH/24 If the recipient of an error message contained an @ in the local part
1444 (suitably quoted, of course), incorrect values were put in $domain and
1445 $local_part during the evaluation of errors_copy.
1451 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1452 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1455 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1456 issue a MAIL command.
1458 PH/02 In an ACL statement such as
1460 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1462 if a client was not listed at all, or was listed with a value other than
1463 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1464 the condition was not true (as it should be), so access was not denied.
1465 The bug was that the ! inversion was incorrectly passed on to the second
1466 item. This has been fixed.
1468 PH/03 Added additional dnslists conditions == and =& which are different from
1469 = and & when the dns lookup returns more than one IP address.
1471 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1472 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1474 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1475 FSYNC, which compiles an option called disable_fsync that allows for
1476 bypassing fsync(). The documentation is heavily laced with warnings.
1478 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1480 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1481 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1482 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1483 including adding "make clean"; (3) Added -fPIC when compiling the test
1484 dynamically loaded module, to get rid of a warning.
1486 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1487 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1488 = 0s. The bug is otherwise harmless.
1490 PH/07 There was a bug in the dovecot authenticator such that the value of
1491 $auth1 could be overwritten, and so not correctly preserved, after a
1492 successful authentication. This usually meant that the value preserved by
1493 the server_setid option was incorrect.
1495 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1497 PH/09 Installed PCRE release 7.0.
1499 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1500 run for batched SMTP input. It is now run at the start of every message
1501 in the batch. While fixing this I discovered that the process information
1502 (output by running exiwhat) was not always getting set for -bs and -bS
1503 input. This is fixed, and it now also says "batched" for BSMTP.
1505 PH/11 Added control=no_pipelining.
1507 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1508 patch, slightly modified), and move the expansion of helo_data till after
1509 the connection is made in the smtp transport (so it can use these
1512 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1514 PH/14 Added log_selector = +pid.
1516 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1518 PH/16 Add ${if forany and ${if forall.
1520 PH/17 Added dsn_from option to vary the From: line in DSNs.
1522 PH/18 Flush SMTP output before performing a callout, unless control =
1523 no_callout_flush is set.
1525 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1526 was true (the default) a successful delivery failed to delete the retry
1527 item, thus causing premature timeout of the address. The bug is now
1530 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1532 PH/21 Long custom messages for fakedefer and fakereject are now split up
1533 into multiline reponses in the same way that messages for "deny" and
1534 other ACL rejections are.
1536 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1537 with slight modification.
1539 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1540 draft, changing the syntax and factoring some duplicate code".
1542 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1543 for deliveries of the second and subsequent messages over the same SMTP
1546 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1547 ${reduce, with only minor "tidies".
1549 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1551 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1552 expansion side effects.
1554 PH/27 When a message times out after an over-quota error from an Exim-imposed
1555 quota, the bounce message says "mailbox is full". This message was not
1556 being given when it was a system quota that was exceeded. It now should
1559 MH/03 Made $recipients available in local_scan(). local_scan() already has
1560 better access to the recipient list through recipients_list[], but
1561 $recipients can be useful in postmaster-provided expansion strings.
1563 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1564 in the case of a MAIL command with additional options following the
1565 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1566 were accidentally chopped off.
1568 PH/29 SMTP synchronization checks are implemented when a command is read -
1569 there is a check that no more input is waiting when there shouldn't be
1570 any. However, for some commands, a delay in an ACL can mean that it is
1571 some time before the response is written. In this time, more input might
1572 arrive, invalidly. So now there are extra checks after an ACL has run for
1573 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1574 pipelining has not been advertised.
1576 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1578 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1579 returned in the SMTP response when smtp_return_error_details was false.
1580 This has been fixed.
1582 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1583 instead of the C I/O that was originally supplied, because problems were
1584 reported on Solaris.
1586 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1587 Exim which did not show up earlier: it was assuming that a call to
1588 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1589 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1590 was a macro that became an assignment, so it seemed to work. This has
1591 changed to a proper function call with a void return, hence the compile
1592 error. Exim's code has been fixed.
1594 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1597 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1599 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1601 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1602 because it turns out to be too big or there is a timeout) there is no
1603 "Completed" line in the log. When some messages of this type were
1604 selected by exigrep, they were listed as "not completed". Others were
1605 picked up by some special patterns. I have improved the selection
1606 criteria to be more general.
1608 PH/38 The host_find_failed option in the manualroute router can now be set
1609 to "ignore", to completely ignore a host whose IP address cannot be
1610 found. If all hosts are ignored, the behaviour is controlled by the new
1611 host_all_ignored option.
1613 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1614 homing or because of multiple MX records with /mx) generated more than
1615 one IP address, and the following item turned out to be the local host,
1616 all the secondary addresses of the first item were incorrectly removed
1617 from the list, along with the local host and any following hosts (which
1618 is what is supposed to happen).
1620 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1621 whoever called Exim into the -H file. In the case of the daemon it was
1622 behaving confusingly. When first started, it used values for whoever
1623 started the daemon, but after a SIGHUP it used the Exim user (because it
1624 calls itself on a restart). I have changed the code so that it now always
1627 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1628 message are rejected with the same error (e.g. no authentication or bad
1629 sender address), and a DATA command is nevertheless sent (as can happen
1630 with PIPELINING or a stupid MUA), the error message that was given to the
1631 RCPT commands is included in the rejection of the DATA command. This is
1632 intended to be helpful for MUAs that show only the final error to their
1635 PH/42 Another patch from the Sieve maintainer.
1637 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1638 Eximstats - Fixed some broken HTML links and added missing column headers
1640 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1641 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1643 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1645 PH/43 Yet another patch from the Sieve maintainer.
1647 PH/44 I found a way to check for a TCP/IP connection going away before sending
1648 the response to the final '.' that terminates a message, but only in the
1649 case where the client has not sent further data following the '.'
1650 (unfortunately, this is allowed). However, in many cases there won't be
1651 any further data because there won't be any more messages to send. A call
1652 to select() can be used: if it shows that the input is "ready", there is
1653 either input waiting, or the socket has been closed. An attempt to read
1654 the next input character can distinguish the two cases. Previously, Exim
1655 would have sent an OK response which the client would never have see.
1656 This could lead to message repetition. This fix should cure that, at
1657 least in a lot of common cases.
1659 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1660 advertised in response to EHLO.
1666 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1667 fixed by 4.65/MH/01 (is this a record?) are fixed:
1669 (i) An empty string was always treated as zero by the numeric comparison
1670 operators. This behaviour has been restored.
1672 (ii) It is documented that the numeric comparison operators always treat
1673 their arguments as decimal numbers. This was broken in that numbers
1674 starting with 0 were being interpreted as octal.
1676 While fixing these problems I realized that there was another issue that
1677 hadn't been noticed. Values of message_size_limit (both the global option
1678 and the transport option) were treated as octal if they started with 0.
1679 The documentation was vague. These values are now always treated as
1680 decimal, and I will make that clear in the documentation.
1686 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1687 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1690 MH/01 Don't check that the operands of numeric comparison operators are
1691 integers when their expansion is in "skipping" mode (fixes bug
1692 introduced by 4.64-PH/07).
1694 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1695 child addresses, Exim now panics and dies. Previously, because the count
1696 is held in a short int, deliveries were likely to be lost. As such a
1697 large number of recipients for a single message is ridiculous
1698 (performance will be very, very poor), I have chosen to impose a limit
1699 rather than extend the field.
1705 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1706 leftover -K file (the existence of which was triggered by #402).
1707 While we were at it, introduced process PID as part of the -K
1708 filename. This should rule out race conditions when creating
1711 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1712 processing considerably. Previous code took too long for large mails,
1713 triggering a timeout which in turn triggers #401.
1715 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1716 in the DK code in transports.c. sendfile() is not really portable,
1717 hence the _LINUX specificness.
1719 TF/01 In the add_headers option to the mail command in an Exim filter,
1720 there was a bug that Exim would claim a syntax error in any
1721 header after the first one which had an odd number of characters
1724 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1725 callout verification, Exim cached a "reject" for the entire domain. This
1726 is correct for most verifications, but it is not correct for a recipient
1727 verification with use_sender or use_postmaster set, because in that case
1728 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1729 case of MAIL FROM:<> rejection from other early rejections (e.g.
1730 rejection of HELO). When verifying a recipient using a non-null MAIL
1731 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1732 Whatever the result of the callout, the value of the domain cache is
1733 left unchanged (for any other kind of callout, getting as far as trying
1734 RCPT means that the domain itself is ok).
1736 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1739 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1740 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1741 session, the daemon ignores ECONNECT errors and logs others; it now
1742 ignores EPIPE as well.
1744 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1745 (quoted-printable decoding).
1747 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1748 later the small subsequent patch to fix an introduced bug.
1750 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1752 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1754 PH/08 An error is now given if message_size_limit is specified negative.
1756 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1757 to be given (somewhat) arbitrary names.
1759 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1762 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1763 miscellaneous code fixes
1765 PH/10 Added the log_reject_target ACL modifier to specify where to log
1768 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1769 hostname. This is wrong, because it relates to the incoming message (and
1770 probably the interface on which it is arriving) and not to the outgoing
1771 callout (which could be using a different interface). This has been
1772 changed to use the value of the helo_data option from the smtp transport
1773 instead - this is what is used when a message is actually being sent. If
1774 there is no remote transport (possible with a router that sets up host
1775 addresses), $smtp_active_hostname is used.
1777 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1778 tweaks were necessary in order to get it to work (see also 21 below):
1779 (a) The code assumed that strncpy() returns a negative number on buffer
1780 overflow, which isn't the case. Replaced with Exim's string_format()
1782 (b) There were several signed/unsigned issues. I just did the minimum
1783 hacking in of casts. There is scope for a larger refactoring.
1784 (c) The code used strcasecmp() which is not a standard C function.
1785 Replaced with Exim's strcmpic() function.
1786 (d) The code set only $1; it now sets $auth1 as well.
1787 (e) A simple test gave the error "authentication client didn't specify
1788 service in request". It would seem that Dovecot has changed its
1789 interface. Fortunately there's a specification; I followed it and
1790 changed what the client sends and it appears to be working now.
1792 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1795 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1796 address A is aliased to B and C, where B exists and C does not. Without
1797 -v the output is "A verified" because verification stops after a
1798 successful redirection if more than one address is generated. However,
1799 with -v the child addresses are also verified. Exim was outputting "A
1800 failed to verify" and then showing the successful verification for C,
1801 with its parentage. It now outputs "B failed to verify", showing B's
1802 parentage before showing the successful verification of C.
1804 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1805 look up a TXT record in a specific list after matching in a combined
1808 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1809 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1810 they consult the DNS. I had assumed they would set it the way they
1811 wanted; and indeed my experiments on Linux seem to show that in some
1812 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1813 To be on the safe side, however, I have now made the interface to
1814 host_find_byname() similar to host_find_bydns(), with an argument
1815 containing the DNS resolver options. The host_find_byname() function now
1816 sets these options at its start, just as host_find_bydns() does. The smtp
1817 transport options dns_qualify_single and dns_search_parents are passed to
1818 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1819 of host_find_byname() use the default settings of RES_DEFNAMES
1820 (qualify_single) but not RES_DNSRCH (search_parents).
1822 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1823 spool_read_header() do less string testing, by means of a preliminary
1824 switch on the second character of optional "-foo" lines. (This is
1825 overdue, caused by the large number of possibilities that now exist.
1826 Originally there were few.) While I was there, I also converted the
1827 str(n)cmp tests so they don't re-test the leading "-" and the first
1828 character, in the hope this might squeeze out yet more improvement.
1830 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1831 flag allowing group syntax was set by the header_syntax check but not
1832 turned off, possible causing trouble later; (2) The flag was not being
1833 set at all for the header_verify test, causing "group"-style headers to
1834 be rejected. I have now set it in this case, and also caused header_
1835 verify to ignore an empty address taken from a group. While doing this, I
1836 came across some other cases where the code for allowing group syntax
1837 while scanning a header line wasn't quite right (mostly, not resetting
1838 the flag correctly in the right place). These bugs could have caused
1839 trouble for malformed header lines. I hope it is now all correct.
1841 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1842 with the "reply" argument non-NULL. The code, however (which originally
1843 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1844 but it didn't always do it. This confused somebody who was copying the
1845 code for some other use. I have removed all the tests.
1847 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1848 feature that was used to support insecure browsers during the U.S. crypto
1849 embargo. It requires special client support, and Exim is probably the
1850 only MTA that supported it -- and would never use it because real RSA is
1851 always available. This code has been removed, because it had the bad
1852 effect of slowing Exim down by computing (never used) parameters for the
1853 RSA_EXPORT functionality.
1855 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1856 authenticator to fail if there's a tab character in the incoming data
1857 (there should never be unless someone is messing about, as it's supposed
1858 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1859 if the connection is using TLS or if the remote IP is the same as the
1860 local IP, and the "valid-client-cert option" if a client certificate has
1863 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1864 authenticators. This can be used for authorization after authentication
1865 succeeds. (In the case of plaintext, it servers for both authentication
1868 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1869 if any retry times were supplied.
1871 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1872 connection, where there is no client IP address to check. In this
1873 situation, the verify now always succeeds.
1875 PH/25 Applied John Jetmore's -Mset patch.
1877 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1879 PH/27 In a string expansion for a processed (not raw) header when multiple
1880 headers of the same name were present, leading whitespace was being
1881 removed from all of them, but trailing whitespace was being removed only
1882 from the last one. Now trailing whitespace is removed from each header
1883 before concatenation. Completely empty headers in a concatenation (as
1884 before) are ignored.
1886 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1887 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1889 PH/29 [Removed. This was a change that I later backed out, and forgot to
1890 correct the ChangeLog entry (that I had efficiently created) before
1891 committing the later change.]
1893 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1894 address errors (4xx response to RCPT) over the same connection as other
1895 messages routed to the same hosts. Such deliveries are always "forced",
1896 so retry times are not inspected. This resulted in far too many retries
1897 for the affected addresses. The effect occurred only when there were more
1898 hosts than the hosts_max_try setting in the smtp transport when it had
1899 the 4xx errors. Those hosts that it had tried were not added to the list
1900 of hosts for which the message was waiting, so if all were tried, there
1901 was no problem. Two fixes have been applied:
1903 (i) If there are any address or message errors in an SMTP delivery, none
1904 of the hosts (tried or untried) are now added to the list of hosts
1905 for which the message is waiting, so the message should not be a
1906 candidate for sending over the same connection that was used for a
1907 successful delivery of some other message. This seems entirely
1908 reasonable: after all the message is NOT "waiting for some host".
1909 This is so "obvious" that I'm not sure why it wasn't done
1910 previously. Hope I haven't missed anything, but it can't do any
1911 harm, as the worst effect is to miss an optimization.
1913 (ii) If, despite (i), such a delivery is accidentally attempted, the
1914 routing retry time is respected, so at least it doesn't keep
1915 hammering the server.
1917 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1918 in ${readsocket because some servers need this prod.
1920 PH/32 Added some extra debug output when updating a wait-xxx database.
1922 PH/33 The hint "could be header name not terminated by colon", which has been
1923 given for certain expansion errors for a long time, was not being given
1924 for the ${if def:h_colon_omitted{... case.
1926 PH/34 The spec says: "With one important exception, whenever a domain list is
1927 being scanned, $domain contains the subject domain." There was at least
1928 one case where this was not true.
1930 PH/35 The error "getsockname() failed: connection reset by peer" was being
1931 written to the panic log as well as the main log, but it isn't really
1932 panic-worthy as it just means the connection died rather early on. I have
1933 removed the panic log writing for the ECONNRESET error when getsockname()
1936 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1937 runs only) independently of the message's sender address. This meant
1938 that, if the 4xx error was in fact related to the sender, a different
1939 message to the same recipient with a different sender could confuse
1940 things. In particualar, this can happen when sending to a greylisting
1941 server, but other circumstances could also provoke similar problems.
1942 I have changed the default so that the retry time for these errors is now
1943 based a combination of the sender and recipient addresses. This change
1944 can be overridden by setting address_retry_include_sender=false in the
1947 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1948 remote server are returned as part of bounce messages. This was not
1949 happening for LMTP over a pipe (the lmtp transport), but now it is the
1950 same for both kinds of LMTP.
1952 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1953 in header lines that were in fact CNAMEs. This is no longer the case.
1955 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1956 and queue runs started by the daemon processed all messages. This has
1957 been fixed so that -R and -S can now usefully be given with -q<time>.
1959 PH/40 Import PCRE release 6.7 (fixes some bugs).
1961 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1963 PH/42 Give an error if -q is specified more than once.
1965 PH/43 Renamed the variables $interface_address and $interface_port as
1966 $received_ip_address and $received_port, to make it clear that these
1967 values apply to message reception, and not to the outgoing interface when
1968 a message is delivered. (The old names remain recognized, of course.)
1970 PH/44 There was no timeout on the connect() call when using a Unix domain
1971 socket in the ${readsocket expansion. There now is.
1973 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1974 be meaningful with "accept".
1976 SC/01 Eximstats V1.43
1977 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1979 SC/02 Eximstats V1.44
1980 Use a glob alias rather than an array ref in the generated
1981 parser. This improves both readability and performance.
1983 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1984 Collect SpamAssassin and rejection statistics.
1985 Don't display local sender or destination tables unless
1986 there is data to show.
1987 Added average volumes into the top table text output.
1989 SC/04 Eximstats V1.46
1990 Collect data on the number of addresses (recipients)
1991 as well as the number of messages.
1993 SC/05 Eximstats V1.47
1994 Added 'Message too big' to the list of mail rejection
1995 reasons (thanks to Marco Gaiarin).
1997 SC/06 Eximstats V1.48
1998 Mainlog lines which have GMT offsets and are too short to
1999 have a flag are now skipped.
2001 SC/07 Eximstats V1.49 (Alain Williams)
2002 Added the -emptyok flag.
2004 SC/08 Eximstats V1.50
2005 Fixes for obtaining the IP address from reject messages.
2007 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2008 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2009 whitesspace changes from 4.64-PH/27
2011 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2014 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2015 are found, allow negative numbers in numeric criteria)
2017 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2019 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2020 to match changes made in 4.64-PH/43
2022 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2024 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2025 "domain" is actually an IP address, and force a failure. This locks out
2026 those revolvers/nameservers that support "A-for-A" lookups, in
2027 contravention of the specifications.
2029 PH/48 When a host name was looked up from an IP address, and the subsequent
2030 forward lookup of the name timed out, the host name was left in
2031 $sender_host_name, contrary to the specification.
2033 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2034 restricted to single-key lookups, Exim was not diagnosing an error if
2035 * or *@ was used with a query-style lookup.
2037 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2039 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2040 long ago, but noone interested enough thought of it. Let's just say that
2041 the "1.1" means that there are some new functions that weren't there at
2042 some point in the past.
2044 PH/51 Error processing for expansion failure of helo_data from an smtp
2045 transport during callout processing was broken.
2047 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2048 tested/used via the -bh/-bhc/-bs options.
2050 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2051 bug, fixed in subsequent PCRE releases).
2053 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2054 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2056 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2062 SC/01 Use a glob alias rather than an array ref in eximstats generated
2063 parser. This improves both readability and performance.
2065 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2066 Don't display local sender or destination tables in eximstats unless
2067 there is data to show.
2068 Added average volumes into the eximstats top table text output.
2070 SC/03 Collect data on the number of addresses (recipients) as well
2071 as the number of messages in eximstats.
2073 TF/01 Correct an error in the documentation for the redirect router. Exim
2074 does (usually) call initgroups() when daemonizing.
2076 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2077 with consistent privilege compared to when running as a daemon.
2079 TF/03 Note in the spec that $authenticated_id is not set for local
2080 submissions from trusted users.
2082 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2083 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2085 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2086 by adding some example configuration directives to the default
2087 configuration file. A little bit of work is required to uncomment the
2088 directives and define how usernames and passwords are checked, but
2089 there is now a framework to start from.
2091 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2092 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2093 without this. I don't know how relevant this is to other LDAP libraries.
2095 PH/02 Add the verb name to the "unknown ACL verb" error.
2097 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2099 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2101 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2102 directory not expanded when it should be if an expanded home directory
2103 was set for the address (which is overridden by the transport).
2105 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2108 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2109 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2110 because it is too late at that time, and has no effect.
2112 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2113 security issue with \' (bugzilla #107). I could not use the
2114 PQescapeStringConn() function, because it needs a PGconn value as one of
2117 PH/08 When testing addresses using -bt, indicate those final addresses that
2118 are duplicates that would not cause an additional delivery. At least one
2119 person was confused, thinking that -bt output corresponded to deliveries.
2120 (Suppressing duplicates isn't a good idea as you lose the information
2121 about possibly different redirections that led to the duplicates.)
2123 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2124 systems where poll() doesn't work, in particular OS X.
2126 PH/10 Added more information to debugging output for retry time not reached.
2128 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2129 operations in malware.c.
2131 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2134 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2135 syslog_duplication set false, log lines that would normally be written
2136 both the the main log and to the reject log were not written to syslog at
2139 PH/14 In the default configuration, change the use of "message" in ACL warn
2140 statements to "add_header".
2142 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2143 not followed by a command (e.g. "seen endif").
2145 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2146 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2149 PH/17 Added extra conditions to the default value of delay_warning_condition
2153 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2154 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2155 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2158 The Auto-Submitted: and various List- headers are standardised, whereas I
2159 don't think Precedence: ever was.
2161 PH/18 Refactored debugging code in route_finduser() to show more information,
2162 in particular, the error code if getpwnam() issues one.
2164 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2165 This is apparently needed in addition to the PH/07 change above to avoid
2166 any possible encoding problems.
2168 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2169 but not after initializing Perl.
2171 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2172 output them only if debugging. By default they are written stderr,
2173 apparently, which is not desirable.
2175 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2178 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2181 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2183 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2184 authenticated or an ident call has been made. Suppress the default
2185 values for $authenticated_id and $authenticated_sender (but permit -oMai
2186 and -oMas) when testing with -bh.
2188 PH/24 Re-jigged the order of the tests in the default configuration so that the
2189 tests for valid domains and recipients precede the DNS black list and CSA
2190 tests, on the grounds that those ones are more expensive.
2192 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2193 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2194 This bug exists in every version of Exim that I still have, right back to
2197 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2198 However, an attempt to turn on case-sensitivity in a regex key by
2199 including (?-i) didn't work because the subject string was already
2200 lowercased, and the effects were non-intuitive. It turns out that a
2201 one-line patch can be used to allow (?-i) to work as expected.
2207 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2208 other effects) broke the use of negated acl sub-conditions.
2210 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2213 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2214 "Deny" causes Exim to reject the incoming connection with a 554 error.
2215 Unfortunately, if there is a major crisis, such as a disk failure,
2216 tcp-wrappers gives "deny", whereas what one would like would be some
2217 kind of temporary error. A kludge has been added to help with this.
2218 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2219 554 error is used if errno is still zero or contains ENOENT (which occurs
2220 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2223 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2225 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2226 errors. Otherwise a message that provokes a temporary error (when other
2227 messages do not) can cause a whole host to time out.
2229 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2230 addresses were routed directly to files or pipes from a redirect router.
2231 File deliveries just didn't batch; pipe deliveries might have suffered
2234 PH/06 A failure to get a lock for a hints database would erroneously always say
2235 "Failed to get write lock", even when it was really a read lock.
2237 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2238 of 0600. This has been changed to use the value of the lockfile_mode
2239 option (which defaults to 0600).
2241 PH/08 Applied small patch from the Sieve maintainer.
2243 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2244 folder from quota calculations, a direct delivery into this folder messed
2245 up the contents of the maildirsize file. This was because the regex was
2246 used only to exclude .Trash (or whatever) when the size of the mailbox
2247 was calculated. There was no check that a delivery was happening into an
2248 excluded directory. This bug has been fixed by ignoring all quota
2249 processing for deliveries into excluded directories.
2251 PH/10 Added the maildirfolder_create_regex option to appendfile.
2257 PH/01 The code for finding all the local interface addresses on a FreeBSD
2258 system running IPv6 was broken. This may well have applied to all BSD
2259 systems, as well as to others that have similar system calls. The broken
2260 code found IPv4 interfaces correctly, but gave incorrect values for the
2261 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2262 that it would not match correctly against @[] and not recognize the IPv6
2265 PH/02 The ipliteral router was not recognizing addresses of the form user@
2266 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2268 PH/03 Added disable_ipv6.
2270 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2271 decoded form, because it is most often used to construct To: headers
2272 lines in autoreplies, and the decoded form may well be syntactically
2273 invalid. However, $reply_address has leading white space removed, and all
2274 newlines turned into spaces so that the autoreply transport does not
2277 PH/05 If group was specified without a user on a router, and no group or user
2278 was specified on a transport, the group from the router was ignored.
2280 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2281 for visible compile-time settings that can be used to change these
2282 numbers, for those that want even more. Backwards compatibility with old
2283 spool files has been maintained. However, going back to a previous Exim
2284 release will lost any variables that are in spool files.
2286 PH/07 Two small changes when running in the test harness: increase delay when
2287 passing a TCP/IP connection to a new process, in case the original
2288 process has to generate a bounce, and remove special handling of
2289 127.0.0.2 (sic), which is no longer necessary.
2291 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2292 be the same on different OS.
2294 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2297 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2298 whether --show-vars was specified or not
2300 JJ/02 exipick: Added support for new ACL variable spool format introduced
2303 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2304 syntactically invalid From: or Reply-to: line, and a filter used this to
2305 generate an autoreply, and therefore failed to obtain an address for the
2306 autoreply, Exim could try to deliver to a non-existent relative file
2307 name, causing unrelated and misleading errors. What now happens is that
2308 it logs this as a hard delivery error, but does not attempt to create a
2311 PH/11 The exinext utility has a -C option for testing purposes, but although
2312 the given file was scanned by exinext itself; it wasn't being passed on
2313 when Exim was called.
2315 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2316 an end-of-file indication when reading a command response.
2318 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2319 compiled. In many other places in Exim, IPv6 addresses are always
2320 recognized, so I have changed this. It also means that IPv4 domain
2321 literals of the form [IPV4:n.n.n.n] are now always recognized.
2323 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2324 used if the router is not running as root, for example, when verifying at
2325 ACL time, or when using -bh. The debugging output from this situation was
2326 non-existent - all you got was a failure to exec. I have made two
2329 (a) Failures to set uid/gid, the current directory, or a process leader
2330 in a subprocess such as that created by queryprogram now generate
2331 suitable debugging ouput when -d is set.
2333 (b) The queryprogram router detects when it is not running as root,
2334 outputs suitable debugging information if -d is set, and then runs
2335 the subprocess without attempting to change uid/gid.
2337 PH/15 Minor change to Makefile for building test_host (undocumented testing
2340 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2341 additional section of a DNS packet that returns MX or SRV records.
2342 Instead, it always explicitly searches for A/AAAA records. This avoids
2343 major problems that occur when a DNS server includes only records of one
2344 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2345 fixed another bug: if SRV records were looked up and the corresponding
2346 address records were *not* found in the additional section, the port
2347 values from the SRV records were lost.
2349 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2350 using the correct key (the original address) when searching the retry
2351 rules in order to find which one to use for generating the retry hint.
2353 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2354 adding the default one. Similarly, if it contains a Reply-To: header, the
2355 errors_reply_to option, if set, is not used.
2357 PH/19 When calculating a retry time, Exim used to measure the "time since
2358 failure" by looking at the "first failed" field in the retry record. Now
2359 it does not use this if it is later than than the arrival time of the
2360 message. Instead it uses the arrival time. This makes for better
2361 behaviour in cases where some deliveries succeed, thus re-setting the
2362 "first failed" field. An example is a quota failure for a huge message
2363 when small messages continue to be delivered. Without this change, the
2364 "time since failure" will always be short, possible causing more frequent
2365 delivery attempts for the huge message than are intended.
2366 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2368 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2369 $1, $2, $3) because the numerical variables can be reset during some
2370 expansion items (e.g. "match"), thereby losing the authentication data.
2372 PH/21 Make -bV show the size of off_t variables so that the test suite can
2373 decide whether to run tests for quotas > 2G.
2375 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2376 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2377 filecount value is greater than 2G or if a quota value is greater than 2G
2378 on a system where the size of off_t is not greater than 4, a panic error
2381 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2382 never match. The debug and -bh output now contains an explicit error
2383 message indicating a malformed IPv4 address or mask.
2385 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2386 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2387 PH/23 above applies.
2389 PH/25 Do not write to syslog when running in the test harness. The only
2390 occasion when this arises is a failure to open the main or panic logs
2391 (for which there is an explicit test).
2393 PH/26 Added the /no_tell option to "control=freeze".
2395 PH/27 If a host name lookup failed very early in a connection, for example, if
2396 the IP address matched host_lookup and the reverse lookup yielded a name
2397 that did not have a forward lookup, an error message of the form "no IP
2398 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2399 could be logged. Now it outputs the IP address instead of "NULL".
2401 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2402 allows the sender and the authenticated sender to be set when
2403 submitting a message from within Exim. Since child_open_exim() is
2404 documented for local_scan(), the new function should be too.
2406 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2407 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2408 results in an empty string is now treated as unset.
2410 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2412 PH/31 Added sender_verify_fail logging option.
2414 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2415 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2416 tidied the source and removed it altogether.
2418 PH/33 When a queue run was abandoned because the load average was too high, a
2419 log line was always written; now it is written only if the queue_run log
2420 selector is set. In addition, the log line for abandonment now contains
2421 information about the queue run such as the pid. This is always present
2422 in "start" and "stop" lines but was omitted from the "abandon" line.
2424 PH/34 Omit spaces between a header name and the colon in the error message that
2425 is given when verify = headers_syntax fails (if there are lots of them,
2426 the message gets confusing).
2428 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2429 names, as there are now some PTR records that contain slashes. This check
2430 is only to protect against broken name servers that fall over on strange
2431 characters, so the fact that it applies to all lookups doesn't matter.
2433 PH/36 Now that the new test suite is complete, we can remove some of the
2434 special code in Exim that was needed for the old test suite. For example,
2435 sorting DNS records because real resolvers return them in an arbitrary
2436 order. The new test suite's fake resolver always returns records in the
2439 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2440 bounces) except when queue_only is set, to avoid logging races between
2441 the different processes.
2443 PH/38 Panic-die if .include specifies a non-absolute path.
2445 PH/39 A tweak to the "H" retry rule from its user.
2447 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2448 a label. They prevented compilation on older perls.
2450 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2451 a warning to be raised on newish perls.
2453 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2454 on queue. Changes to match documented behaviour of showing count of
2455 messages matching specified criteria.
2457 PH/40 Changed the default ident timeout from 30s to 5s.
2459 PH/41 Added support for the use of login_cap features, on those BSD systems
2460 that have them, for controlling the resources used by pipe deliveries.
2462 PH/42 The content-scanning code uses fopen() to create files in which to put
2463 message data. Previously it was not paying any attention to the mode of
2464 the files. Exim runs with umask(0) because the rest of the code creates
2465 files with open(), and sets the required mode explicitly. Thus, these
2466 files were ending up world-writeable. This was not a big issue, because,
2467 being within the spool directory, they were not world-accessible. I have
2468 created a function called modefopen, which takes an additional mode
2469 argument. It sets umask(777), creates the file, chmods it to the required
2470 mode, then resets the umask. All the relevant calls to fopen() in the
2471 content scanning code have been changed to use this function.
2473 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2474 to 24 hours. This avoids potential overflow problems when processing G
2475 and H retry rules. I suspect nobody ever tinkers with this value.
2477 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2479 PH/45 When the plaintext authenticator is running as a client, the server's
2480 challenges are checked to ensure they are valid base64 strings. By
2481 default, the authentication attempt is cancelled if an invalid string is
2482 received. Setting client_ignore_invalid_base64 true ignores these errors.
2483 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2484 they are received. Thus, the responses can be made to depend on the
2485 challenges. If an invalid string is ignored, an empty string is placed in
2488 PH/46 Messages that are created by the autoreply transport now contains a
2489 References: header, in accordance with RFCs 2822 and 3834.
2491 PH/47 Added authenticated_sender_force to the smtp transport.
2493 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2495 PH/49 Installed latest patch from the Sieve maintainer.
2497 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2498 was also set, the appendfile transport was unnecessarily scanning a
2499 directory of message files (e.g. for maildir delivery) to find the count
2500 of files (along with the size), even though it did not need this
2501 information. It now does the scan only if it needs to find either the
2502 size of the count of files.
2504 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2506 PH/52 Two bugs concerned with error handling when the smtp transport is
2509 (i) Exim was not creating retry information for temporary errors given
2510 for individual recipients after the DATA command when the smtp transport
2511 was used in LMTP mode. This meant that they could be retried too
2512 frequently, and not timed out correctly.
2514 (ii) Exim was setting the flag that allows error details to be returned
2515 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2516 recipients that were returned after the DATA command.
2518 PH/53 This is related to PH/52, but is more general: for any failing address,
2519 when detailed error information was permitted to be returned to the
2520 sender, but the error was temporary, then after the final timeout, only
2521 "retry timeout exceeded" was returned. Now it returns the full error as
2522 well as "retry timeout exceeded".
2524 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2525 do this, and (what is worse) MTAs that accept it.
2527 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2528 will now be deprecated.
2530 PH/56 New os.c-cygwin from the Cygwin maintainer.
2532 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2533 formats (previously only available in exim formats via -bpr, -bpru,
2534 and -bpra. Now also available in native and exiqgrep formats)
2536 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2537 with very large, slow to parse queues
2539 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2541 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2543 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2544 responses to authentication challenges, though it was showing the
2545 challenges; (ii) I've removed the CR characters from the debug output for
2548 PH/58 Allow for the insertion of a newline as well as a space when a string
2549 is turned into more than one encoded-word during RFC 2047 encoding. The
2550 Sieve code now uses this.
2552 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2553 data_4xx, lost_connection, tls_required.
2555 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2556 message was being sent as an SMTP response.
2558 PH/61 Add -l and -k options to exicyclog.
2560 PH/62 When verifying, if an address was redirected to one new address, so that
2561 verification continued, and the new address failed or deferred after
2562 having set something in $address_data, the value of $address_data was not
2563 passed back to the ACL. This was different to the case when no
2564 redirection occurred. The value is now passed back in both cases.
2566 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2567 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2568 use login_cap.h, so on its own it isn't the distinguishing feature. The
2569 new name refers directly to the setclassresources() function.
2571 PH/65 Added configuration files for NetBSD3.
2573 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2575 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2576 is preferred over IPv4.
2578 PH/68 The bounce_return_message and bounce_return_body options were not being
2579 honoured for bounces generated during the reception of non-SMTP messages.
2580 In particular, this applied to messages rejected by the ACL. This bug has
2581 been fixed. However, if bounce_return_message is true and bounce_return_
2582 body is false, the headers that are returned for a non-SMTP message
2583 include only those that have been read before the error was detected.
2584 (In the case of an ACL rejection, they have all been read.)
2586 PH/69 The HTML version of the specification is now built in a directory called
2587 spec_html instead of spec.html, because the latter looks like a path with
2588 a MIME-type, and this confuses some software.
2590 PH/70 Catch two compiler warnings in sieve.c.
2592 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2593 function verify_get_ident() calls ip_connect() to connect a socket, but
2594 if the "connect()" function timed out, ip_connect() used to close the
2595 socket. However, verify_get_ident() also closes the socket later, and in
2596 between Exim writes to the log, which may get opened at this point. When
2597 the socket was closed in ip_connect(), the log could get the same file
2598 descriptor number as the socket. This naturally causes chaos. The fix is
2599 not to close the socket in ip_connect(); the socket should be closed by
2600 the function that creates it. There was only one place in the code where
2601 this was missing, in the iplookup router, which I don't think anybody now
2602 uses, but I've fixed it anyway.
2604 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2605 well as to direct DNS lookups. Otherwise the handling of names in host
2606 lists is inconsistent and therefore confusing.
2612 PH/01 Two changes to the default runtime configuration:
2614 (1) Move the checks for relay_from_hosts and authenticated clients from
2615 after to before the (commented out) DNS black list checks.
2617 (2) Add control=submission to the relay_from_hosts and authenticated
2618 clients checks, on the grounds that messages accepted by these
2619 statements are most likely to be submissions.
2621 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2623 (1) Generate an error if the third argument for the ${prvs expansion is
2626 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2629 (3) Reset the variables that are obtained from the first argument of
2630 ${prvscheck and used in the second argument before leaving the code,
2631 because their memory is reclaimed, so using them afterwards may do
2634 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2635 one (it's much easier than Tom thought :-).
2637 (5) Because of (4), we can now allow for the use of $prvscheck_result
2638 inside the third argument.
2640 PH/03 For some reason, the default setting of PATH when running a command from
2641 a pipe transport was just "/usr/bin". I have changed it to
2644 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2645 anything to be listed in the output from -bV.
2647 PH/05 When a filter generated an autoreply, the entire To: header line was
2648 quoted in the delivery log line, like this:
2650 => >A.N.Other <ano@some.domain> <original@ddress> ...
2652 This has been changed so that it extracts the operative address. There
2653 may be more than one such address. If so, they are comma-separated, like
2656 => >ano@some.domain,ona@other.domain <original@ddress> ...
2658 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2659 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2660 not being looked up in the rDNS to get a host name, Exim was showing the
2661 IP address twice in Received: lines, even though the IP addresses were
2662 identical. For example:
2664 Received: from [1.2.3.4] (helo=[1.2.3.4])
2666 However, if the real host name was known, it was omitting the HELO data
2667 if it matched the actual IP address. This has been tidied up so that it
2668 doesn't show the same IP address twice.
2670 PH/07 When both +timestamp and +memory debugging was on, the value given by
2671 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2672 called by the debug printing, thereby overwriting the timestamp buffer.
2673 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2675 PH/08 When the original message was included in an autoreply transport, it
2676 always said "this is a copy of the message, including all the headers",
2677 even if body_only or headers_only was set. It now gives an appropriate
2680 PH/09 Applied a patch from the Sieve maintainer which:
2682 o fixes some comments
2683 o adds the (disabled) notify extension core
2684 o adds some debug output for the result of if/elsif tests
2685 o points to the current vacation draft in the documentation
2686 and documents the missing references header update
2690 o fixes a bug in processing the envelope test (when testing
2691 multiple envelope elements, the last element determinted the
2694 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2695 Electronic Mail") by including:
2697 Auto-submitted: auto-generated
2699 in the messages that it generates (bounce messages and others, such as
2700 warnings). In the case of bounce messages for non-SMTP mesages, there was
2701 also a typo: it was using "Auto_submitted" (underscore instead of
2702 hyphen). Since every message generated by Exim is necessarily in response
2703 to another message, thes have all been changed to:
2705 Auto-Submitted: auto-replied
2707 in accordance with these statements in the RFC:
2709 The auto-replied keyword:
2711 - SHOULD be used on messages sent in direct response to another
2712 message by an automatic process,
2714 - MUST NOT be used on manually-generated messages,
2716 - MAY be used on Delivery Status Notifications (DSNs) and Message
2717 Disposition Notifications (MDNs),
2719 - MUST NOT be used on messages generated by automatic or periodic
2720 processes, except for messages which are automatic responses to
2723 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2724 to the default Received: header definition.
2726 PH/12 Added log selector acl_warn_skipped (default on).
2728 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2729 variables because (a) they are in the wrong storage pool and (b) even if
2730 they were copied, it wouldn't work properly because of the caching.
2732 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2733 checking when decoding. Apparently there are clients that generate
2734 overlong encoded strings. Why am I not surprised?
2736 PH/15 If the first argument of "${if match_address" was not empty, but did not
2737 contain an "@" character, Exim crashed. Now it writes a panic log message
2738 and treats the condition as false.
2740 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2742 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2743 extension "envelope-auth". The code is finished and in agreement with
2744 other implementations, but there is no documentation so far and in fact,
2745 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2746 not changing the active code.
2748 Print executed "if" and "elsif" statements when debugging is used. This
2749 helps a great deal to understand what a filter does.
2751 Document more things not specified clearly in RFC3028. I had all this
2752 sorted out, when out of a sudden new issues came to my mind. Oops."
2754 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2757 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2758 canonical form (as documented). However, after a host name lookup from
2759 the IP address, check_host() was doing a simple string comparison with
2760 addresses acquired from the DNS when checking that the found name did
2761 have the original IP as one of its addresses. Since any found IPv6
2762 addresses are likely to be in abbreviated form, the comparison could
2763 fail. Luckily, there already exists a function for doing the comparison
2764 by converting both addresses to binary, so now that is used instead of
2765 the text comparison.
2767 PH/20 There was another similar case to PH/19, when a complete host name was
2768 given in a host list; looking up its IP address could give an abbreviated
2769 form, whereas the current host's name might or might not be abbreviated.
2770 The same fix has been applied.
2776 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2777 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2780 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2781 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2783 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2785 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2786 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2787 Also, for "Auto-Submitted", treat anything other than "no" as
2788 non-personal, in accordance with RFC 3834. (Previously it treated
2789 anything starting "auto-" as non-personal.)
2791 TF/01 The control=submission/name=... option had a problem with syntax
2792 errors if the name included a slash character. The /name= option
2793 now slurps the rest of the string, so it can include any characters
2794 but it must come last in the list of options (after /sender_retain
2797 PH/05 Some modifications to the interface to the fake nameserver for the new
2805 TK/01 Added the "success_on_redirect" address verification option. See
2806 NewStuff for rationale and an example.
2808 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2810 PH/02 Patch to exigrep to allow it to work on syslog lines.
2812 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2813 fread() to skip over the body file's header line, because in Cygwin the
2814 header line is locked and is inaccessible.
2816 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2817 co-exist for some time) to make it clear that it is the Exim ID that is
2818 referenced, not the Message-ID: header line.
2820 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2821 string_format() function, because snprintf() does not exist on all
2824 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2825 ${stat: expansion item.
2827 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2828 protocol synchronization error", to keep the pedants happy.
2830 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2831 well as for IRIX systems, when gcc is being used. See the host.c source
2834 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2836 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2839 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2840 transport and to the smtp transport in LMTP mode.
2842 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2844 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2845 run a filter in a subprocess. This could lead to confusion in subsequent
2846 lookups in the parent process. There should also be a search_tidyup() at
2847 the end of the subprocess.
2849 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2850 only if the host matched helo_try_verify_hosts, which caused the
2851 verification to occur when the EHLO/HELO command was issued. The ACL just
2852 tested the remembered result. Now, if a previous verification attempt has
2853 not happened, "verify = helo" does it there and then.
2855 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2857 TK/03 Fix log output including CR from clamd.
2859 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2860 exist provoked a memory error which could cause a segfault.
2862 PH/15 Installed PCRE 6.2
2864 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2866 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2867 of the problem. Specifically, suggested +O2 rather than +O1 for the
2870 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2872 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2873 record, the port was not correctly logged when the outgoing_port log
2874 selector was set (it logged the transort's default port).
2876 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2877 fallback_hosts, and "hosts" in the smtp transport.
2879 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2880 host errors such as "Connection refused".
2882 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2883 authentication with radiusclient 0.4.9:
2885 - Error returned from rc_read_config was caught wrongly
2886 - Username/password not passed on to radius server due to wrong length.
2888 The presumption is that some radiusclient API changes for 4.51/PH/17
2889 were not taken care of correctly. The code is still untested by me (my
2890 Linux distribution still has 0.3.2 of radiusclient), but it was
2891 contributed by a Radius user.
2893 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2894 expanding the "port" option of the smtp transport.
2896 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2897 while reading a MIME header. Thanks to Tom Hughes for a patch.
2899 PH/24 Include config.h inside local_scan.h so that configuration settings are
2902 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2903 that in an ACL for RCPT (for example), you can examine exactly what was
2906 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2907 commands, but it was not correctly comparing the address with the actual
2908 client host address. Thus, it would show the EHLO address in Received:
2909 header lines when this was not necessary.
2911 PH/27 Added the % operator to ${eval:}.
2913 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2914 it should be ignoring failures (because with -C, for example, it has lost
2915 privilege). It wasn't ignoring creation failures other than "already
2918 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2921 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2922 because an input line was too long, either on its own, or by virtue of
2923 too many continuations, the temporary file was not being removed, and the
2924 return code was incorrect.
2926 PH/31 Missing "BOOL" in function definition in filtertest.c.
2928 PH/32 Applied Sieve patches from the maintainer.
2930 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2932 PH/33 Added "verify = not_blind".
2934 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2935 Local/Makefile (with some defaults set). These are used in built scripts
2936 such as exicyclog, but they have never been used in the exim_install
2937 script (though there are many overriding facilities there). I have
2938 arranged that the exim_install script now takes note of these two
2941 PH/35 Installed configuration files for Dragonfly.
2943 PH/36 When a locally submitted message by a trusted user did not contain a
2944 From: header, and the sender address was obtained from -f or from an SMTP
2945 MAIL command, and the trusted user did not use -F to supply a sender
2946 name, $originator_name was incorrectly used when constructing a From:
2947 header. Furthermore, $originator_name was used for submission mode
2948 messages from external hosts without From: headers in a similar way,
2949 which is clearly wrong.
2951 PH/37 Added control=suppress_local_fixups.
2953 PH/38 When log_selector = +received_sender was set, and the addition of the
2954 sender made the log line's construction buffer exactly full, or one byte
2955 less than full, an overflow happened when the terminating "\n" was
2958 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2959 when the result of a list match is failure because a DNS lookup failed.
2961 PH/40 RM_COMMAND is now used in the building process.
2963 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2964 the "build-* directories that it finds.
2966 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2967 address in a domain literal was a prefix of an interface address.
2969 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2970 when verifying a sender address, unless rewrite_headers is false.
2972 PH/44 Wrote a long comment about why errors_to addresses are verified as
2973 recipients, not senders.
2975 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2976 the ratelimit ACL was added.
2978 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2980 PH/46 Added extra information about PostgreSQL errors to the error string.
2982 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2983 avoiding the need to install special zones in a real server. This is
2984 backwards compatible; if it can't find the fake resolver, it drops back.
2985 Thus, both old and new test suites can be run.
2987 TF/02 Added util/ratelimit.pl
2989 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2990 clock is set back in time.
2992 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2993 Candler <B.Candler@pobox.com>.
2995 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2996 OK for addresses that are the result of redirections.
2998 PH/48 A number of further additions for the benefit of the new test suite,
2999 including a fake gethostbyname() that interfaces to the fake DNS resolver
3002 TF/06 The fix for widen_domains has also been applied to qualify_single and
3003 search_parents which are the other dnslookup options that can cause
3006 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3009 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3011 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3012 service" if it encountered an MX record with an empty target hostname.
3013 The message is now "an MX or SRV record indicated no SMTP service".
3015 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3016 if the DNS of the sending site is misconfigured. This is quite a
3017 common situation. This change restores the behaviour of treating a
3018 helo verification defer as a failure.
3020 PH/51 If self=fail was set on a router, the bounce message did not include the
3021 actual error message.
3027 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3029 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3030 command itself ended in error, the underlying message about the transport
3031 filter timeout was being overwritten with the pipe command error. Now the
3032 underlying error message should be appended to the second error message.
3034 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3036 PH/02 Reduce the amount of output that "make" produces by default. Full output
3037 can still be requested.
3039 PH/03 The warning log line about a condition test deferring for a "warn" verb
3040 was being output only once per connection, rather than after each
3041 occurrence (because it was using the same function as for successful
3042 "warn" verbs). This seems wrong, so I have changed it.
3044 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3045 it should not have, which might have caused a crash in the right
3046 circumstances, but probably never did.
3048 PH/04 Installed a modified version of Tony Finch's patch to make submission
3049 mode fix the return path as well as the Sender: header line, and to
3050 add a /name= option so that you can make the user's friendly name appear
3053 TF/03 Added the control = fakedefer ACL modifier.
3055 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3056 Mark Lowes for thorough testing.
3058 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3060 TK/03 Merged latest SRS patch from Miles Wilton.
3062 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3063 with the definition in sysexits.h (which is #included earlier).
3064 Fortunately, Exim does not actually use EX_OK. The code used to try to
3065 preserve the sysexits.h value, by assumimg that macro definitions were
3066 scanned for macro replacements. I have been disabused of this notion,
3067 so now the code just undefines EX_OK before #including unistd.h.
3069 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3070 in the smtp transport. When a block could not be written in a single
3071 write() function, the timeout was being re-applied to each part-write.
3072 This seems wrong - if the receiver was accepting one byte at a time it
3073 would take for ever. The timeout is now adjusted when this happens. It
3074 doesn't have to be particularly precise.
3076 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3077 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3079 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3080 without a domain if the check to <postmaster@domain> fails.
3082 SC/01 Eximstats: added -xls and the ability to specify output files
3083 (patch written by Frank Heydlauf).
3085 SC/02 Eximstats: use FileHandles for outputing results.
3087 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3089 SC/04 Eximstats: fixed display of large numbers with -nvr option
3091 SC/05 Eximstats: fixed merging of reports with empty tables.
3093 SC/06 Eximstats: added the -include_original_destination flag
3095 SC/07 Eximstats: removed tabs and trailing whitespace.
3097 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3099 TK/06 MBOX spool code: Add real "From " MBOX separator line
3100 so the .eml file is really in mbox format (even though
3101 most programs do not really care). Patch from Alex Miller.
3103 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3104 The latter is generated from $received_to and is only set if the
3105 message has one envelope recipient. SA can use these headers,
3106 obviously out-of-the-box. Patch from Alex Miller.
3108 PH/08 The ${def test on a variable was returning false if the variable's
3109 value was "0", contrary to what the specification has always said!
3110 The result should be true unless the variable is empty.
3112 PH/09 The syntax error of a character other than { following "${if
3113 def:variable_name" (after optional whitespace) was not being diagnosed.
3114 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3115 accidental colon was present, for example, could give incorrect results.
3117 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3118 result is used (not including appendfile, where other changes are about
3121 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3122 This involved changing a lot of size variables from int to off_t. It
3123 should work with maildirs and everything.
3125 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3126 spamd dying while we are connected to it.
3128 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3131 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3132 function for BDB 4.3.
3134 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3136 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3137 the sources for PCRE have been re-arranged and more files are now
3140 PH/15 The code I had for printing potentially long long variables in PH/11
3141 above was not the best (it lost precision). The length of off_t variables
3142 is now inspected at build time, and an appropriate printing format (%ld
3143 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3144 to be "long long int" or "long int". This is needed for the internal
3145 formatting function string_vformat().
3147 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3148 the configuration file to be ":syslog", then the script "guesses" where
3149 the logs files are, rather than using the compiled in default. In our
3150 case the guess is not the same as the compiled default, so the script
3151 suddenly stopped working when I started to use syslog. The patch checks
3152 to see if log_file_path is "". If so, it attempts to read it from exim
3153 with no configuration file to get the compiled in version, before it
3154 falls back to the previous guessing code."
3156 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3157 implementing BATV in an Exim configuration. See NewStuff for the gory
3160 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3161 Makefile that are specific to HP-UX.
3163 PH/18 If the "use_postmaster" option was set for a recipient callout together
3164 with the "random" option, the postmaster address was used as the MAIL
3165 FROM address for the random test, but not for the subsequent recipient
3166 test. It is now used for both.
3168 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3169 patch removes a few documentation additions to RFC 3028, because the
3170 latest draft now contains them. It adds the new en;ascii-case comparator
3171 and a new error check for 8bit text in MIME parts. Comparator and
3172 require names are now matched exactly. I enabled the subaddress
3173 extension, but it is not well tested yet (read: it works for me)."
3175 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3176 rework some of the code of TK/09 above to avoid the hardwired use of
3177 "%lld" and "long long". Replaced the call to snprintf() with a call to
3180 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3181 records point to non-existent hosts", "retry timeout exceeded", and
3182 "retry time not reached for any host after a long failure period".
3184 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3185 experimental DomainKeys support:
3187 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3188 (2) On an error such as an illegally used "control", the wrong name for
3189 the control was given.
3191 These problems did NOT occur unless DomainKeys support was compiled.
3193 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3195 PH/24 Added ${if match_ip condition.
3197 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3198 databases so that it will be absolutely obvious if a crash occurs in the
3199 DB library. This is a regular occurrence (often caused by mis-matched
3202 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3203 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3204 picked up on a user's system that detects such things. There doesn't seem
3205 to be a gcc warning option for this - only an attribute that has to be
3206 put on the function's prototype. It seems that in Fedora Core 4 they have
3207 set this on a number of new functions. No doubt there will be more in due
3210 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3211 specify a transport. However, if an address that was verified by such a
3212 router was the subject of a callout, Exim crashed because it tried to
3213 read the rcpt_include_affixes from the non-existent transport. Now it
3214 just assumes that the setting of that option is false. This bug was
3215 introduced by 4.51/PH/31.
3217 PH/28 Changed -d+all to exclude +memory, because that information is very
3218 rarely of interest, but it makes the output a lot bigger. People tend to
3219 do -d+all out of habit.
3221 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3222 code in os-type was giving problems when libc.so lives in lib64, like on
3225 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3226 aren't the modern standard, and it seems that some systems' include files
3227 don't always have them. Exim was already checking for some of the newer
3228 ones like T_AAAA, and defining it itself. I've added checks for all the
3229 record types that Exim uses.
3231 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3232 not automatically generating a new one, as it is supposed to. This
3233 prevented TLS from working. If the file did exist, but contained invalid
3234 data, a new version was generated, as expected. It was only the case of a
3235 non-existent file that was broken.
3237 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3238 with a change in libdomainkeys > 0.64.
3240 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3241 from DNS. If the selector record carries the flag, it now has
3242 precedence over the domain-wide flag.
3244 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3246 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3247 the use of a port name that isn't defined in /etc/services) occurred, the
3248 message was deferred as in a normal delivery, and thus remained on the
3249 spool, instead of being failed because of the mua_wrapper setting. This
3250 is now fixed, and I tidied up some of the mua_wrapper messages at the
3253 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3254 the messages in a hash of arrays rather than using individual hashes.
3255 This is a bit cleaner and results in dramatic memory savings, albeit
3256 at a slight CPU cost.
3258 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3259 as requested by Marc Sherman.
3261 SC/10 Eximstats: added histograms for user specified patterns as requested
3264 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3266 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3267 fopen() in the content-scanning modules that did not already have it.
3273 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3274 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3276 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3278 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3280 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3281 to test Sieve filters that use "vacation".
3283 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3284 that changes the way the GnuTLS parameters are stored in the cache file.
3285 The new format can be generated externally. For backward compatibility,
3286 if the data in the cache doesn't make sense, Exim assumes it has read an
3287 old-format file, and it generates new data and writes a new file. This
3288 means that you can't go back to an older release without removing the
3291 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3292 work if there are any delivery delays because "one_time" forces the
3293 parent to be marked "delivered", so its unseen clone is never tried
3294 again. For this reason, Exim now forbids the simultaneous setting of
3297 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3298 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3299 entry if you want to know the details. The fix, however, neglected to
3300 consider the case when local delivery batching is involved. The test for
3301 "previously delivered" was not happening when checking to see if an
3302 address could be batched with a previous (undelivered) one; under
3303 certain circumstances this could lead to multiple deliveries to the same
3306 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3307 in its include files, and this causes problems building Exim.
3309 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3310 header_syntax) but Exim was just ignoring anything given after a slash.
3311 In particular, this caused confusion with an attempt to use "verify =
3312 reverse_host_lookup/defer_ok". An error is now given when options are
3313 supplied for verify items that do not have them. (Maybe reverse_host_
3314 lookup should have a defer_ok option, but that's a different point.)
3316 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3317 defined by RFC 821) to 2048, because there were problems with some AUTH
3318 commands, and RFC 1869 says the size should be increased for extended
3319 SMTP commands that take arguments.
3321 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3324 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3325 "unknown" error; now it says that the functionality isn't in the binary.
3327 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3328 an address' error message when a string expansion fails (syntax or
3329 whatever). Otherwise the password may appear in the log. Following change
3330 PH/42 below, there is no longer a chance of it appearing in a bounce
3333 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3335 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3336 of its addresses was ever tried. (Bugzilla bug #2.)
3338 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3339 the result incorrectly in the debug output. (It correctly added a newline
3340 to what was transported.)
3342 TF/01 Added $received_time.
3344 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3345 commented out examples of how to interface to a virus scanner and to
3346 SpamAssassin. Also added commented examples of av_scanner and
3347 spamd_address settings.
3349 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3350 and controls are allowed in which ACLs. There were a couple of minor
3351 errors. Some of the entries in the conditions table (which is a table of
3352 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3353 negation of where the condition IS allowed.
3355 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3357 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3358 header file does not have a version number, so I've had to invent a new
3359 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3360 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3361 radiusclient), but it was contributed by a Radius user.
3363 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3364 files or directories, for OpenSSL.
3366 PH/19 When an Exim process that is running as root has to create an Exim log
3367 file, it does so in a subprocess that runs as exim:exim so as to get the
3368 ownership right at creation (otherwise, other Exim processes might see
3369 the file with the wrong ownership). There was no test for failure of this
3370 fork() call, which would lead to the process getting stuck as it waited
3371 for a non-existent subprocess. Forks do occasionally fail when resources
3372 run out. I reviewed all the other calls to fork(); they all seem to check
3375 PH/20 When checking for unexpected SMTP input at connect time (before writing
3376 the banner), Exim was not dealing correctly with a non-positive return
3377 from the read() function. If the client had disconnected by this time,
3378 the result was a log entry for a synchronization error with an empty
3379 string after "input=" when read() returned zero. If read() returned -1
3380 (an event I could not check), uninitialized data bytes were printed.
3381 There were reports of junk text (parts of files, etc) appearing after
3384 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3386 PH/22 Added support for macro redefinition, and (re)definition in between
3387 driver and ACL definitions.
3389 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3390 forgetting to use the resulting value; it was using the unexpanded value.
3392 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3393 hadn't been configured. The fix is from Juergen Kreileder, who
3394 understands it better than I do:
3396 "Here's what I see happening with three configured cyrus_sasl
3397 authenticators configured (plain, login, cram-md5):
3399 On startup auth_cyrus_sasl_init() gets called for each of these.
3400 This means three calls to sasl_listmech() without a specified mech_list.
3401 => SASL tests which mechs of all available mechs actually work
3402 => three warnings about OTP not working
3403 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3405 With the patch, sasl_listmech() also gets called three times. But now
3406 SASL's mech_list option is set to the server_mech specified in the the
3407 authenticator. Or in other words, the answer from sasl_listmech()
3408 gets limited to just the mech you're testing for (which is different
3410 => the return list contains just 'plain' or 'login', 'cram-md5' or
3411 nothing depending on the value of ob->server_mech.
3413 I've just tested the patch: Authentication still works fine,
3414 unavailable mechs specified in the exim configuration are still
3415 caught, and the auth.log warnings about OTP are gone."
3417 PH/25 When debugging is enabled, the contents of the command line are added
3418 to the debugging output, even when log_selector=+arguments is not
3421 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3422 answer is "GNU", and only if the return is "GNU/something" is the answer
3425 PH/27 $acl_verify_message is now set immediately after the failure of a
3426 verification in an ACL, and so is available in subsequent modifiers. In
3427 particular, the message can be preserved by coding like this:
3429 warn !verify = sender
3430 set acl_m0 = $acl_verify_message
3432 Previously, $acl_verify_message was set only while expanding "message"
3433 and "log_message" when a very denied access.
3435 PH/28 Modified OS/os.c-Linux with
3437 -#ifndef OS_LOAD_AVERAGE
3438 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3440 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3441 nomenclature these days.)
3443 PH/29 Installed patch from the Sieve maintainer that adds the options
3444 sieve_useraddress and sieve_subaddress to the redirect router.
3446 PH/30 In these circumstances:
3447 . Two addresses routed to the same list of hosts;
3448 . First host does not offer TLS;
3449 . First host accepts first address;
3450 . First host gives temporary error to second address;
3451 . Second host offers TLS and a TLS session is established;
3452 . Second host accepts second address.
3453 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3454 and peerdn, if requested) that were in fact used only for the second
3457 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3458 attention to any local part prefix or suffix that was matched by the
3459 router that accepted the address. It now behaves in the same way as it
3460 does for delivery: the affixes are removed from the local part unless
3461 rcpt_include_affixes is set on the transport.
3463 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3464 timeout during the DATA phase of an incoming message.
3466 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3467 applied a patch sanctioned by the Sieve maintainer.
3469 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3470 the uid or gid is negative. A case of a negative gid caused this to be
3471 noticed. The fix allows for either to be negative.
3473 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3474 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3477 PH/36 The change PH/12 above was broken. Fixed it.
3479 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3480 the grounds that routing the same address twice would always produce the
3481 same answer. This might have been true once, but it is certainly no
3482 longer true now. Routing a child address may depend on the previous
3483 routing that produced that child. Some complicated redirection strategies
3484 went wrong when messages had multiple recipients, and made Exim's
3485 behaviour dependent on the order in which the addresses were given.
3487 I have moved the duplicate checking until after the routing is complete.
3488 Exim scans the addresses that are assigned to local and remote
3489 transports, and removes any duplicates. This means that more work will be
3490 done, as duplicates will always all be routed, but duplicates are
3491 presumably rare, so I don't expect this is of any significance.
3493 For deliveries to pipes, files, and autoreplies, the duplicate checking
3494 still happens during the routing process, since they are not going to be
3497 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3498 It corrects a timeout issue with spamd. This is Ian's comment: "The
3499 background is that sometimes spamd either never reads data from a
3500 connection it has accepted, or it never writes response data. The exiscan
3501 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3502 blindly assumes that writes won't block so it may never time out."
3504 PH/39 Allow G after quota size as well as K and M.
3506 PH/40 The value set for $authenticated_id in an authenticator may not contain
3507 binary zeroes or newlines because the value is written to log lines and
3508 to spool files. There was no check on this. Now the value is run through
3509 the string_printing() function so that such characters are converted to
3510 printable escape sequences.
3512 PH/41 $message_linecount is a new variable that contains the total number of
3513 lines in the message. Compare $body_linecount, which is the count for the
3516 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3517 bounce and delay warning messages, except in certain special cases, which
3520 (a) An SMTP error message from a remote host;
3521 (b) A message specified in a :fail: redirection;
3522 (c) A message specified in a "fail" command in a system filter;
3523 (d) A message specified in a FAIL return from the queryprogram router;
3524 (e) A message specified by the cannot_route_message router option.
3526 In these cases only, Exim does include the error details in bounce and
3527 warning messages. There are also a few cases where bland messages such
3528 as "unrouteable address" or "local delivery error" are given.
3530 PH/43 $value is now also set for the "else" part of a ${run expansion.
3532 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3533 being worked on, but at least Exim now implements the latest version to
3536 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3537 process to complete was treated as a delivery failure, a timeout while
3538 writing the message to the pipe was logged, but erroneously treated as a
3539 successful delivery. Such timeouts include transport filter timeouts. For
3540 consistency with the overall process timeout, these timeouts are now
3541 treated as errors, giving rise to delivery failures by default. However,
3542 there is now a new Boolean option for the pipe transport called
3543 timeout_defer, which, if set TRUE, converts the failures into defers for
3544 both kinds of timeout. A transport filter timeout is now identified in
3547 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3548 systems where "make" and "gmake" are different, calling "gmake" at top
3549 level broke things. I've arranged for the value of $(MAKE) to be passed
3550 from the Makefile to this script so that it can call the same version of
3554 A note about Exim versions 4.44 and 4.50
3555 ----------------------------------------
3557 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3558 changes of various kinds. As a consequence, a big documentation update was
3559 needed. This delayed the release for rather longer than seemed good, especially
3560 in the light of a couple of (minor) security issues. Therefore, the changes
3561 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3562 release. So 4.44 and 4.50 are in effect two different branches that both start
3565 I have left the 4.50 change log unchanged; it contains all the changes since
3566 4.43. The change log for 4.44 is below; many of its items are identical to
3567 those for 4.50. This seems to be the most sensible way to preserve the
3568 historical information.
3574 1. Minor wording change to the doc/README.SIEVE file.
3576 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3577 computation of the current number of files was incorrect.
3579 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3580 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3583 4. Give more explanation in the error message when the command for a transport
3584 filter fails to execute.
3586 5. There are several places where Exim runs a non-Exim command in a
3587 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3588 was being done only for the command run by the queryprogram router. It is
3589 now done for all such subprocesses. The other cases are: ${run, transport
3590 filters, and the commands run by the lmtp and pipe transports.
3592 6. Added CONFIGURE_GROUP build-time option.
3594 7. Some older OS have a limit of 256 on the maximum number of file
3595 descriptors. Exim was using setrlimit() to set 1000 as a large value
3596 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3597 systems. I've change it so that if it can't get 1000, it tries for 256.
3599 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3600 was an oversight, and furthermore, ever since the addition of extra
3601 controls (e.g. 4.43/32), the checks on when to allow different forms of
3602 "control" were broken. There should now be diagnostics for all cases when a
3603 control that does not make sense is encountered.
3605 9. Added the /retain_sender option to "control=submission".
3607 10. $recipients is now available in the predata ACL (oversight).
3609 11. Tidy the search cache before the fork to do a delivery from a message
3610 received from the command line. Otherwise the child will trigger a lookup
3611 failure and thereby defer the delivery if it tries to use (for example) a
3612 cached ldap connection that the parent has called unbind on.
3614 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3615 of $address_data from the recipient verification was clobbered by the
3616 sender verification.
3618 13. The value of address_data from a sender verification is now available in
3619 $sender_address_data in subsequent conditions in the ACL statement.
3621 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3623 15. Added a new option "connect=<time>" to callout options, to set a different
3626 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3627 was its contents. (It was OK if the option was not defined at all.)
3629 17. A "Completed" log line is now written for messages that are removed from
3630 the spool by the -Mrm option.
3632 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3633 information about exactly what failed.
3635 19. Added -dd to debug only the daemon process.
3637 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3638 handles timeouts, both on the server side and network timeouts. Renamed the
3639 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3641 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3642 It is now set to "smtps".
3644 22. $host_address is now set to the target address during the checking of
3645 ignore_target_hosts.
3647 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3648 being passed; this would have caused $sender_host_name to have been used if
3649 matching the list had actually called for a host name (not very likely,
3650 since this list is usually IP addresses). A host name is now passed as
3653 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3654 code that specifies a non-restarting handler (typically sigaction() in
3655 modern systems) in an attempt to fix a rare and obscure crash bug.
3657 25. Narrowed the window for a race in the daemon that could cause it to ignore
3658 SIGCHLD signals. This is not a major problem, because they are used only to
3659 wake it up if nothing else does.
3661 26. A malformed maildirsize file could cause Exim to calculate negative values
3662 for the mailbox size or file count. Odd effects could occur as a result.
3663 The maildirsize information is now recalculated if the size or filecount
3666 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3667 support for a long time. Removed HAVE_SYS_VFS_H.
3669 28. Installed the latest version of exipick from John Jetmore.
3671 29. In an address list, if the pattern was not a regular expression, an empty
3672 subject address (from a bounce message) matched only if the pattern was an
3673 empty string. Non-empty patterns were not even tested. This was the wrong
3674 because it is perfectly reasonable to use an empty address as part of a
3675 database query. An empty address is now tested by patterns that are
3676 lookups. However, all the other forms of pattern expect the subject to
3677 contain a local part and a domain, and therefore, for them, an empty
3678 address still always fails if the pattern is not itself empty.
3680 30. Exim went into a mad DNS loop when attempting to do a callout where the
3681 host was specified on an smtp transport, and looking it up yielded more
3682 than one IP address.
3684 31. Re-factored the code for checking spool and log partition space into a
3685 function that finds that data and another that does the check. The former
3686 is then used to implement four new variables: $spool_space, $log_space,
3687 $spool_inodes, and $log_inodes.
3689 32. The RFC2047 encoding function was originally intended for short strings
3690 such as real names; it was not keeping to the 75-character limit for
3691 encoded words that the RFC imposes. It now respects the limit, and
3692 generates multiple encoded words if necessary. To be on the safe side, I
3693 have increased the buffer size for the ${rfc2047: expansion operator from
3696 33. It is now permitted to omit both strings after an "if" condition; if the
3697 condition is true, the result is "true". As before, when the second string
3698 is omitted, a false condition yields an empty string. This makes it less
3699 cumbersome to write custom ACL and router conditions.
3701 34. Failure to deliver a bounce message always caused it to be frozen, even if
3702 there was an errors_to setting on the router. The errors_to setting is now
3705 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3706 canonical form (fully expanded) before being placed in
3707 $sender_host_address.
3709 36. The table in the code that translates DNS record types into text (T_A to
3710 "A" for instance) was missing entries for NS and CNAME. It is just possible
3711 that this could have caused confusion if both these types were looked up
3712 for the same domain, because the text type is used as part of Exim's
3713 per-process caching. But the chance of anyone hitting this buglet seems
3716 37. The dnsdb lookup has been extended in a number of ways.
3718 (1) There is a new type, "zns", which walks up the domain tree until it
3719 finds some nameserver records. It should be used with care.
3721 (2) There is a new type, "mxh", which is like "mx" except that it returns
3722 just the host names, not the priorities.
3724 (3) It is now possible to give a list of domains (or IP addresses) to be
3725 looked up. The behaviour when one of the lookups defers can be
3726 controlled by a keyword.
3728 (4) It is now possible to specify the separator character for use when
3729 multiple records are returned.
3731 38. The dnslists ACL condition has been extended: it is now possible to supply
3732 a list of IP addresses and/or domains to be looked up in a particular DNS
3735 39. Added log_selector=+queue_time_overall.
3737 40. When running the queue in the test harness, wait just a tad after forking a
3738 delivery process, to get repeatability of debugging output.
3740 41. Include certificate and key file names in error message when GnuTLS fails
3741 to set them up, because the GnuTLS error message doesn't include the name
3742 of the failing file when there is a problem reading it.
3744 42. Allow both -bf and -bF in the same test run.
3746 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3748 44. The "Exiscan patch" is now merged into the mainline Exim source.
3750 45. Sometimes the final signoff response after QUIT could fail to get
3751 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3752 < 0 before doing a fflush(). This bug looks as though it goes back to the
3753 introduction of TLS in release 3.20, but "sometimes" must have been rare
3754 because the tests only now provoked it.
3756 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3757 (this can affect the format of dates).
3759 47. exim_tidydb, when checking for the continued existence of a message for
3760 which it has found a message-specific retry record, was not finding
3761 messages that were in split spool directories. Consequently, it was
3762 deleting retry records that should have stayed in existence.
3764 48. Steve fixed some bugs in eximstats.
3766 49. The SPA authentication driver was not abandoning authentication and moving
3767 on to the next authenticator when an expansion was forced to fail,
3768 contradicting the general specification for all authenticators. Instead it
3769 was generating a temporary error. It now behaves as specified.
3771 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3772 (the order specifies the preference for clients). The order is now AES256,
3773 AES128, 3DES, ARCFOUR128.
3775 51. Small patch to Sieve code - explicitly set From: when generating an
3778 52. Exim crashed if a remote delivery caused a very long error message to be
3779 recorded - for instance if somebody sent an entire SpamAssassin report back
3780 as a large number of 550 error lines. This bug was coincidentally fixed by
3781 increasing the size of one of Exim's internal buffers (big_buffer) that
3782 happened as part of the Exiscan merge. However, to be on the safe side, I
3783 have made the code more robust (and fixed the comments that describe what
3786 53. Now that there can be additional text after "Completed" in log lines (if
3787 the queue_time_overall log selector is set), a one-byte patch to exigrep
3788 was needed to allow it to recognize "Completed" as not the last thing in
3791 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3792 patch that reportedly fixes this has been added. I am not expert enough to
3793 create a test for it. This is what the patch creator wrote:
3795 "I found a little strange behaviour of ldap code when working with
3796 Windows 2003 AD Domain, where users was placed in more than one
3797 Organization Units. When I tried to give exim partial DN, the exit code
3798 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3799 But simultaneously result of request was absolutely normal ldap result,
3800 so I produce this patch..."
3802 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3803 REFERENCE, so I have modified the code to exclude the patch when that macro
3806 55. Some experimental protocols are using DNS PTR records for new purposes. The
3807 keys for these records are domain names, not reversed IP addresses. The
3808 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3809 leaves it alone. Component reversal etc. now happens only for IP addresses.
3812 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3814 57. Double the size of the debug message buffer (to 2048) so that more of very
3815 long debug lines gets shown.
3817 58. The exicyclog utility now does better if the number of log files to keep
3818 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3820 59. Two changes related to the smtp_active_hostname option:
3822 (1) $smtp_active_hostname is now available as a variable.
3823 (2) The default for smtp_banner uses $smtp_active_hostname instead
3824 of $primary_hostname.
3826 60. The host_aton() function is supposed to be passed a string that is known
3827 to be a valid IP address. However, in the case of IPv6 addresses, it was
3828 not checking this. This is a hostage to fortune. Exim now panics and dies
3829 if the condition is not met. A case was found where this could be provoked
3830 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3831 components; fortuitously, this particular loophole had already been fixed
3832 by change 4.50/55 above.
3834 If there are any other similar loopholes, the new check in host_aton()
3835 itself should stop them being exploited. The report I received stated that
3836 data on the command line could provoke the exploit when Exim was running as
3837 exim, but did not say which command line option was involved. All I could
3838 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3839 running as the user.
3842 61. There was a buffer overflow vulnerability in the SPA authentication code
3843 (which came originally from the Samba project). I have added a test to the
3844 spa_base64_to_bits() function which I hope fixes it.
3847 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3848 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3850 63. The daemon start-up calls getloadavg() while still root for those OS that
3851 need the first call to be done as root, but it missed one case: when
3852 deliver_queue_load_max is set with deliver_drop_privilege. This is
3853 necessary for the benefit of the queue runner, because there is no re-exec
3854 when deliver_drop_privilege is set.
3856 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3857 This has been fixed.
3859 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3860 was in use, was not putting the data itself into the right store pool;
3861 consequently, it could be overwritten for a subsequent message in the same
3862 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3865 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3867 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3868 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3869 the function were treating the return as a boolean value, which happened to
3870 work because 0=false and not-0=true, but is not correct code.
3872 68. The host_aton() function was not handling scoped IPv6 addresses (those
3873 with, for example, "%eth0" on the end) correctly.
3875 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3876 negated items (that is, ~something) in unsigned ints. Some compilers
3877 apparently mutter when there is no cast.
3879 70. If an address verification called from an ACL failed, and did not produce a
3880 user-specific message (i.e. there was only a "system" message), nothing was
3881 put in $acl_verify_message. In this situation, it now puts the system
3884 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3885 session; change 4.31/43 added the unwanted input to the log line - except
3886 that it did not do this in the start of session case. It now does.
3888 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3889 This is wrong and can cause the other end to generate a synchronization
3890 error if it is another Exim or anything else that does the synchronization
3891 check. A QUIT command is no longer sent after a timeout.
3893 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3894 during host lookups.
3896 74. The defer_ok option of callout verification was not working if it was used
3897 when verifying addresses in header lines, that is, for this case:
3899 verify = header_sender/callout=defer_ok
3901 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3902 those file descriptors could be used for SMTP connections. If anything
3903 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3904 could be sent to the SMTP client, causing chaos. The daemon now opens
3905 stdin, stdout, and stderr to /dev/null when it puts itself into the
3908 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3909 log by default. The user can override this with suitable Perl magic.
3911 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3912 the log message when discard triggers, was not working for the DATA ACL or
3913 for the non-SMTP ACL.
3915 78. Error message wording change in sieve.c.
3917 79. If smtp_accept_max_per_host was set, the number of connections could be
3918 restricted to fewer than expected, because the daemon was trying to set up
3919 a new connection before checking whether the processes handling previous
3920 connections had finished. The check for completed processes is now done
3921 earlier. On busy systems, this bug wouldn't be noticed because something
3922 else would have woken the daemon, and it would have reaped the completed
3925 80. If a message was submitted locally by a user whose login name contained one
3926 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3927 It caused a spool format error. I have fixed the spool reading code. A
3928 related problem was that the "from" clause in the Received: line became
3929 illegal because of the space(s). It is now covered by ${quote_local_part.
3931 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3934 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3936 83. Patch from Sieve maintainer to fix unterminated string problem in
3937 "vacation" handling.
3939 84. Some minor changes to the Linux configuration files to help with other
3940 OS variants using glibc.
3942 85. One more patch for Sieve to update vacation handling to latest spec.
3945 ----------------------------------------------------
3946 See the note above about the 4.44 and 4.50 releases.
3947 ----------------------------------------------------
3953 1. Change 4.43/35 introduced a bug that caused file counts to be
3954 incorrectly computed when quota_filecount was set in an appendfile
3957 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3958 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3961 3. Give more explanation in the error message when the command for a transport
3962 filter fails to execute.
3964 4. There are several places where Exim runs a non-Exim command in a
3965 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3966 was being done only for the command run by the queryprogram router. It is
3967 now done for all such subprocesses. The other cases are: ${run, transport
3968 filters, and the commands run by the lmtp and pipe transports.
3970 5. Some older OS have a limit of 256 on the maximum number of file
3971 descriptors. Exim was using setrlimit() to set 1000 as a large value
3972 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3973 systems. I've change it so that if it can't get 1000, it tries for 256.
3975 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3976 was an oversight, and furthermore, ever since the addition of extra
3977 controls (e.g. 4.43/32), the checks on when to allow different forms of
3978 "control" were broken. There should now be diagnostics for all cases when a
3979 control that does not make sense is encountered.
3981 7. $recipients is now available in the predata ACL (oversight).
3983 8. Tidy the search cache before the fork to do a delivery from a message
3984 received from the command line. Otherwise the child will trigger a lookup
3985 failure and thereby defer the delivery if it tries to use (for example) a
3986 cached ldap connection that the parent has called unbind on.
3988 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3989 of $address_data from the recipient verification was clobbered by the
3990 sender verification.
3992 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3993 was its contents. (It was OK if the option was not defined at all.)
3995 11. A "Completed" log line is now written for messages that are removed from
3996 the spool by the -Mrm option.
3998 12. $host_address is now set to the target address during the checking of
3999 ignore_target_hosts.
4001 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4002 being passed; this would have caused $sender_host_name to have been used if
4003 matching the list had actually called for a host name (not very likely,
4004 since this list is usually IP addresses). A host name is now passed as
4007 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4008 code that specifies a non-restarting handler (typically sigaction() in
4009 modern systems) in an attempt to fix a rare and obscure crash bug.
4011 15. Narrowed the window for a race in the daemon that could cause it to ignore
4012 SIGCHLD signals. This is not a major problem, because they are used only to
4013 wake it up if nothing else does.
4015 16. A malformed maildirsize file could cause Exim to calculate negative values
4016 for the mailbox size or file count. Odd effects could occur as a result.
4017 The maildirsize information is now recalculated if the size or filecount
4020 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4021 support for a long time. Removed HAVE_SYS_VFS_H.
4023 18. Updated exipick to current release from John Jetmore.
4025 19. Allow an empty sender to be matched against a lookup in an address list.
4026 Previously the only cases considered were a regular expression, or an
4029 20. Exim went into a mad DNS lookup loop when doing a callout where the
4030 host was specified on the transport, if the DNS lookup yielded more than
4033 21. The RFC2047 encoding function was originally intended for short strings
4034 such as real names; it was not keeping to the 75-character limit for
4035 encoded words that the RFC imposes. It now respects the limit, and
4036 generates multiple encoded words if necessary. To be on the safe side, I
4037 have increased the buffer size for the ${rfc2047: expansion operator from
4040 22. Failure to deliver a bounce message always caused it to be frozen, even if
4041 there was an errors_to setting on the router. The errors_to setting is now
4044 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4045 canonical form (fully expanded) before being placed in
4046 $sender_host_address.
4048 24. Updated eximstats to version 1.33
4050 25. Include certificate and key file names in error message when GnuTLS fails
4051 to set them up, because the GnuTLS error message doesn't include the name
4052 of the failing file when there is a problem reading it.
4054 26. Expand error message when OpenSSL has problems setting up cert/key files.
4057 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4058 (this can affect the format of dates).
4060 28. exim_tidydb, when checking for the continued existence of a message for
4061 which it has found a message-specific retry record, was not finding
4062 messages that were in split spool directories. Consequently, it was
4063 deleting retry records that should have stayed in existence.
4065 29. eximstats updated to version 1.35
4066 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4067 1.35 - bugfix such that pie charts by volume are generated correctly
4069 30. The SPA authentication driver was not abandoning authentication and moving
4070 on to the next authenticator when an expansion was forced to fail,
4071 contradicting the general specification for all authenticators. Instead it
4072 was generating a temporary error. It now behaves as specified.
4074 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4075 (the order specifies the preference for clients). The order is now AES256,
4076 AES128, 3DES, ARCFOUR128.
4078 31. Small patch to Sieve code - explicitly set From: when generating an
4081 32. Exim crashed if a remote delivery caused a very long error message to be
4082 recorded - for instance if somebody sent an entire SpamAssassin report back
4083 as a large number of 550 error lines. This bug was coincidentally fixed by
4084 increasing the size of one of Exim's internal buffers (big_buffer) that
4085 happened as part of the Exiscan merge. However, to be on the safe side, I
4086 have made the code more robust (and fixed the comments that describe what
4089 33. Some experimental protocols are using DNS PTR records for new purposes. The
4090 keys for these records are domain names, not reversed IP addresses. The
4091 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4092 leaves it alone. Component reversal etc. now happens only for IP addresses.
4095 34. The host_aton() function is supposed to be passed a string that is known
4096 to be a valid IP address. However, in the case of IPv6 addresses, it was
4097 not checking this. This is a hostage to fortune. Exim now panics and dies
4098 if the condition is not met. A case was found where this could be provoked
4099 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4100 components; fortuitously, this particular loophole had already been fixed
4101 by change 4.50/55 or 4.44/33 above.
4103 If there are any other similar loopholes, the new check in host_aton()
4104 itself should stop them being exploited. The report I received stated that
4105 data on the command line could provoke the exploit when Exim was running as
4106 exim, but did not say which command line option was involved. All I could
4107 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4108 running as the user.
4111 35. There was a buffer overflow vulnerability in the SPA authentication code
4112 (which came originally from the Samba project). I have added a test to the
4113 spa_base64_to_bits() function which I hope fixes it.
4116 36. The daemon start-up calls getloadavg() while still root for those OS that
4117 need the first call to be done as root, but it missed one case: when
4118 deliver_queue_load_max is set with deliver_drop_privilege. This is
4119 necessary for the benefit of the queue runner, because there is no re-exec
4120 when deliver_drop_privilege is set.
4122 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4123 was in use, was not putting the data itself into the right store pool;
4124 consequently, it could be overwritten for a subsequent message in the same
4125 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4128 38. Sometimes the final signoff response after QUIT could fail to get
4129 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4130 < 0 before doing a fflush(). This bug looks as though it goes back to the
4131 introduction of TLS in release 3.20, but "sometimes" must have been rare
4132 because the tests only now provoked it.
4138 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4139 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4140 Now it returns an integer. A number of calls were still expecting a T/F
4141 return. Fortuitously, in all cases, the tests worked in OK situations,
4142 which is the norm. However, things would have gone wrong on any write
4143 failures on the smtp file descriptor. This function is used when sending
4144 messages over SMTP and also when doing verify callouts.
4146 2. When Exim is called to do synchronous delivery of a locally submitted
4147 message (the -odf or -odi options), it no longer closes stderr before doing
4150 3. Implemented the mua_wrapper option.
4152 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4154 5. Implemented the functions header_remove(), header_testname(),
4155 header_add_at_position(), and receive_remove_recipient(), and exported them
4158 6. If an ACL "warn" statement specified the addition of headers, Exim already
4159 inserted X-ACL-Warn: at the start if there was no header name. However, it
4160 was not making this test for the second and subsequent header lines if
4161 there were newlines in the string. This meant that an invalid header could
4162 be inserted if Exim was badly configured.
4164 7. Allow an ACL "warn" statement to add header lines at the start or after all
4165 the Received: headers, as well as at the end.
4167 8. Added the rcpt_4xx retry error code.
4169 9. Added postmaster_mailfrom=xxx to callout verification option.
4171 10. Added mailfrom=xxxx to the callout verification option, for verify=
4174 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4175 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4177 12. Inserted some casts to stop certain compilers warning when using pointer
4178 differences as field lengths or precisions in printf-type calls (mostly
4179 affecting debugging statements).
4181 13. Added optional readline() support for -be (dynamically loaded).
4183 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4184 same clock tick as a message's arrival, so that its received time was the
4185 same as the "first fail" time on the retry record, and that message
4186 remained on the queue past the ultimate address timeout, every queue runner
4187 would try a delivery (because it was past the ultimate address timeout) but
4188 after another failure, the ultimate address timeout, which should have then
4189 bounced the address, did not kick in. This was a "< instead of <=" error;
4190 in most cases the first failure would have been in the next clock tick
4191 after the received time, and all would be well.
4193 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4194 being recognized when the domain list was tested by the match_domain
4195 condition in an expansion string.
4197 16. Added the ${str2b64: operator.
4199 17. Exim was always calling setrlimit() to set a large limit for the number of
4200 processes, without checking whether the existing limit was already
4201 adequate. (It did check for the limit on file descriptors.) Furthermore,
4202 errors from getrlimit() and setrlimit() were being ignored. Now they are
4203 logged to the main and panic logs, but Exim does carry on, to try to do its
4204 job under whatever limits there are.
4206 18. Imported PCRE 5.0.
4208 19. Trivial typo in log message " temporarily refused connection" (the leading
4211 20. If the log selector return_path_on_delivery was set and an address was
4212 redirected to /dev/null, the delivery process crashed because it assumed
4213 that a return path would always be set for a "successful" delivery. In this
4214 case, the whole delivery is bypassed as an optimization, and therefore no
4217 21. Internal re-arrangement: the function for sending a challenge and reading
4218 a response while authentication was assuming a zero-terminated challenge
4219 string. It's now changed to take a pointer and a length, to allow for
4220 binary data in such strings.
4222 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4224 23. Exim was not respecting finduser_retries when seeking the login of the
4225 uid under which it was called; it was always trying 10 times. (The default
4226 setting of finduser_retries is zero.) Also, it was sleeping after the final
4227 failure, which is pointless.
4229 24. Implemented tls_on_connect_ports.
4231 25. Implemented acl_smtp_predata.
4233 26. If the domain in control=submission is set empty, Exim assumes that the
4234 authenticated id is a complete email address when it generates From: or
4235 Sender: header lines.
4237 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4238 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4239 chown and chgrp in /bin and hostname in /usr/bin.
4241 28. Exim was keeping the "process log" file open after each use, just as it
4242 does for the main log. This opens the possibility of it remaining open for
4243 long periods when the USR1 signal hits a daemon. Occasional processlog
4244 errors were reported, that could have been caused by this. Anyway, it seems
4245 much more sensible not to leave this file open at all, so that is what now
4248 29. The long-running daemon process does not normally write to the log once it
4249 has entered its main loop, and it closes the log before doing so. This is
4250 so that log files can straightforwardly be renamed and moved. However,
4251 there are a couple of unusual error situations where the daemon does write
4252 log entries, and I had neglected to close the log afterwards.
4254 30. The text of an SMTP error response that was received during a remote
4255 delivery was being truncated at 512 bytes. This is too short for some of
4256 the long messages that one sometimes sees. I've increased the limit to
4259 31. It is now possible to make retry rules that apply only when a message has a
4260 specific sender, in particular, an empty sender.
4262 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4263 it possible to be selective about when SMTP synchronization is enforced.
4265 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4267 32. Implemented hosts_connection_nolog.
4269 33. Added an ACL for QUIT.
4271 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4274 35. Added mailbox_size and mailbox_filecount to appendfile.
4276 36. Added control = no_multiline_responses to ACLs.
4278 37. There was a bug in the logic of the code that waits for the clock to tick
4279 in the case where the clock went backwards by a substantial amount such
4280 that the microsecond fraction of "now" was more than the microsecond
4281 fraction of "then" (but the whole seconds number was less).
4283 38. Added support for the libradius Radius client library this is found on
4284 FreeBSD (previously only the radiusclient library was supported).
4290 1. When certain lookups returned multiple values in the form name=value, the
4291 quoting of the values was not always being done properly. Specifically:
4292 (a) If the value started with a double quote, but contained no whitespace,
4294 (b) If the value contained whitespace other than a space character (i.e.
4295 tabs or newlines or carriage returns) it was not quoted.
4296 This fix has been applied to the mysql and pgsql lookups by writing a
4297 separate quoting function and calling it from the lookup code. The fix
4298 should probably also be applied to nisplus, ibase and oracle lookups, but
4299 since I cannot test any of those, I have not disturbed their existing code.
4301 2. A hit in the callout cache for a specific address caused a log line with no
4302 reason for rejecting RCPT. Now it says "Previous (cached) callout
4303 verification failure".
4305 3. There was an off-by-one bug in the queryprogram router. An over-long
4306 return line was truncated at 256 instead of 255 characters, thereby
4307 overflowing its buffer with the terminating zero. As well as fixing this, I
4308 have increased the buffer size to 1024 (and made a note to document this).
4310 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4311 when Exim is waiting for an SMTP response from a remote server, Exim
4312 restarts its select() call on the socket, thereby resetting its timeout.
4313 This is not a problem when such interrupts are rare. Somebody set up a cron
4314 job to run exiwhat every 2 minutes, which is less than the normal select()
4315 timeout (5 or 10 minutes). This meant that the select() timeout never
4316 kicked in because it was always reset. I have fixed this by comparing the
4317 time when an interrupt arrives with the time at the start of the first call
4318 to select(). If more time than the timeout has elapsed, the interrupt is
4319 treated as a timeout.
4321 5. Some internal re-factoring in preparation for the addition of Sieve
4322 extensions (by MH). In particular, the "personal" test is moved to a
4323 separate function, and given an option for scanning Cc: and Bcc: (which is
4324 not set for Exim filters).
4326 6. When Exim created an email address using the login of the caller as the
4327 local part (e.g. when creating a From: or Sender: header line), it was not
4328 quoting the local part when it contained special characters such as @.
4330 7. Installed new OpenBSD configuration files.
4332 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4333 try to make them clearer.
4335 9. Callout options, other than the timeout value, were being ignored when
4336 verifying sender addresses in header lines. For example, when using
4338 verify = header_sender/callout=no_cache
4340 the cache was (incorrectly) being used.
4342 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4343 only the Cygwin environment.
4345 11. When return_path_on_delivery was set as a log selector, if different remote
4346 addresses in the same message used different return paths and parallel
4347 remote delivery occurred, the wrong values would sometimes be logged.
4348 (Whenever a remote delivery process finished, the return path value from
4349 the most recently started remote delivery process was logged.)
4351 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4352 lines when AUTH and/or TLS are in use. This is the "received protocol"
4353 field. Exim used to use "asmtp" for authenticated SMTP, without any
4354 indication (in the protocol name) for TLS use. Now it follows the RFC and
4355 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4356 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4357 names appear in log lines as well as in Received: header lines.
4359 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4360 extensions, and comparison tests, and to fix some bugs.
4362 14. Changes to the "personal" filter test:
4364 (1) The test was buggy in that it was just doing the equivalent of
4365 "contains" tests on header lines. For example, if a user's address was
4366 anne@some.where, the "personal" test would incorrectly be true for
4368 To: susanne@some.where
4370 This test is now done by extracting each address from the header in turn,
4371 and checking the entire address. Other tests that are part of "personal"
4372 are now done using regular expressions (for example, to check local parts
4373 of addresses in From: header lines).
4375 (2) The list of non-personal local parts in From: addresses has been
4376 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4377 taken from the Sieve specification recommendations.
4379 (3) If the message contains any header line starting with "List-" it is
4380 treated as non-personal.
4382 (4) The test for "circular" in the Subject: header line has been removed
4383 because it now seems ill-conceived.
4385 15. Minor typos in src/EDITME comments corrected.
4387 16. Installed latest exipick from John Jetmore.
4389 17. If headers_add on a router specified a text string that was too long for
4390 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4391 of string_sprintf() is now avoided.
4393 18. $message_body_size was not set (it was always zero) when running the DATA
4394 ACL and the local_scan() function.
4396 19. For the "mail" command in an Exim filter, no default was being set for
4397 the once_repeat time, causing a random time value to be used if "once" was
4398 specified. (If the value happened to be <= 0, no repeat happened.) The
4399 default is now 0s, meaning "never repeat". The "vacation" command was OK
4400 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4401 (I found it when inspecting the code).
4403 20. There is now an overall timeout for performing a callout verification. It
4404 defaults to 4 times the callout timeout, which applies to individual SMTP
4405 commands during the callout. The overall timeout applies when there is more
4406 than one host that can be tried. The timeout is checked before trying the
4407 next host. This prevents very long delays if there are a large number of
4408 hosts and all are timing out (e.g. when the network connections are timing
4409 out). The value of the overall timeout can be changed by specifying an
4410 additional sub-option for "callout", called "maxwait". For example:
4412 verify = sender/callout=5s,maxwait=20s
4414 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4415 to the end before writing, but this should make it even safer).
4417 22. Exim was forgetting that it had advertised PIPELINING for the second and
4418 subsequent messages on an SMTP connection. It was also not resetting its
4419 memory on STARTTLS and an internal HELO.
4421 23. When Exim logs an SMTP synchronization error within a session, it now
4422 records whether PIPELINING has been advertised or not.
4424 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4425 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4426 rather than long int.
4428 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4430 26. Added the never_mail option to autoreply.
4436 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4437 crash if the getsockname() call failed; this can happen if a connection is
4438 closed very soon after it is established. The problem was simply in the
4439 order in which certain operations were done, causing Exim to try to write
4440 to the SMTP stream before it had set up the file descriptor. The bug has
4441 been fixed by making things happen in the correct order.
4447 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4448 before the connection was closed, thus losing the rejection response.
4450 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4451 some early Solaris releases, but causes trouble in current releases where
4452 socklen_t is defined.
4454 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4457 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4460 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4461 that releasing the top of it at the end releases what was used for sub-
4462 expansions (unless the block got too big). However, discard this block if
4463 the first thing is a variable or header, so that we can use its block when
4464 it is dynamic (useful for very large $message_headers, for example).
4466 6. Lookups now cache *every* query, not just the most recent. A new, separate
4467 store pool is used for this. It can be recovered when all lookup caches are
4468 flushed. Lookups now release memory at the end of their result strings.
4469 This has involved some general refactoring of the lookup sources.
4471 7. Some code has been added to the store_xxx() functions to reduce the amount
4472 of flapping under certain conditions.
4474 8. log_incoming_interface used to affect only the <= reception log lines. Now
4475 it causes the local interface and port to be added to several more SMTP log
4476 lines, for example "SMTP connection from", and rejection lines.
4478 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4480 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4482 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4483 could be overwritten at the end of the current message (or the start of a
4484 new message if it was set in a HELO ACL). The value is now preserved for
4485 the duration of the SMTP connection.
4487 12. If a transport had a headers_rewrite setting, and a matching header line
4488 contained an unqualified address, that address was qualified, even if it
4489 did not match any rewriting rules. The underlying bug was that the values
4490 of the flags that permit the existence of unqualified sender and recipient
4491 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4492 non-local messages, and by -bnq for local messages) were not being
4493 preserved with the message after it was received.
4495 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4496 "next input=" as part of the text comprising the host identity instead of
4497 the correct text. The code was using the same buffer for two different
4498 strings. However, depending on which order the printing function evaluated
4499 its arguments, the bug did not always show up. Under Linux, for example, my
4500 test suite worked just fine.
4502 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4503 doesn't work with some older versions of Perl. It has been changed to "my",
4504 which in any case is probably the better facility to use.
4506 15. A really picky compiler found some instances of statements for creating
4507 error messages that either had too many or two few arguments for the format
4510 16. The size of the buffer for calls to the DNS resolver has been increased
4511 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4512 for addresses that have a lot of PTR records. This alleviates a problem; it
4513 does not fully solve it.
4515 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4516 buffer now truncates the list and logs the incident, which is the same
4517 action as happens when Exim is looking up a host name and its aliases.
4518 Previously in this situation something unpredictable would happen;
4519 sometimes it was "internal error: store_reset failed".
4521 18. If a server dropped the connection unexpectedly when an Exim client was
4522 using GnuTLS and trying to read a response, the client delivery process
4523 crashed while trying to generate an error log message.
4525 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4526 string, for example:
4528 warn message = H1: something\nH2: something
4530 the text was added as a single header line from Exim's point of view
4531 though it ended up OK in the delivered message. However, searching for the
4532 second and subsequent header lines using $h_h2: did not work. This has been
4533 fixed. Similarly, if a system filter added multiple headers in this way,
4534 the routers could not see them.
4536 20. Expanded the error message when iplsearch is called with an invalid key to
4537 suggest using net-iplsearch in a host list.
4539 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4540 ACLs are no longer actually imposed (and a message to that effect is
4543 22. If a "gecos" field in a passwd entry contained escaped characters, in
4544 particular, if it contained a \" sequence, Exim got it wrong when building
4545 a From: or a Sender: header from that name. A second bug also caused
4546 incorrect handling when an unquoted " was present following a character
4547 that needed quoting.
4549 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4550 was not being matched caselessly.
4552 24. Arranged for all hyphens in the exim.8 source to be escaped with
4555 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4556 properly. Recipient callout cache records were still being keyed to include
4557 the sender, even when use_sender was set false. This led to far more
4558 callouts that were necessary. The sender is no longer included in the key
4559 when use_sender is false.
4561 26. Added "control = submission" modifier to ACLs.
4563 27. Added the ${base62d: operator to decode base 62 numbers.
4565 28. dnsdb lookups can now access SRV records.
4567 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4568 the configuration file.
4570 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4571 (-v) mode. This makes the output for a verbose queue run more intelligible.
4573 31. Added a use_postmaster feature to recipient callouts.
4575 32. Added the $body_zerocount variable, containing the number of binary zero
4576 bytes in the message body.
4578 33. The time of last modification of the "new" subdirectory is now used as the
4579 "mailbox time last read" when there is a quota error for a maildir
4582 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4584 35. Added +ignore_unknown as a special item in host lists.
4586 36. Code for decoding IPv6 addresses in host lists is now included, even if
4587 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4588 address was recognized as an IP address, but was then not correctly decoded
4589 into binary, causing unexpected and incorrect effects when compared with
4596 1. Very minor rewording of debugging text in manualroute to say "list of
4597 hosts" instead of "hostlist".
4599 2. If verify=header_syntax was set, and a header line with an unqualified
4600 address (no domain) and a large number of spaces between the end of the
4601 name and the colon was received, the reception process suffered a buffer
4602 overflow, and (when I tested it) crashed. This was caused by some obsolete
4603 code that should have been removed. The fix is to remove it!
4605 3. When running in the test harness, delay a bit after writing a bounce
4606 message to get a bit more predictability in the log output.
4608 4. Added a call to search_tidyup() just before forking a reception process. In
4609 theory, someone could use a lookup in the expansion of smtp_accept_max_
4610 per_host which, without the tidyup, could leave open a database connection.
4612 5. Added the variables $recipient_data and $sender_data which get set from a
4613 lookup success in an ACL "recipients" or "senders" condition, or a router
4614 "senders" option, similar to $domain_data and $local_part_data.
4616 6. Moved the writing of debug_print from before to after the "senders" test
4619 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4620 problems for message scanning, either using a data ACL, or using
4621 local_scan() because the Received: header was not generated till after they
4622 were called (in order to set the time as the time of reception completion).
4623 I have revised the way this works. The header is now generated after the
4624 body is received, but before the ACL or local_scan() are called. After they
4625 are run, the timestamp in the header is updated.
4631 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4632 before starting a queue runner without re-exec. This happened only when
4633 deliver_drop_privilege was set or when the Exim user was set to root. The
4634 effect of the bug was that timeouts during subsequent deliveries caused
4635 crashes instead of being properly handled. The handler is now left at its
4636 default (and expected) setting.
4638 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4639 message, again when deliver_drop_privilege is set or Exim is run as root.
4640 The bug described in (1) was not present in this case, but the tidying up
4641 of the other signals was missing. I have made the two cases consistent.
4643 3. The ignore_target_hosts setting on a manualroute router was being ignored
4644 for hosts that were looked up using the /MX notation.
4646 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4649 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4650 operated on the sender address. After changing the $sender_address to <>
4651 for the sender address verify, Exim was re-instated it as the original
4652 (before rewriting) address, but remembering that it had rewritten it, so it
4653 wasn't rewriting it again. This bug also had the effect of breaking the
4654 sender address verification caching when the sender address was rewritten.
4656 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4657 This has been changed so that if the ip literal address matches
4658 ignore_target_hosts, the router declines.
4660 7. Added expansion conditions match_domain, match_address, and match_local_
4661 part (NOT match_host).
4663 8. The placeholder for the Received: header didn't have a length field set.
4665 9. Added code to Exim itself and to exim_lock to test for a specific race
4666 condition that could lead to file corruption when using MBX delivery. The
4667 issue is with the lockfile that is created in /tmp. If this file is removed
4668 after a process has opened it but before that process has acquired a lock,
4669 there is the potential for a second process to recreate the file and also
4670 acquire a lock. This could lead to two Exim processes writing to the file
4671 at the same time. The added code performs the same test as UW imapd; it
4672 checks after acquiring the lock that its file descriptor still refers to
4673 the same named file.
4675 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4676 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4679 11. Added the smtp_active_hostname option. If used, this will typically be made
4680 to depend on the incoming interface address. Because $interface_address is
4681 not set up until the daemon has forked a reception process, error responses
4682 that can happen earlier (such as "too many connections") no longer contain
4685 12. If an expansion in a condition on a "warn" statement fails because a lookup
4686 defers, the "warn" statement is abandoned, and the next ACL statement is
4687 processed. Previously this caused the whole ACL to be aborted.
4689 13. Added the iplsearch lookup type.
4691 14. Added ident_timeout as a log selector.
4693 15. Added tls_certificate_verified as a log selector.
4695 16. Added a global option tls_require_ciphers (compare the smtp transport
4696 option of the same name). This controls incoming TLS connections.
4698 17. I finally figured out how to make tls_require_ciphers do a similar thing
4699 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4700 before starting the TLS session.
4702 18. Tabs are now shown as \t in -bP output.
4704 19. If the log selector return_path_on_delivery was set, Exim crashed when
4705 bouncing a message because it had too many Received: header lines.
4707 20. If two routers both had headers_remove settings, and the first one included
4708 a superfluous trailing colon, the final name in the first list and the
4709 first name in the second list were incorrectly joined into one item (with a
4710 colon in the middle).
4716 1. Added -C and -D options to the exinext utility, mainly to make it easier
4717 to include in the automated testing, but these could be helpful when
4718 multiple configurations are in use.
4720 2. The exinext utility was not formatting the output nicely when there was
4721 an alternate port involved in the retry record key, nor when there was a
4722 message id as well (for retries that were specific to a specific message
4723 and a specific host). It was also confused by IPv6 addresses, because of
4724 the additional colons they contain. I have fixed the IPv4 problem, and
4725 patched it up to do a reasonable job for IPv6.
4727 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4728 delivery, the log line now contains "pipelined" if PIPELINING was used.
4730 4. An SMTP transport process used to panic and die if the bind() call to set
4731 an explicit outgoing interface failed. This has been changed; it is now
4732 treated in the same way as a connect() failure.
4734 5. A reference to $sender_host_name in the part of a conditional expansion
4735 that was being skipped was still causing a DNS lookup. This no longer
4738 6. The def: expansion condition was not recognizing references to header lines
4739 that used bh_ and bheader_.
4741 7. Added the _cache feature to named lists.
4743 8. The code for checking quota_filecount in the appendfile transport was
4744 allowing one more file than it should have been.
4746 9. For compatibility with Sendmail, the command line option
4754 and sets the incoming protocol and host name (for trusted callers). The
4755 host name and its colon can be omitted when only the protocol is to be set.
4756 Note the Exim already has two private options, -pd and -ps, that refer to
4757 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4758 "s", but I don't think that's a major issue.
4760 10. A number of refactoring changes to the code, none of which should affect
4763 (a) The number of logging options was getting close to filling up the
4764 32-bit word that was used as a bit map. I have split them into two classes:
4765 those that are passed in the argument to log_write(), and those that are
4766 only ever tested independently outside of that function. These are now in
4767 separate 32-bit words, so there is plenty of room for expansion again.
4768 There is no change in the user interface or the logging behaviour.
4770 (b) When building, for example, log lines, the code previously used a
4771 macro that called string_cat() twice, in order to add two strings. This is
4772 not really sufficiently general. Furthermore, there was one instance where
4773 it was actually wrong because one of the argument was used twice, and in
4774 one call a function was used. (As it happened, calling the function twice
4775 did not affect the overall behaviour.) The macro has been replaced by a
4776 function that can join an arbitrary number of extra strings onto a growing
4779 (c) The code for expansion conditions now uses a table and a binary chop
4780 instead of a serial search (which was left over from when there were very
4781 few conditions). Also, it now recognizes conditions like "pam" even when
4782 the relevant support is not compiled in: a suitably worded error message is
4783 given if an attempt is made to use such a condition.
4785 11. Added ${time_interval:xxxxx}.
4787 12. A bug was causing one of the ddress fields not to be passed back correctly
4788 from remote delivery subprocesses. The field in question was not being
4789 subsequently used, so this caused to problems in practice.
4791 13. Added new log selectors queue_time and deliver_time.
4793 14. Might have fixed a bug in maildirsizefile handling that threw up
4794 "unexpected character" debug warnings, and recalculated the data
4795 unnecessarily. In any case, I expanded the warning message to give more
4798 15. Added the message "Restricted characters in address" to the statements in
4799 the default ACL that block characters like @ and % in local parts.
4801 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4802 Three changes have been made:
4804 (a) There was a serious bug; a negative response to MAIL caused the whole
4805 recipient domain to be cached as invalid, thereby blocking all messages
4806 to all local parts at the same domain, from all senders. This bug has
4807 been fixed. The domain is no longer cached after a negative response to
4808 MAIL if the sender used is not empty.
4810 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4813 (c) A new callout option, "use_sender" has been added for people who want
4814 the modified behaviour.
4820 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4823 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4824 indeed breaks things for older releases.
4826 3. Added additional logging to the case where there is a problem reading data
4827 from a filter that is running in a subprocess using a pipe, in order to
4828 try to track down a specific problem.
4830 4. Testing facility fudge: when running in the test harness and attempting
4831 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4832 getting "No route to host". Convert this to a timeout.
4834 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4837 6. Some OS don't have socklen_t but use size_t instead. This affects the
4838 fifth argument of getsockopt() amongst other things. This is now
4839 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4840 can be set for individual OS. I have set it for SunOS5, OSF1, and
4841 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4842 some earlier ones do not.
4844 7. Change 4.30/15 was not doing the test caselessly.
4846 8. The standard form for an IPv6 address literal was being rejected by address
4847 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4848 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4849 this, as well as the form without the "IPv6" on the front (but only when
4850 address literals are enabled, of course).
4852 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4854 10. Exim crashed if a message with an empty sender address specified by -f
4855 encountered a router with an errors_to setting. This could be provoked only
4856 by a command such as
4860 where an empty string was supplied; "<>" did not hit this bug.
4862 11. Installed PCRE release 4.5.
4864 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4865 remained set. It is now erased.
4867 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4868 times from message ids (which are base 36 rather than the normal 62).
4870 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4871 were being counted as actual protocol errors, and logged if the log
4872 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4873 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4874 and DATA following a set of rejected RCPTs do not count as protocol errors.
4875 In other words, Exim assumes they were pipelined, though this may not
4876 actually be the case. Of course, in all cases the client gets an
4877 appropriate error code.
4879 15. If a lookup fails in an ACL condition, a message about the failure may
4880 be available; it is used if testing the ACL cannot continue, because most
4881 such messages specify what the cause of the deferral is. However, some
4882 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4883 that caused an old message to be retained and used if a later statement
4884 caused a defer, replacing the real cause of the deferral.
4886 16. If an IP address had so many PTR records that the DNS lookup buffer
4887 was not large enough to hold them, Exim could crash while trying to process
4888 the truncated data. It now detects and logs this case.
4890 17. Further to 4.21/58, another change has been made: if (and only if) the
4891 first line of a message (the first header line) ends with CRLF, a bare LF
4892 in a subsequent header line has a space inserted after it, so as not to
4893 terminate the header.
4895 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4896 unnecessarily, used atoi() instead of strtol(), and didn't check the
4897 termination when getting file sizes from file names by regex.
4899 19. Completely re-implemented the support for maildirsize files, in the light
4900 of a number of problems with the previous contributed implementation
4901 (4.30/29). In particular:
4903 . If the quota is zero, the maildirsize file is maintained, but no quota is
4906 . If the maildir directory does not exist, it is created before any attempt
4907 to write a maildirsize file.
4909 . The quota value in the file is just a cache; if the quota is changed in
4910 the transport, the new value overrides.
4912 . A regular expression is available for excluding directories from the
4915 20. The autoreply transport checks the characters in options that define the
4916 message's headers; it allows continued headers, but it was checking with
4917 isspace() after an embedded newline instead of explicitly looking for a
4920 21. If all the "regular" hosts to which an address was routed had passed their
4921 expiry times, and had not reached their retry times, the address was
4922 bounced, even if fallback hosts were defined. Now Exim should go on to try
4925 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4926 equivalent code in the SMTP transport. Some hosts send humungous responses
4927 to HELO/EHLO, more than 1024 it seems.
4929 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4930 gives compiler warnings in some environments. I've now done it "properly",
4933 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4934 (because of problems with the built-in one) was declared to return uschar *
4935 instead of char *, causing compiler failure.
4937 25. Fixed a file descriptor leak when processing alias/forward files.
4939 26. Fixed a minor format string issue in dbfn.c.
4941 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4943 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4944 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4947 29. When a sender address is verified, it is cached, to save repeating the test
4948 when there is more than one recipient in a message. However, when the
4949 verification involves a callout, it is possible for different callout
4950 options to be set for different recipients. It is too complicated to keep
4951 track of this in the cache, so now Exim always runs a verification when a
4952 callout is required, relying on the callout cache for the optimization.
4953 The overhead is duplication of the address routing, but this should not be
4956 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4957 to be verified with callout=postmaster, and the main callout worked but the
4958 postmaster check failed, the verification correctly failed. However, if a
4959 subsequent RCPT command asked for sender verification *without* the
4960 postmaster check, incorrect caching caused this verification also to fail,
4963 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4964 it was not caching the DNS options (qualify_single, search_parents) that
4965 were used when the lookup failed. A subsequent lookup with different
4966 options therefore always gave the same answer, though there were cases
4967 where it should not have. (Example: a "domains = !$mx_any" option on a
4968 dnslookup router: the "domains" option is always processed without any
4969 widening, but the router might have qualify_single set.) Now Exim uses the
4970 cached value only when the same options are set.
4972 32. Added John Jetmore's "exipick" utility to the distribution.
4974 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4975 than a timeout (e.g. a certificate is required, and is not provided), an
4976 Exim server now closes the connection immediately. Previously it waited for
4977 the client to close - but if the client is SSL, it seems that they each
4978 wait for each other, leading to a delay before one of them times out.
4980 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4981 maintained 0.8.x compatibility because I don't think many are using it, and
4982 it is clearly obsolete.
4984 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4987 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4988 client certificate was expired. A simple patch fixes this, though I don't
4989 understand the full logic of why the verify callback is called multiple
4992 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4993 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4994 which causes problems with some clients (such as the Certicom SSL Plus
4995 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4996 disables the coutermeasure allowing Eudora to connect."
4998 38. Exim was not checking that a write() to a log file succeeded. This could
4999 lead to Bad Things if a log got too big, in particular if it hit a file
5000 size limit. Exim now panics and dies if it cannot write to a log file, just
5001 as it does if it cannot open a log file.
5003 39. Modified OS/Makefile-Linux so that it now contains
5005 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5007 The two -D definitions ensure that Exim is compiled with large file
5008 support, which makes it possible to handle log files that are bigger than
5011 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5012 instance) a domain was checked against a named list that involved a lookup,
5013 causing $domain_data to be set, then another domain was checked against the
5014 same list, then the first domain was re-checked, the value of $domain_data
5015 after the final check could be wrong. In particular, if the second check
5016 failed, it could be set empty. This bug probably also applied to
5019 41. The strip_trailing_dot option was not being applied to the address given
5020 with the -f command-line option.
5022 42. The code for reading a message's header from the spool was incrementing
5023 $received_count, but never initializing it. This meant that the value was
5024 incorrect (doubled) while delivering a message in the same process in which
5025 it was received. In the most common configuration of Exim, this never
5026 happens - a fresh exec is done - but it can happen when
5027 deliver_drop_privilege is set.
5029 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5030 it now includes up to 150 characters of the unexpected data in the log
5033 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5034 and building data strings. The size of both of these buffers was 10 000
5035 bytes - far larger than anybody would *ever* want, thought I. Needless to
5036 say, somebody hit the limit. I have increased the maximum line length to
5037 20 000 and the maximum data length of concatenated lines to 100 000. I have
5038 also fixed two bugs, because there was no checking on these buffers. Tsk,
5039 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5040 buffer is too small.
5042 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5043 lsearch lookups. Now it does.
5045 46. When parsing a route_list item in a manualroute router, a fixed-length
5046 buffer was used for the list of hosts. I made this 1024 bytes long,
5047 thinking that nobody would ever have a list of hosts that long. Wrong.
5048 Somebody had a whole pile of complicated expansion conditions, and the
5049 string was silently truncated, leading to an expansion error. It turns out
5050 that it is easier to change to an unlimited length (owing to other changes
5051 that have happened since this code was originally written) than to build
5052 structure for giving a limitation error. The length of the item that
5053 expands into the list of hosts is now unlimited.
5055 47. The lsearch lookup could not handle data where the length of text line was
5056 more than 4095 characters. Such lines were truncated, leading to shortened
5057 data being returned. It should now handle lines of any length.
5059 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5060 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5063 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5064 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5065 no longer happens in comment lines. A list of replacements is now placed
5066 at the head of all of the source files, except those whose only change is
5067 to replace PERL_COMMAND in the very first #! line.
5069 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5070 messages on the queue, with a bottom-up merge sort, using code contributed
5071 by Michael Haardt. This should make operations like -bp somewhat faster on
5072 large queues. It won't affect queue runners, except when queue_run_in_order
5075 51. Installed eximstats 1.31 in the distribution.
5077 52. Added support for SRV lookups to the dnslookup router.
5079 53. If an ACL referred to $message_body or $message_body_end, the value was not
5080 reset for any messages that followed in the same SMTP session.
5082 54. The store-handling optimization for building very long strings was not
5083 differentiating between the different store pools. I don't think this
5084 actually made any difference in practice, but I've tidied it.
5086 55. While running the routers to verify a sender address, $sender_address
5087 was still set to the sender address. This is wrong, because when routing to
5088 send a bounce to the sender, it would be empty. Therefore, I have changed
5089 it so that, while verifying a sender address, $sender_address is set to <>.
5090 (There is no change to what happens when verifying a recipient address.)
5092 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5093 target A or AAAA records (if not already returned) without resetting the
5094 qualify_single or search_parents options of the DNS resolver. These are
5095 inappropriate in this case because the targets of MX and SRV records must
5096 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5097 target that, when qualified, matched something in the local domain. These
5098 two options are now turned off when doing these lookups.
5100 57. It seems that at least some releases of Reiserfs (which does not have the
5101 concept of a fixed number of inodes) returns zero and not -1 for the
5102 number of available inodes. This interacted badly with check_spool_inodes,
5103 which assumed that -1 was the "no such thing" setting. What I have done is
5104 to check that the total number of inodes is greater than zero before doing
5105 the test of how many are available.
5107 58. When a "warn" ACL statement has a log_message modifier, the message is
5108 remembered, and not repeated. This is to avoid a lot of repetition when a
5109 message has many recipients that cause the same warning to be written.
5110 Howewer, Exim was preserving the list of already written lines for an
5111 entire SMTP session, which doesn't seem right. The memory is now reset if a
5112 new message is started.
5114 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5115 debugging output unless log_rewrite was also set.
5117 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5118 of (int)(handle) when we know that handle contains (void *)(-1).
5120 61. The Exim daemon panic-logs an error return when it closes the incoming
5121 connection. However "connection reset by peer" seems to be common, and
5122 isn't really an error worthy of noting specially, so that particular error
5125 62. When Exim is trying to find all the local interfaces, it used to panic and
5126 die if the ioctl to get the interface flags failed. However, it seems that
5127 on at least one OS (Solaris 9) it is possible to have an interface that is
5128 included in the list of interfaces, but for which you get a failure error
5129 for this call. This happens when the interface is not "plumbed" into a
5130 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5131 failure of the "get flags" call assumes that the interface is down.
5133 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5134 makes life easier for people who are doing arithmetic on fields extracted
5135 from dates, where you often get leading zeros that should not be
5136 interpreted as octal.
5138 64. Added qualify_domain to the redirect router, to override the global
5141 65. If a pathologically long header line contained very many addresses (the
5142 report of this problem mentioned 10 000) and each of them was rewritten,
5143 Exim could use up a very large amount of memory. (It kept on making new
5144 copies of the header line as it rewrote, and never released the old ones.)
5145 At the expense of a bit more processing, the header rewriting function has
5146 been changed so that it no longer eats memory in this way.
5148 66. The generation of the Received: header has been moved from the time that a
5149 message starts to be received, to the time that it finishes. The timestamp
5150 in the Received: header should now be very close to that of the <= log
5151 line. There are two side-effects of this change:
5153 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5154 logged header lines no longer include the local Received: line, because
5155 it has not yet been created. The same applies to a copy of the message
5156 that is returned to a non-SMTP sender when a message is rejected.
5158 (b) When a filter file is tested using -bf, no additional Received: header
5159 is added to the test message. After some thought, I decided that this
5162 This change does not affect the value of $received_for. It is still set
5163 after address rewriting, but before local_scan() is called.
5165 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5167 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5168 gave an unhelpful panic error message, and a defer error. I have managed to
5169 change this behaviour so that it now rejects any supplied certificate,
5170 which seems right, as the list of acceptable certificates is empty.
5172 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5173 gave an unhelpful defer error. I have not managed to make this reject any
5174 supplied certificates, but the error message it gives is "no certificate
5175 supplied", which is not helpful.
5177 70. exigrep's output now also includes lines that are not associated with any
5178 message, but which match the given pattern. Implemented by a patch from
5179 Martin Sluka, which also tidied up the Perl a bit.
5181 71. Recipient callout verification, like sender verification, was using <> in
5182 the MAIL FROM command. This isn't really the right thing, since the actual
5183 sender may affect whether the remote host accepts the recipient or not. I
5184 have changed it to use the actual sender in the callout; this means that
5185 the cache record is now keyed on a recipient/sender pair, not just the
5186 recipient address. There doesn't seem to be a real danger of callout loops,
5187 since a callout by the remote host to check the sender would use <>.
5188 [SEE ABOVE: changed after hitting problems.]
5190 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5191 temporary errors. However, in the case of such a code being given after
5192 the end of a data transmission (i.e. after ".") Exim was failing to write
5193 a retry record for the message. (Yes, there was some broken host that was
5194 actually sending 8xx at this point.)
5196 73. An unknown lookup type in a host list could cause Exim to panic-die when
5197 the list was checked. (An example that provoked this was putting <; in the
5198 middle of a list instead of at the start.) If this happened during a DATA
5199 ACL check, a -D file could be left lying around. This kind of configuration
5200 error no longer causes Exim to die; instead it causes a defer errror. The
5201 incident is still logged to the main and panic logs.
5203 74. Buglet left over from Exim 3 conversion. The message "too many messages
5204 in one connection" was written to the rejectlog but not the mainlog, except
5205 when address rewriting (yes!) was being logged.
5207 75. Added write_rejectlog option.
5209 76. When a system filter was run not as root (that is, when system_filter_user
5210 was set), the values of the $n variables were not being returned to the
5211 main process; thus, they were not subsequently available in the $sn
5214 77. Added +return_path_on_delivery log selector.
5216 78. A connection timeout was being treated differently from recipients deferred
5217 when testing hosts_max_try with a message that was older than the host's
5218 retry timeout. (The host should not be counted, thus allowing all hosts to
5219 be tried at least once before bouncing.) This may have been the cause of an
5220 occasionally reported bug whereby a message would remain on the queue
5221 longer than the retry timeout, but would be bounced if a delivery was
5222 forced. I say "may" because I never totally pinned down the problem;
5223 setting up timeout/retry tests is difficult. See also the next item.
5225 79. The ultimate address timeout was not being applied to errors that involved
5226 a combination of host plus message (for example, a timeout on a MAIL
5227 command). When an address resolved to a number of possible hosts, and they
5228 were not all tried for each delivery (e.g. because of hosts_max_try), a
5229 message could remain on the queue longer than the retry timeout.
5231 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5234 81. Fixed an obscure SMTP outgoing bug which required at least the following
5235 conditions: (a) there was another message waiting for the same server;
5236 (b) the server returned 5xx to all RCPT commands in the first message so
5237 that the message was not completed; (c) the server dropped the connection
5238 or gave a negative response to the RSET that Exim sends to abort the
5239 transaction. The observed case was a dropped connection after DATA that had
5240 been sent in pipelining mode. That is, the server had advertised PIPELINING
5241 but was not implementing it correctly. The effect of the bug was incorrect
5242 behaviour, such as trying another host, and this could lead to a crash.
5248 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5249 and daemon.c were passed as pointers to ints; they should have been
5250 pointers to socklen_t variables (which are typically unsigned ints).
5252 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5255 3. Fixed a really odd bug that affected only the testing scheme; patching a
5256 certain fixed string in the binary changed the value of another string that
5257 happened to be identical to the end of the original first string.
5259 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5260 name", it returns that address as the IP address. On some operating
5261 systems (e.g. Solaris), it also passes back the IP address string as the
5262 "host name". However, on others (e.g. Linux), it passes back an empty
5263 string. Exim wasn't checking for this, and was changing the host name to an
5264 empty string, assuming it had been canonicized.
5266 5. Although rare, it is permitted to have more than one PTR record for a given
5267 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5268 all the names associated with an address, because they do in Solaris.
5269 However, it seems that they do not in Linux for data that comes from the
5270 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5271 I found this out when I moved to a new Linux workstation and tried to run
5272 the Exim test suite.
5274 To get round this problem I have changed the code so that it now does its
5275 own call to the DNS to look up PTR records when searching for a host name.
5276 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5277 addresses that are only in /etc/hosts are still found.
5279 This behaviour is, however, controlled by an option called host_lookup_
5280 order, which defaults to "bydns:byaddr". If people want to use the other
5281 order, or indeed, just use one or the other means of lookup, they can
5282 specify it in this variable.
5284 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5285 some operating systems, this comes back from gethostbyaddr() as an empty
5286 string, and this is what Exim used to test for. However, it seems that in
5287 other systems, "." is yielded. Exim now tests for this case too.
5289 7. The values of check_spool_space and check_log_space are now held internally
5290 as a number of kilobytes instead of an absolute number of bytes. If a
5291 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5292 kilobyte. This means that much larger values can be stored.
5294 8. Exim monitor: an attempt to get the action menu when not actually pointing
5295 at a message produces an empty menu entitled "No message selected". This
5296 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5297 no entries in it ("Shell widget menu has zero width and/or height"). So I
5298 have added a single, blank menu entry in this case.
5300 9. Added ${quote_local_part.
5302 10. MIME decoding is now applied to the contents of Subject: header lines when
5305 11. Now that a reference to $sender_host_address automatically causes a reverse
5306 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5307 host lookup before query-style lookups in lists that might use this
5308 variable. This has therefore been abolished, and the "net-" prefix is no
5309 longer necessary for query-style lookups.
5311 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5312 have been a typo for LFLAGS, so it has been changed.
5314 13. The install script calls Exim with "-C /dev/null" in order to find the
5315 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5316 to be output. Howeve, since Exim outputs its version number before the
5317 error, it didn't break the script. It just looked ugly. I fixed this by
5318 always allowing "-C /dev/null" if the caller is root.
5320 14. Ignore overlarge ACL variable number when reading spool file - insurance
5321 against a later release with more variables having written the file.
5323 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5324 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5325 this, as well as the form without the "IPv6" on the front.
5327 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5328 OS/Makefile-Darwin file.
5330 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5331 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5333 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5334 variable. However, some OS use ints for inodes. I've added cast to long int
5335 to get rid of the compiler warning.
5337 19. I had forgotten to lock out "/../" in configuration file names when
5338 ALT_CONFIG_PREFIX was set.
5340 20. Routers used for verification do not need to specify transports. However,
5341 if such a router generated a host list, and callout was configured, Exim
5342 crashed, because it could not find a port number from the (non-existent)
5343 transport. It now assumes port 25 in this circumstance.
5345 21. Added the -t option to exigrep.
5347 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5348 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5349 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5350 them is set, LOOKUP_LSEARCH is forced.
5352 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5354 24. Added sender and host information to the "rejected by local_scan()" log
5355 line; previously there was no indication of these.
5357 25. Added .include_if_exists.
5359 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5360 Linux. It turns out that not all file systems support this. Apparently some
5361 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5362 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5363 not supported on the file descriptor, is now ignored when Exim is trying to
5364 sync a directory. This applies only to Linux.
5366 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5368 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5369 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5372 29. Added support for maildirsize files from supplied patch (modified a bit).
5374 30. The use of :fail: followed by an empty string could lead Exim to respond to
5375 sender verification failures with (e.g.):
5377 550 Verification failed for <xxx>
5378 550 Sender verify failed
5380 where the first response line was missing the '-' that indicates it is not
5381 the final line of the response.
5383 31. The loop for finding the name of the user that called Exim had a hardwired
5384 limit of 10; it now uses the value of finduser_retries, which is used for
5385 all other user lookups.
5387 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5390 33. Exim was neglecting to zero errno before one call of strtol() when
5391 expanding a string and expecting an integer value. On some systems this
5392 resulted in spurious "integer overflow" errors. Also, it was casting the
5393 result into an int without checking.
5395 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5396 did not work. The code looks as if it has *never* worked, though it appears
5397 to have been documented since at least releast 1.62. I have made it work.
5399 35. The "timeout_DNS" error in retry rules, also documented since at least
5400 1.62, also never worked. As it isn't clear exactly what this means, and
5401 clearly it isn't a major issue, I have abolished the feature by treating it
5402 as "timeout", and writing a warning to the main and panic logs.
5404 36. The display of retry rules for -brt wasn't always showing the error code
5407 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5408 timeout_connect_A, timeout_connect_MX.
5410 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5411 to the empty sender.
5413 39. The daemon was not analysing the content of -oX till after it had closed
5414 stderr and disconnected from the controlling terminal. This meant that any
5415 syntax errors were only noted on the panic log, and the return code from
5416 the command was 0. By re-arranging the code a little, I've made the
5417 decoding happen first, so such errors now appear on stderr, and the return
5418 code is 1. However, the actual setting up of the sockets still happens in
5419 the disconnected process, so errors there are still only recorded on the
5422 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5423 connections (as happens on some IP stacks) was logged at start up time as
5424 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5425 it from "IPv6 and IPv4", which means that two separate sockets are being
5428 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5429 says whether AF_INET or AF_INET6 was passed as an argument.
5431 42. Exiwhat output was messed up when time zones were included in log
5434 43. Exiwhat now gives more information about the daemon's listening ports,
5435 and whether -tls-on-connect was used.
5437 44. The "port" option of the smtp transport is now expanded.
5439 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5440 silently ignored. Now an error message is written to the main and panic
5443 46. There's a new ACL modifier called "logwrite" which writes to a log file
5444 as soon as it is encountered.
5446 47. Added $local_user_uid and $local_user_gid at routing time.
5448 48. Exim crashed when trying to verify a sender address that was being
5451 49. Exim was recognizing only a space character after ".include". It now also
5452 recognizes a tab character.
5454 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5455 extracting the relevant information from the specification. The man page no
5456 longer contains scrambled data for the -d option, and I've added a section
5457 at the front about calling Exim under different names.
5459 51. Added "extra_headers" argument to the "mail" command in filter files.
5461 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5464 53. Installed eximstats 1.29.
5466 54. Added transport_filter_timeout as a generic transport option.
5468 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5469 Cc: header lines. This was required by RFC 822, but it not required by RFC
5472 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5473 incoming messages that did not have them. Now it does so only if the
5474 message originates locally, that is, if there is no associated remote host
5475 address. When Resent- header lines are present, this applies to the Resent-
5476 lines rather than the non-Resent- lines.
5478 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5479 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5481 58. Messages for configuration errors now include the name of the main
5482 configuration file - useful now that there may be more than one file in a
5483 list (.included file names were always shown).
5485 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5486 for those rare installations that do not start the daemon as root or run it
5487 setuid root. I've cut out the call to initgroups() if the daemon is not
5490 60. The Exim user and group can now be bound into the binary as text strings
5491 that are looked up at the start of Exim's processing.
5493 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5495 62. Added $mailstore_basename variable.
5497 63. Installed patch to sieve.c from Michael Haardt.
5499 64. When Exim failed to open the panic log after failing to open the main log,
5500 the original message it was trying to log was written to stderr and debug
5501 output, but if they were not available (the usual case in production), it
5502 was lost. Now it is written to syslog before the two lines that record the
5503 failures to open the logs.
5505 65. Users' Exim filters run in subprocesses under the user's uid. It is
5506 possible for a "deliver" command or an alias in a "personal" command to
5507 provoke an address rewrite. If logging of address rewriting is configured,
5508 this fails because the process is not running as root or exim. There may be
5509 a better way of dealing with this, but for the moment (because 4.30 needs
5510 to be released), I have disabled address rewrite logging when running a
5511 filter in a non-root, non-exim process.
5517 1. The buildconfig auxiliary program wasn't quoting the value set for
5518 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5519 not defined. This bug was masked in 4.22 by the effect that was fixed in
5522 2. Some messages that were rejected after a message id was allocated were
5523 shown as "incomplete" by exigrep. It no longer does this for messages that
5524 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5526 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5527 have allow_domain_literals set, the ID did not get logged in the <= line.
5528 Domain literals are now always recognized in Message-ID: header lines.
5530 4. The first argument for a ${extract expansion item is the key name or field
5531 number. Leading and trailing spaces in this item were not being ignored,
5532 causing some misleading effects.
5534 5. When deliver_drop_privilege was set, single queue runner processes started
5535 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5536 same command in the process it spins off) were not dropping privilege.
5538 6. When the daemon running as "exim" started a queue runner, it always
5539 re-executed Exim in the spun-off process. This is a waste of effort when
5540 deliver_drop_privilege is set. The new process now just calls the
5541 queue-runner function directly.
5547 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5550 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5551 ignored. Though the use of -oP was forcing the writing of a pid file, it
5552 was always written to the default place.
5554 3. If the message "no IP address found for host xxxx" is generated during
5555 incoming verification, it is now followed by identification of the incoming
5556 connection (so you can more easily find what provoked it).
5558 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5560 5. Added some features to "harden" Exim a bit more against certain attacks:
5562 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5563 be put in Local/Makefile. This is like the never_users runtime option,
5564 but it cannot be overridden. The default setting is "root".
5566 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5567 prefix string with which any file named in a -C command line option
5570 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5571 is retained for -C and -D only if the caller of Exim is root. Without
5572 it, the exim user may also use -C and -D and retain privilege.
5574 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5575 command line option is disabled.
5577 6. Macro names set by the -D option must start with an upper case letter, just
5578 like macro names defined in the configuration file.
5580 7. Added "dereference=" facility to LDAP.
5582 8. Two instances of the typo "uknown" in the source files are fixed.
5584 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5585 the Configure-Makefile script screwed up while processing it.
5587 10. Incorporated PCRE 4.4.
5589 11. The SMTP synchronization check was not operating right at the start of an
5590 SMTP session. For example, it could not catch a HELO sent before the client
5591 waited for the greeting. There is now a check for outstanding input at the
5592 point when the greeting is written. Because of the duplex, asynchronous
5593 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5594 way, but not yet received, when the check is performed.
5596 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5597 on TCP/IP sockets, because this apparently causes some broken clients to
5600 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5601 unchanged) from the Cygwin maintainer.
5603 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5604 was supported instead of testing for mbx. Effectively a typo.
5606 15. The spa authenticator server code was not checking that the input it
5607 received was valid base64.
5609 16. The debug output line for the "set" modifier in ACLs was not showing the
5610 name of the variable that was being set.
5612 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5614 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5615 lookup on its own. Something else was needed to trigger the lookup. For
5616 example, a match in host_lookup or the need for a host name in a host list.
5617 Now, if $sender_host_name is referenced and the host name has not yet been
5618 looked up, a lookup is performed. If the lookup fails, the variable remains
5619 empty, and $host_lookup_failed is set to "1".
5621 19. Added "eqi" as a case-independent comparison operator.
5623 20. The saslauthd authentication condition could segfault if neither service
5624 nor realm was specified.
5626 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5627 error message that was logged was misleading, and incoming SMTP
5628 connections were dropped. The message is now more accurate, and temporary
5629 errors are given to SMTP connections.
5631 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5632 (except RSET) with a 421 error, until QUIT is received. However, it was
5633 failing to send a response to QUIT.
5635 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5636 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5637 ACL is now run after the helo_try_verify_hosts code.
5639 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5640 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5641 case-independent, so other case variants are also recognized). Apparently
5642 some systems use these upper case variants.
5644 25. If more than two messages were waiting for the same host, and a transport
5645 filter was specified for the transport, Exim sent two messages over the
5646 same TCP/IP connection, and then failed with "socket operation on non-
5647 socket" when it tried to send the third.
5649 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5651 27. The extern definition of crypt16() in expand.c was not being excluded when
5652 the OS had its own crypt16() function.
5654 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5655 as a preferred synonym for return_size_limit, both as an option and as an
5658 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5660 30. Changed the default configuration ACL to relax the local part checking rule
5661 for addresses that are not in any local domains. For these addresses,
5662 slashes and pipe symbols are allowed within local parts, but the sequence
5663 /../ is explicitly forbidden.
5665 31. SPA server authentication was not clearing the challenge buffer before
5668 32. log_message in a "warn" ACL statement was writing to the reject log as
5669 well as to the main log, which contradicts the documentation and doesn't
5670 seem right (because no rejection is happening). So I have stopped it.
5672 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5673 However, I am unable to do any testing of this.
5675 34. Fixed an infelicity in the appendfile transport. When checking directories
5676 for a mailbox, to see if any needed to be created, it was accidentally
5677 using path names with one or more superfluous leading slashes; tracing
5678 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5680 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5681 contents are added to the log line that is written for every discarded
5682 recipient. (Previously a log_message setting was ignored.)
5684 36. The ${quote: operator now quotes the string if it is empty.
5686 37. The install script runs exim in order to find its version number. If for
5687 some reason other than non-existence or emptiness, which it checks, it
5688 could not run './exim', it was installing it with an empty version number,
5689 i.e. as "exim-". This error state is now caught, and the installation is
5692 38. An argument was missing from the function that creates an error message
5693 when Exim fails to connect to the socket for saslauthd authentication.
5694 This could cause Exim to crash, or give a corrupted message.
5696 39. Added isip, isip4, and isip6 to ${if conditions.
5698 40. The ACL variables $acl_xx are now saved with the message, and can be
5699 accessed later in routers, transports, and filters.
5701 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5702 strings in the file are not string-expanded.
5704 42. If a MAIL command specified a SIZE value that was too large to fit into an
5705 int variable, the check against message_size_limit failed. Such values are
5706 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5707 day this will have to be increased, but I don't think I want to be around
5708 when emails are that large.
5715 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5716 iconv() is not standard in FreeBSD.
5718 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5719 IPv6 enabled. The observed symptom was a segmentation fault on return from
5720 the function os_common_find_running_interfaces() in src/os.c.
5722 3. In the check_special_case() function in daemon.c I had used "errno" as an
5723 argument name, which causes warnings on some systems. This was basically a
5724 typo, since it was named "eno" in the comments!
5726 4. The code that waits for the clock to tick (at a resolution of some fraction
5727 of a second) so as to ensure message-id uniqueness was always waiting for
5728 at least one whole tick, when it could have waited for less. [This is
5729 almost certainly not relevant at current processor speeds, where it is
5730 unlikely to ever wait at all. But we try to future-proof.]
5732 5. The function that sleeps for a time interval that includes fractions of a
5733 second contained a race. It did not block SIGALRM between setting the
5734 timer, and suspending (a couple of lines later). If the interval was short
5735 and the sigsuspend() was delayed until after it had expired, the suspension
5736 never ended. On busy systems this could lead to processes getting stuck for
5739 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5740 process, before it forks any delivery processes. The open lookup caching
5741 mechanism meant that the open file or database connection was passed into
5742 the delivery process. The problem was that delivery processes always tidy
5743 up cached lookup data. This could cause a problem for the next delivery
5744 process started by the queue runner, because the external queue runner
5745 process does not know about the closure. So the next delivery process
5746 still has data in the lookup cache. In the case of a file lookup, there was
5747 no problem because closing a file descriptor in a subprocess doesn't affect
5748 the parent. However, if the lookup was caching a connection to a database,
5749 the connection was closed, and the second delivery process was likely to
5750 see errors such as "PGSQL: query failed: server closed the connection
5751 unexpectedly". The problem has been fixed by closing all cached lookups
5752 in a queue runner before running a delivery process.
5754 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5755 seem to have the "const" qualifier which it has on other OS. I've
5758 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5759 *of the same type* (client or server) with the same public name that an
5760 error should be diagnosed.
5762 9. When Exim looked up a host name for an IP address, but failed to find the
5763 original IP address when looking up the host name (a safety check), it
5764 output the message "<ip address> does not match any IP for NULL", which was
5765 confusing, to say the least. The bug was that the host name should have
5766 appeared instead of "NULL".
5768 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5769 user that is built into the binary, and the -C or -D options is used, root
5770 privilege is dropped before the configuration file is read. In addition,
5771 logging is switched to stderr instead of the normal log files. If the
5772 configuration then re-defines the Exim user, the unprivileged environment
5773 is probably not what is expected, so Exim logs a panic warning message (but
5776 However, if deliver_drop_privilege is set, the unprivileged state may well
5777 be exactly what is intended, so the warning has been cut out in that case,
5778 and Exim is allowed to try to write to its normal log files.
5784 1. smtp_return_error_details was not giving details for temporary sender
5785 or receiver verification errors.
5787 2. Diagnose a configuration error if two authenticators have the same public
5790 3. Exim used not to create the message log file for a message until the first
5791 delivery attempt. This could be confusing when incoming messages were held
5792 for policy or load reasons. The message log file is now created at the time
5793 the message is received, and an initial "Received" line is written to it.
5795 4. The automatically generated man page for command line options had a minor
5796 bug that caused no ill effects; however, a more serious problem was that
5797 the procedure for building the man page automatically didn't always
5798 operate. Consequently, release 4.20 contains an out-of-date version. This
5799 shouldn't happen again.
5801 5. When building Exim with embedded Perl support, the script that builds the
5802 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5803 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5805 6. The freeze_tell option was not being used for messages that were frozen on
5806 arrival, either by an ACL or by local_scan().
5808 7. Added the smtp_incomplete_transaction log selector.
5810 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5811 was accepting AUTH without a new EHLO.
5813 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5814 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5817 10. Logging of TCP/IP connections (when configured) now happens in the main
5818 daemon process instead of the child process, so that the TCP/IP connection
5819 count is more accurate (but it can never be perfect).
5821 11. The use of "drop" in a nested ACL was not being handled correctly in the
5822 outer ACL. Now, if condition failure induced by the nested "drop" causes
5823 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5824 or "require"), the connection is dropped.
5826 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5827 that yield "discard" can now be used with an "accept" or a "discard" verb,
5828 but an error is generated for any others (because I can't see a useful way
5829 to define what should happen).
5831 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5832 are now processed in the same way as lines in the Exim configuration file.
5833 In particular, continuation lines are supported.
5835 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5837 15. Added -ti meaning -t -i.
5839 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5840 domains, and warn by logging if others are found.
5842 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5843 returned from GIFCONF when Exim is trying to find the list of interfaces on
5844 a host. The code in os.c has been modified to copy each ifreq to an aligned
5845 structure in all cases.
5847 Also, in some cases, the returned ifreq's were being copied to a 'struct
5848 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5849 means the last couple of bytes of an IPv6 address could be chopped if the
5850 ifreq contained only a normal sockaddr (14 bytes storage).
5852 18. Named domain lists were not supported in the hosts_treat_as_local option.
5853 An entry such as +xxxx was not recognized, and was treated as a literal
5856 19. Ensure that header lines added by a DATA ACL are included in the reject log
5857 if the ACL subsequently rejects the message.
5859 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5860 MD5 (which is deprecated).
5862 21. When testing a filter file using -bf, Exim was writing a message when it
5863 took the sender from a "From " line in the message, but it was not doing so
5864 when it took $return_path from a Return-Path: header line. It now does.
5866 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5867 with a valid header line field name (a series of printing characters
5868 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5870 23. Changed "disc" in the source to "disk" to conform to the documentation and
5871 the book and for uniformity.
5873 24. Ignore Sendmail's -Ooption=value command line item.
5875 25. When execve() failed while trying to run a command in a pipe transport,
5876 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5877 could be confused with a return value of 69 from the command itself. This
5878 has been changed to 127, the value the shell returns if it is asked to run
5879 a non-existent command. The wording for the related log line suggests a
5880 non-existent command as the problem.
5882 26. If received_header_text expands to an empty string, do not add a Received:
5883 header line to the message. (Well, it adds a token one on the spool, but
5884 marks it "old" so that it doesn't get used or transmitted.)
5886 27. Installed eximstats 1.28 (addition of -nt option).
5888 28. There was no check for failure on the call to getsockname() in the daemon
5889 code. This can fail if there is a shortage of resources on the system, with
5890 ENOMEM, for example. A temporary error is now given on failure.
5892 29. Contrary to the C standard, it seems that in some environments, the
5893 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5894 program. Exim now does this explicitly; it affects the formatting of
5895 timestamps using strftime().
5897 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5898 complaints. I've now initialized all the variables, to avoid this.
5900 32. Header lines added by a system filter were not being "seen" during
5901 transport-time rewrites.
5903 33. The info_callback() function passed to OpenSSL is set up with type void
5904 (*)(SSL *, int, int), as described somewhere. However, when calling the
5905 function (actually a macro) that sets it up, the type void(*)() is
5906 expected. I've put in a cast to prevent warnings from picky compilers.
5908 34. If a DNS black list lookup found a CNAME record, but there were no A
5909 records associated with the domain it pointed at, Exim crashed.
5911 35. If a DNS black list lookup returned more than one A record, Exim ignored
5912 all but the first. It now scans all returned addresses if a particular IP
5913 value is being sought. In this situation, the contents of the
5914 $dnslist_value variable are a list of all the addresses, separated by a
5917 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5918 to accept a host name and all its aliases if the forward lookup for any of
5919 them yielded the IP address of the incoming connection. Now it accepts only
5920 those names whose forward lookup yields the correct IP address. Any other
5921 names are discarded. This closes a loophole whereby a rogue DNS
5922 administrator could create reverse DNS records to break through a
5923 wildcarded host restriction in an ACL.
5925 37. If a user filter or a system filter that ran in a subprocess used any of
5926 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5927 the wrong values were passed to the pipe command ($thisaddress had the
5928 value of $0, $0 had the value of $1, etc). This bug was introduced by
5929 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5931 38. Improved the line breaking for long SMTP error messages from ACLs.
5932 Previously, if there was no break point between 40 and 75 characters, Exim
5933 left the rest of the message alone. Two changes have been made: (a) I've
5934 reduced the minimum length to 35 characters; (b) if it can't find a break
5935 point between 35 and 75 characters, it looks ahead and uses the first one
5936 that it finds. This may give the occasional overlong line, but at least the
5937 remaining text gets split now.
5939 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5940 file descriptors might be low, and that setting 1000 would always raise it.
5941 It turns out that in some environments, the limit is already over 1000 and
5942 that lowering it causes trouble. So now Exim takes care not to decrease it.
5944 40. When delivering a message, the value of $return_path is set to $sender_
5945 address at the start of routing (routers may change the value). By an
5946 oversight, this default was not being set up when an address was tested by
5947 -bt or -bv, which affected the outcome if any router or filter referred to
5950 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5951 log only when "message" or "log_message" are set. However, if one of the
5952 conditions was an address verification, or a call to a nested ACL, the
5953 messages generated by the underlying test were being passed through. This
5954 no longer happens. The underlying message is available in $acl_verify_
5955 message for both "message" and "log_message" expansions, so it can be
5956 passed through if needed.
5958 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5959 new expansion $bh_ to give the encoded byte string without charset
5960 translation. Translation happens only if iconv() is available; HAVE_ICONV
5961 indicates this at build time. HEADERS_CHARSET gives the charset to
5962 translate to; headers_charset can change it in the configuration, and
5963 "headers charset" can change it in an individual filter file.
5965 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5966 that creates RFC 2047 encoded "words" labels them as that charset instead
5967 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5968 expansion operator; (ii) when Exim creates a From: line for a local
5969 message; (iii) when a header line is rewritten to include a "phrase" part.
5971 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5972 buggy, causing it to skip the first lines of messages whose message ID
5973 ended in 'D'. This would not have bitten before Exim release 4.14, because
5974 message IDs were unlikely to end in 'D' before then. The effect was to have
5975 incorrect size information for certain domains.
5977 45. #include "config.h" was missing at the start of the crypt16.c module. This
5978 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5981 46. If there was a timeout during a "random" callout check, Exim treated it as
5982 a failure of the random address, and carried on sending RSET and the real
5983 address. If the delay was just some slowness somewhere, the response to the
5984 original RCPT would be taken as a response to RSET and so on, causing
5985 mayhem of various kinds.
5987 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5988 when I implemented it. It didn't allow for the fact that some option values
5989 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5990 the right test for positive values.
5992 48. Domain names in DNS records are case-independent. Exim always looks them up
5993 in lower case. Some resolvers return domain names in exactly the case they
5994 appear in the zone file, that is, they may contain uppercase letters. Not
5995 all resolvers do this - some return always lower case. Exim was treating a
5996 change of case by a resolver as a change of domain, similar to a widening
5997 of a domain abbreviation. This triggered its re-routing code and so it was
5998 trying to route what was effectively the same domain again. This normally
5999 caused routing to fail (because the router wouldn't handle the domain
6000 twice). Now Exim checks for this case specially, and just changes the
6001 casing of the domain that it ultimately uses when it transmits the message
6004 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6007 50. If a filter generated a file delivery with a non-absolute name (possible if
6008 no home directory exists for the router), the forbid_file option was not
6011 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6012 the existing equality matching.
6014 52. Exim was using ints instead of ino_t variables in some places where it was
6015 dealing with inode numbers.
6017 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6018 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6019 called TMPDIR, and if it finds it is different, it changes its value.
6021 54. The smtp_printf() function is now made available to local_scan() so
6022 additional output lines can be written before returning. There is also an
6023 smtp_fflush() function to enable the detection of a dropped connection.
6024 The variables smtp_input and smtp_batched_input are exported to
6027 55. Changed the default runtime configuration: the message "Unknown user"
6028 has been removed from the ACL, and instead placed on the localuser router,
6029 using the cannot_route_message feature. This means that any verification
6030 failures that generate their own messages won't get overridden. Similarly,
6031 the "Unrouteable address" message that was in the ACL for unverifiable
6032 relay addresses has also been removed.
6034 56. Added hosts_avoid_esmtp to the smtp transport.
6036 57. The exicyclog script was not checking for the esoteric option
6037 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6038 will work only if exicyclog is run under the appropriate euid.
6040 58. Following a discussion on the list, the rules by which Exim recognises line
6041 endings on incoming messages have been changed. The -dropcr and drop_cr
6042 options are now no-ops, retained only for backwards compatibility. The
6043 following line terminators are recognized: LF CRLF CR. However, special
6044 processing applies to CR:
6046 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6047 nor a local message in the state where . is a terminator.
6049 (ii) If a bare CR is encountered in a header line, an extra space is added
6050 after the line terminator so as not to end the header. The reasoning
6051 behind this is that bare CRs in header lines are most likely either
6052 to be mistakes, or people trying to play silly games.
6054 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6055 was being incorrectly given as 18 bytes larger than it should have been.
6056 This is a VOB (very old bug).
6058 60. This may never have affected anything current, but just in case it has:
6059 When the local host is found other than at the start of a list of hosts,
6060 the local host, those with the same MX, and any that follow, are discarded.
6061 When the list in question was part of a longer list of hosts, the following
6062 hosts (not currently being processed) were also being discarded. This no
6063 longer happens. I'm not sure if this situation could ever has previously
6066 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6069 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6070 header. This is something that is recommended in a new Internet Draft, and
6071 is something that is documented as being done by Sendmail. There are two
6072 possible values. For messages generated by the autoreply transport, Exim
6075 Auto-Submitted: auto-replied
6077 whereas for all other generated messages (e.g. bounces) it adds
6079 Auto-Submitted: auto-generated
6081 63. The "personal" condition in filters now includes a test for the
6082 Auto-Submitted: header. If it contains the string "auto-" the message it
6083 not considered personal.
6085 64. Added rcpt_include_affixes as a generic transport option.
6087 65. Added queue_only_override (default true).
6089 66. Added the syslog_duplication option.
6091 67. If what should have been the first header line of a message consisted of
6092 a space followed by a colon, Exim was mis-interpreting it as a header line.
6093 It isn't of course - it is syntactically invalid and should therefore be
6094 treated as the start of the message body. The misbehaviour could have
6095 caused a number of strange effects, including loss of data in subsequent
6096 header lines, and spool format errors.
6098 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6099 client host had authenticated. This control can now be exercised by an ACL
6100 for more flexibility.
6102 69. By default, callouts do not happen when testing with -bh. There is now a
6103 variant, -bhc, which does actually run the callout code, including
6104 consulting and updating the callout cache.
6106 70. Added support for saslauthd authentication, courtesy of Alexander
6109 71. If statvfs() failed on the spool or log directories while checking their
6110 size for availability, Exim confusingly gave the error "space shortage".
6111 Furthermore, in debugging mode it crashed with a floating point exception.
6112 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6113 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6114 serious problem, Exim now writes to the main and panic logs when this
6115 happens, with details of the failure. It then refuses to accept the
6116 incoming message, giving the message "spool directory problem" or "log
6117 directory problem" with a 421 code for SMTP messages.
6119 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6120 0, 1, and 2 exist, because some OS complain for execs without them (see
6121 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6122 descriptors. However, the code omitted to check that the open succeeded,
6123 causing mysterious errors if for some reason the permissions on /dev/null
6124 got screwed. Now Exim writes a message to the main and panic logs, and
6125 bombs out if it can't open /dev/null.
6127 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6128 interact so that it is all more flexible. It is supposed to remain
6129 backwards compatible. Also added extra_local_interfaces.
6131 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6132 to bomb out with an assertion failure - to the client this appears as a
6133 connection drop. This problem occurs in the part of the code that was taken
6134 from the Samba project. Fortunately, the assertion is in a very simple
6135 function, so I have fixed this by reproducing the function inline in the
6136 one place where it is called, and arranging for authentication to fail
6137 instead of killing the process with assert().
6139 75. The SPA client code was not working when the server requested OEM rather
6140 than Unicode encoding.
6142 76. Added code to make require_files with a specific uid setting more usable in
6143 the case where statting the file as root fails - usually a non-root-mounted
6144 NFS file system. When this happens and the failure is EACCES, Exim now
6145 forks a subprocess and does the per-uid checking as the relevant uid.
6147 77. Added process_log_path.
6149 78. If log_file_path was not explicitly set, a setting of check_log_space or
6150 check_log_inodes was ignored.
6152 79. If a space check for the spool or log partitions fails, the incident is now
6153 logged. Of course, in the latter case the data may get lost...
6155 80. Added the %p formatting code to string_format() so that it can be used to
6156 print addresses in debug_print(). Adjusted all the address printing in the
6157 debugging in store.c to use %p rather than %d.
6159 81. There was a concern that a line of code in smtp_in.c could overflow a
6160 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6161 initially expressed, the concern was not well-founded, because trailing
6162 spaces are removed early. However, if the trailing spaces were followed by
6163 a NULL, they did not get removed, so the overflow was possible. Two fixes
6166 (a) I re-wrote the offending code in a cleaner fashion.
6167 (b) If an incoming SMTP command contains a NULL character, it is rejected
6170 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6171 runs initgroups(), so that if the Exim user is in any additional groups,
6172 they will be used during message reception.
6178 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob