1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
15 TL/02 The BSD's have an arc4random API. One of the functions to induce
16 adding randomness was arc4random_stir(), but it has been removed in
17 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
18 function when detected.
20 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
21 cause callback expansion.
23 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
24 syntax errors in an expansion can be treated as a string instead of
25 logging or causing an error, due to the internal use of bool_lax
26 instead of bool when processing it.
28 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
29 server certificates when making smtp deliveries.
31 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
33 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
35 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
37 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
38 Merged patch from Sebastian Wiedenroth.
40 JH/05 Fix results-pipe from transport process. Several recipients, combined
41 with certificate use, exposed issues where response data items split
42 over buffer boundaries were not parsed properly. This eventually
43 resulted in duplicates being sent. This issue only became common enough
44 to notice due to the introduction of conection certificate information,
45 the item size being so much larger. Found and fixed by Wolfgang Breyha.
47 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
48 size buffer was used, resulting in syntax errors when an expansion
51 JH/07 Add support for directories of certificates when compiled with a GnuTLS
52 version 3.3.6 or later.
54 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
55 is EXPERIMENTAL_EVENT, the main-configuration and transport options
56 both become "event_action", the variables become $event_name, $event_data
57 and $event_defer_errno. There is a new variable $verify_mode, usable in
58 routers, transports and related events. The tls:cert event is now also
59 raised for inbound connections, if the main configuration event_action
62 TL/06 In test suite, disable OCSP for old versions of openssl which contained
63 early OCSP support, but no stapling (appears to be less than 1.0.0).
65 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
66 server certificate names available under the smtp transport option
67 "tls_verify_cert_hostname" now do not permit multi-component wildcard
70 JH/10 Time-related extraction expansions from certificates now use the main
71 option "timezone" setting for output formatting, and are consistent
72 between OpenSSL and GnuTLS compilations. Bug 1541.
74 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
75 encoded parameter in the incoming message. Bug 1558.
77 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
78 include certificate info, eximon was claiming there were spoolfile
81 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
83 JH/14 Log delivery-related information more consistently, using the sequence
84 "H=<name> [<ip>]" wherever possible.
86 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
87 are problematic for Debian distribution, omit them from the release
90 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
92 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
95 JH/17 Fix a null-indirection in certextract expansions when a nondefault
96 output list separator was used.
101 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
102 checkers that were complaining about end of non-void function with no
105 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
106 This was a regression intruduced in 4.83 by another bugfix.
108 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
110 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
111 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
117 TF/01 Correctly close the server side of TLS when forking for delivery.
119 When a message was received over SMTP with TLS, Exim failed to clear up
120 the incoming connection properly after forking off the child process to
121 deliver the message. In some situations the subsequent outgoing
122 delivery connection happened to have the same fd number as the incoming
123 connection previously had. Exim would try to use TLS and fail, logging
124 a "Bad file descriptor" error.
126 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
127 utilities have not been installed.
129 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
130 temporary space as the ACL may create new global variables.
132 TL/01 LDAP support uses per connection or global context settings, depending
133 upon the detected version of the libraries at build time.
135 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
136 to extract and use the src ip:port in logging and expansions as if it
137 were a direct connection from the outside internet. PPv2 support was
138 updated based on HAProxy spec change in May 2014.
140 JH/02 Add ${listextract {number}{list}{success}{fail}}.
142 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
143 Properly escape header and check for NULL return.
145 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
148 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
150 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
151 characters in header names, implemented as a verify condition.
152 Contributed by Michael Fischer v. Mollard.
154 TL/05 Rename SPF condition results err_perm and err_temp to standardized
155 results permerror and temperror. Previous values are deprecated but
156 still accepted. In a future release, err_perm and err_temp will be
157 completely removed, which will be a backward incompatibility if the
158 ACL tests for either of these two old results. Patch contributed by
159 user bes-internal on the mailing list.
161 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
163 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
164 selectors, in both main and reject logs.
166 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
169 JH/07 Add malware type "sock" for talking to simple daemon.
171 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
173 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
174 routers/transports under cutthrough routing.
176 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
177 numbers. Touch up "bool" conditional to keep the same definition.
179 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
181 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
183 JH/12 Expand items in router/transport headers_add or headers_remove lists
184 individually rather than the list as a whole. Bug 1452.
186 Required for reasonable handling of multiple headers_ options when
187 they may be empty; requires that headers_remove items with embedded
188 colons must have them doubled (or the list-separator changed).
190 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
191 view the policy declared in the DMARC record. Currently, $dmarc_status
192 is a combined value of both the record presence and the result of the
195 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
197 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
198 dnslookup router and the smtp transport (applying to the forward
201 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
202 of ldap servers used for a specific lookup. Patch provided by Heiko
205 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
206 New variable $lookup_dnssec_authenticated for observability.
208 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
209 Patch submitted by Lars Timman.
211 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
213 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
214 Requires trusted mode and valid format message id, aborts otherwise.
215 Patch contributed by Heiko Schlichting.
217 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
218 certextract with support for various fields. Bug 1358.
220 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
221 is requested by default, modifiable by smtp transport option
224 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
225 operate on certificate variables to give certificate fingerprints
226 Also new ${sha256:cert_variable}.
228 JH/23 The PRDR feature is moved from being Experimental into the mainline.
230 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
231 Christian Aistleitner.
233 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
235 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
236 file. Patch from Wolfgang Breyha.
238 JH/25 Expand the coverage of the delivery $host and $host_address to
239 client authenticators run in verify callout. Bug 1476.
241 JH/26 Port service names are now accepted for tls_on_connect_ports, to
242 align with daemon_smtp_ports. Bug 72.
244 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
245 support and error reporting did not work properly.
247 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
248 and is readable. Patch from Andrew Colin Kissa.
250 TL/14 Enhance documentation of ${run expansion and how it parses the
251 commandline after expansion, particularly in the case when an
252 unquoted variable expansion results in an empty value.
254 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
256 PP/02 Fix internal collision of T_APL on systems which support RFC3123
257 by renaming away from it. Addresses GH issue 15, reported by
260 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
262 TL/15 SECURITY: prevent double expansion in math comparison functions
263 (can expand unsanitized data). Not remotely exploitable.
270 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
272 PP/02 Make -n do something, by making it not do something.
273 When combined with -bP, the name of an option is not output.
275 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
278 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
279 $sender_host_name and config options to manage this, and basic check
282 PP/05 DSCP support for outbound connections and control modifier for inbound.
284 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
285 (Only plugin which currently uses this is kerberos4, which nobody should
286 be using, but we should make it available and other future plugins might
287 conceivably use it, even though it would break NAT; stuff *should* be
288 using channel bindings instead).
290 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
291 name; added for Sendmail compatibility; requires admin caller.
292 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
293 just ignore it); requires trusted caller.
294 Also parse but ignore: -Ac -Am -X<logfile>
297 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
299 TL/02 Add +smtp_confirmation as a default logging option.
301 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
302 Patch by Magnus Holmgren from 2007-02-20.
304 TL/04 Bugzilla 1281 - Spec typo.
305 Bugzilla 1283 - Spec typo.
306 Bugzilla 1290 - Spec grammar fixes.
308 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
310 TL/06 Add Experimental DMARC support using libopendmarc libraries.
312 TL/07 Fix an out of order global option causing a segfault. Reported to dev
313 mailing list by by Dmitry Isaikin.
315 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
317 JH/02 Support "G" suffix to numbers in ${if comparisons.
319 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
321 NM/01 Bugzilla 1197 - Spec typo
322 Bugzilla 1196 - Spec examples corrections
324 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
326 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
327 gnutls_enable_pkcs11, but renamed to more accurately indicate its
330 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
331 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
333 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
334 "acl {{name}{arg}...}", and optional args on acl condition
337 JH/05 Permit multiple router/transport headers_add/remove lines.
339 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
341 JH/07 Avoid using a waiting database for a single-message-only transport.
342 Performance patch from Paul Fisher. Bugzilla 1262.
344 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
347 JH/09 Add $headers_added variable, with content from use of ACL modifier
348 add_header (but not yet added to the message). Bugzilla 199.
350 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
351 Pulled from Bugzilla 817 by Wolfgang Breyha.
353 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
355 (nb: this is the same fix as in Exim 4.80.1)
357 JH/11 Add A= logging on delivery lines, and a client_set_id option on
360 JH/12 Add optional authenticated_sender logging to A= and a log_selector
363 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
365 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
366 advertise SMTP AUTH mechanism to us, instead of a generic
367 protocol violation error. Also, make Exim more robust to bad
368 data from the Dovecot auth socket.
370 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
372 When a queue runner is handling a message, Exim first routes the
373 recipient addresses, during which it prunes them based on the retry
374 hints database. After that it attempts to deliver the message to
375 any remaining recipients. It then updates the hints database using
378 So if a recipient address works intermittently, it can get repeatedly
379 deferred at routing time. The retry hints record remains fresh so the
380 address never reaches the final cutoff time.
382 This is a fairly common occurrence when a user is bumping up against
383 their storage quota. Exim had some logic in its local delivery code
384 to deal with this. However it did not apply to per-recipient defers
385 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
387 This change adds a proper retry rule check during routing so that the
388 final cutoff time is checked against the message's age. We only do
389 this check if there is an address retry record and there is not a
390 domain retry record; this implies that previous attempts to handle
391 the address had the retry_use_local_parts option turned on. We use
392 this as an approximation for the destination being like a local
393 delivery, as in LMTP.
395 I suspect this new check makes the old local delivery cutoff check
396 redundant, but I have not verified this so I left the code in place.
398 TF/02 Correct gecos expansion when From: is a prefix of the username.
400 Test 0254 submits a message to Exim with the header
404 When I ran the test suite under the user fanf2, Exim expanded
405 the header to contain my full name, whereas it should have added
406 a Resent-Sender: header. It erroneously treats any prefix of the
407 username as equal to the username.
409 This change corrects that bug.
411 GF/01 DCC debug and logging tidyup
412 Error conditions log to paniclog rather than rejectlog.
413 Debug lines prefixed by "DCC: " to remove any ambiguity.
415 TF/03 Avoid unnecessary rebuilds of lookup-related code.
417 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
418 Bug spotted by Jeremy Harris; was flawed since initial commit.
419 Would have resulted in OCSP responses post-SNI triggering an Exim
420 NULL dereference and crash.
422 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
424 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
425 Bug detection, analysis and fix by Samuel Thibault.
426 Bugzilla 1331, Debian bug #698092.
428 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
430 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
431 Server implementation by Todd Lyons, client by JH.
432 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
433 config variable "prdr_enable" controls whether the server
434 advertises the facility. If the client requests PRDR a new
435 acl_data_smtp_prdr ACL is called once for each recipient, after
436 the body content is received and before the acl_smtp_data ACL.
437 The client is controlled by bolth of: a hosts_try_prdr option
438 on the smtp transport, and the server advertisement.
439 Default client logging of deliveries and rejections involving
440 PRDR are flagged with the string "PRDR".
442 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
443 fclose(). Diagnosis by Todd Lyons.
445 PP/17 Update configure.default to handle IPv6 localhost better.
446 Patch by Alain Williams (plus minor tweaks).
449 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
450 This is now consistent with GnuTLS, and is now documented: the
451 previous undocumented portable approach to treating the option as
452 unset was to force an expansion failure. That still works, and
453 an empty string is now equivalent.
455 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
456 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
457 not performing validation itself.
459 PP/20 Added force_command boolean option to pipe transport.
460 Patch from Nick Koston, of cPanel Inc.
462 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
465 TF/04 Added udpsend ACL modifer and hexquote expansion operator
467 PP/21 Fix eximon continuous updating with timestamped log-files.
468 Broken in a format-string cleanup in 4.80, missed when I repaired the
469 other false fix of the same issue.
470 Report and fix from Heiko Schlichting.
473 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
474 Report from Prashanth Katuri.
476 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
477 It's SecureTransport, so affects any MacOS clients which use the
478 system-integrated TLS libraries, including email clients.
480 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
481 using a MIME ACL for non-SMTP local injection.
482 Report and assistance in diagnosis by Warren Baker.
484 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
486 JH/16 Fix comparisons for 64b. Bugzilla 1385.
488 TL/09 Add expansion variable $authenticated_fail_id to keep track of
489 last id that failed so it may be referenced in subsequent ACL's.
491 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
494 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
495 ldap library initialization, allowing self-signed CA's to be
496 used. Also properly sets require_cert option later in code by
497 using NULL (global ldap config) instead of ldap handle (per
498 session). Bug diagnosis and testing by alxgomz.
500 TL/12 Enhanced documentation in the ratelimit.pl script provided in
501 the src/util/ subdirectory.
503 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
504 renamed to Transport Post Delivery Action by Jeremy Harris, as
507 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
508 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
509 redis_servers = needs to be configured which will be used by the redis
510 lookup. Patch from Warren Baker, of The Packet Hub.
512 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
514 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
515 hostname or reverse DNS when processing a host list. Used suggestions
516 from multiple comments on this bug.
518 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
520 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
521 Missed a few lines, added it to make the runtest require no keyboard
524 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
525 contains upper case chars. Make router use caseful_local_part.
527 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
528 support when GnuTLS has been built with p11-kit.
534 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
536 This, or similar/improved, will also be change PP/11 of 4.82.
542 PP/01 Handle short writes when writing local log-files.
543 In practice, only affects FreeBSD (8 onwards).
544 Bugzilla 1053, with thanks to Dmitry Isaikin.
546 NM/01 Bugzilla 949 - Documentation tweak
548 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
551 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
553 PP/02 Implemented gsasl authenticator.
555 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
557 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
558 `pkg-config foo` for cflags/libs.
560 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
561 with rest of GSASL and with heimdal_gssapi.
563 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
564 `pkg-config foo` for cflags/libs for the TLS implementation.
566 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
567 properties get this fed in as external SSF. A number of robustness
568 and debugging improvements to the cyrus_sasl authenticator.
570 PP/08 cyrus_sasl server now expands the server_realm option.
572 PP/09 Bugzilla 1214 - Log authentication information in reject log.
573 Patch by Jeremy Harris.
575 PP/10 Added dbmjz lookup type.
577 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
579 PP/12 MAIL args handles TAB as well as SP, for better interop with
580 non-compliant senders.
581 Analysis and variant patch by Todd Lyons.
583 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
584 Bug report from Lars Müller <lars@samba.org> (via SUSE),
585 Patch from Dirk Mueller <dmueller@suse.com>
587 PP/13 tls_peerdn now print-escaped for spool files.
588 Observed some $tls_peerdn in wild which contained \n, which resulted
589 in spool file corruption.
591 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
592 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
593 or write after TLS renegotiation, which otherwise led to messages
596 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
597 as a tracking header (ie: a signed header comes before the signature).
598 Patch from Wolfgang Breyha.
600 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
601 comma-sep list; embedded commas doubled.
603 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
605 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
607 Report and patch from Dmitry Banschikov.
609 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
610 Removed SSL_clear() after SSL_new() which led to protocol negotiation
611 failures. We appear to now support TLS1.1+ with Exim.
613 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
614 lets Exim select keys and certificates based upon TLS SNI from client.
615 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
616 before an outbound SMTP session. New log_selector, +tls_sni.
618 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
619 NULL dereference. Report and patch from Alun Jones.
621 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
622 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
623 Not seeing resolver debug output on NetBSD, but suspect this is a
624 resolver implementation change.
626 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
627 Left warnings. Added "eximon gdb" invocation mode.
629 PP/21 Defaulting "accept_8bitmime" to true, not false.
631 PP/22 Added -bw for inetd wait mode support.
633 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
634 locate the relevant includes and libraries. Made this the default.
636 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
637 Bugzilla 1246, report and most of solution from Tomasz Kusy.
639 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
640 This may cause build issues on older platforms.
642 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
643 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
644 gnutls_require_mac & gnutls_require_protocols (no longer supported).
645 Added SNI support via GnuTLS too.
646 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
648 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
650 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
651 Applied second patch from Janne, implementing suggestion to default
652 multiple-strings-in-record handling to match SPF spec.
654 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
656 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
657 read-only, out of scope).
658 Patch from Wolfgang Breyha, report from Stuart Northfield.
660 PP/29 Fix three issues highlighted by clang analyser static analysis.
661 Only crash-plausible issue would require the Cambridge-specific
662 iplookup router and a misconfiguration.
663 Report from Marcin Mirosław.
665 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
667 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
668 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
669 As part of this, removing so much warning spew let me fix some minor
670 real issues in debug logging.
672 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
673 assignment on my part. Fixed.
675 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
676 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
677 Janne Snabb (who went above and beyond: thank you).
679 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
680 string otherwise requires a connection and a bunch more work and it's
681 relatively easy to get wrong. Should also expose TLS library linkage
684 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
685 64-bit ${eval} (JH/03).
687 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
688 GNU libc to support some of the 64-bit stuff, should not lead to
689 conflicts. Defined before os.h is pulled in, so if a given platform
690 needs to override this, it can.
692 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
693 protection layer was required, which is not implemented.
694 Bugzilla 1254, patch from Wolfgang Breyha.
696 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
697 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
698 tls_dhparam take prime identifiers. Also unbreak combination of
699 OpenSSL+DH_params+TLSSNI.
701 PP/39 Disable SSLv2 by default in OpenSSL support.
707 PP/01 Solaris build fix for Oracle's LDAP libraries.
708 Bugzilla 1109, patch from Stephen Usher.
710 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
712 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
715 TF/02 Fix a couple more cases where we did not log the error message
716 when unlink() failed. See also change 4.74-TF/03.
718 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
719 lock up or crash if it happened to be inside a call to libc when it
720 got a SIGUSR1 from exiwhat.
722 The SIGUSR1 handler appends the current process status to the process
723 log which is later printed by exiwhat. It used to use the general
724 purpose logging code to do this, but several functions it calls are
725 not safe for signals.
727 The new output code in the SIGUSR1 handler is specific to the process
728 log, and simple enough that it's easy to inspect for signal safety.
729 Removing some special cases also simplifies the general logging code.
730 Removing the spurious timestamps from the process log simplifies
733 TF/04 Improved ratelimit ACL condition.
735 The /noupdate option has been deprecated in favour of /readonly which
736 has clearer semantics. The /leaky, /strict, and /readonly update modes
737 are mutually exclusive. The update mode is no longer included in the
738 database key; it just determines when the database is updated. (This
739 means that when you upgrde Exim will forget old rate measurements.)
741 Exim now checks that the per_* options are used with an update mode that
742 makes sense for the current ACL. For example, when Exim is processing a
743 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
744 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
745 must specify per_mail/readonly. If you omit the update mode it defaults to
746 /leaky where that makes sense (as before) or /readonly where required.
748 The /noupdate option is now undocumented but still supported for
749 backwards compatibility. It is equivalent to /readonly except that in
750 ACLs where /readonly is required you may specify /leaky/noupdate or
751 /strict/noupdate which are treated the same as /readonly.
753 A useful new feature is the /count= option. This is a generalization
754 of the per_byte option, so that you can measure the throughput of other
755 aggregate values. For example, the per_byte option is now equivalent
756 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
758 The per_rcpt option has been generalized using the /count= mechanism
759 (though it's more complicated than the per_byte equivalence). When it is
760 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
761 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
762 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
763 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
764 Note that using per_rcpt with a non-readonly update mode in more than
765 one ACL will cause the recipients to be double-counted. (The per_mail
766 and per_byte options don't have this problem.)
768 The handling of very low rates has changed slightly. If the computed rate
769 is less than the event's count (usually one) then this event is the first
770 after a long gap. In this case the rate is set to the same as this event's
771 count, so that the first message of a spam run is counted properly.
773 The major new feature is a mechanism for counting the rate of unique
774 events. The new per_addr option counts the number of different
775 recipients that someone has sent messages to in the last time period. It
776 behaves like per_rcpt if all the recipient addresses are different, but
777 duplicate recipient addresses do not increase the measured rate. Like
778 the /count= option this is a general mechanism, so the per_addr option
779 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
780 example, measure the rate that a client uses different sender addresses
781 with the options per_mail/unique=$sender_address. There are further
782 details in the main documentation.
784 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
786 TF/06 Removed a few PCRE remnants.
788 TF/07 Automatically extract Exim's version number from tags in the git
789 repository when doing development or release builds.
791 PP/02 Raise smtp_cmd_buffer_size to 16kB.
792 Bugzilla 879. Patch from Paul Fisher.
794 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
795 Heavily based on revision 40f9a89a from Simon Arlott's tree.
798 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
800 PP/05 Variable $av_failed, true if the AV scanner deferred.
801 Bugzilla 1078. Patch from John Horne.
803 PP/06 Stop make process more reliably on build failure.
804 Bugzilla 1087. Patch from Heiko Schlittermann.
806 PP/07 Make maildir_use_size_file an _expandable_ boolean.
807 Bugzilla 1089. Patch from Heiko Schlittermann.
809 PP/08 Handle ${run} returning more data than OS pipe buffer size.
810 Bugzilla 1131. Patch from Holger Weiß.
812 PP/09 Handle IPv6 addresses with SPF.
813 Bugzilla 860. Patch from Wolfgang Breyha.
815 PP/10 GnuTLS: support TLS 1.2 & 1.1.
817 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
820 PP/11 match_* no longer expand right-hand-side by default.
821 New compile-time build option, EXPAND_LISTMATCH_RHS.
822 New expansion conditions, "inlist", "inlisti".
824 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
826 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
828 PP/14 fix log_write() format string regression from TF/03.
829 Bugzilla 1152. Patch from Dmitry Isaikin.
835 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
837 PP/02 Harmonised TLS library version reporting; only show if debugging.
838 Layout now matches that introduced for other libraries in 4.74 PP/03.
840 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
842 PP/04 New "dns_use_edns0" global option.
844 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
847 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
848 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
850 TK/01 Updated PolarSSL code to 0.14.2.
851 Bugzilla 1097. Patch from Andreas Metzler.
853 PP/07 Catch divide-by-zero in ${eval:...}.
856 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
859 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
860 format-string attack -- SECURITY: remote arbitrary code execution.
862 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
863 time unintentionally subject to list matching rules, letting the header
864 cause arbitrary Exim lookups (of items which can occur in lists, *not*
865 arbitrary string expansion). This allowed for information disclosure.
867 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
868 INT_MIN/-1 -- value coerced to INT_MAX.
874 NM/01 Workround for PCRE version dependency in version reporting
877 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
878 This fixes portability to compilers other than gcc, notably
879 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
881 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
882 makefiles for portability to HP-UX and POSIX correctness.
884 PP/01 Permit LOOKUP_foo enabling on the make command-line.
885 Also via indented variable definition in the Makefile.
886 (Debugging by Oliver Heesakkers).
888 PP/02 Restore caching of spamd results with expanded spamd_address.
889 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
891 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
892 Improves build reliability. Fix from: Frank Elsner
894 NM/02 Fix wide character breakage in the rfc2047 coding
895 Fixes bug 1064. Patch from Andrey N. Oktyabrski
897 NM/03 Allow underscore in dnslist lookups
898 Fixes bug 1026. Patch from Graeme Fowler
900 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
901 Code patches from Adam Ciarcinski of NetBSD.
903 NM/04 Fixed exiqgrep to cope with mailq missing size issue
906 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
907 is logged, to avoid truncation. Patch from John Horne.
909 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
910 Patch from Jakob Hirsch.
912 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
913 SQL string expansion failure details.
914 Patch from Andrey Oktyabrski.
916 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
917 Patch from Simon Arlott.
919 PP/09 New lookups functionality failed to compile on old gcc which rejects
920 extern declarations in function scope.
921 Patch from Oliver Fleischmann
923 PP/10 Use sig_atomic_t for flags set from signal handlers.
924 Check getgroups() return and improve debugging.
925 Fixed developed for diagnosis in bug 927 (which turned out to be
928 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
929 Patch from Mark Zealey.
931 PP/12 Bugzilla 1056: Improved spamd server selection.
932 Patch from Mark Zealey.
934 PP/13 Bugzilla 1086: Deal with maildir quota file races.
935 Based on patch from Heiko Schlittermann.
937 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
938 Patch from Uwe Doering, sign-off by Michael Haardt.
940 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
941 variable declaration deep within a block. Bug and patch from
944 PP/15 lookups-Makefile IRIX compatibilty coercion.
946 PP/16 Make DISABLE_DKIM build knob functional.
948 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
949 Patch by Simon Arlott
951 TF/03 Fix valgrind.h portability to C89 compilers that do not support
952 variable argument macros. Our copy now differs from upstream.
958 TF/01 Failure to get a lock on a hints database can have serious
959 consequences so log it to the panic log.
961 TF/02 Log LMTP confirmation messages in the same way as SMTP,
962 controlled using the smtp_confirmation log selector.
964 TF/03 Include the error message when we fail to unlink a spool file.
966 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
967 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
968 for maintaining out-of-tree patches for some time.
970 PP/01 Bugzilla 139: Documentation and portability issues.
971 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
972 Handle per-OS dynamic-module compilation flags.
974 PP/02 Let /dev/null have normal permissions.
975 The 4.73 fixes were a little too stringent and complained about the
976 permissions on /dev/null. Exempt it from some checks.
977 Reported by Andreas M. Kirchwitz.
979 PP/03 Report version information for many libraries, including
980 Exim version information for dynamically loaded libraries. Created
981 version.h, now support a version extension string for distributors
982 who patch heavily. Dynamic module ABI change.
984 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
985 privilege escalation vulnerability whereby the Exim run-time user
986 can cause root to append content of the attacker's choosing to
989 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
992 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
993 If dropping privileges for untrusted macros, we disabled normal logging
994 on the basis that it would fail; for the Exim run-time user, this is not
995 the case, and it resulted in successful deliveries going unlogged.
996 Fixed. Reported by Andreas Metzler.
1002 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1003 only prepend for the Resent-* case. Fixes regression introduced in
1004 Exim 4.70 by NM/22 for Bugzilla 607.
1006 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1007 increasing numbers of administrators be bitten by this.
1009 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1011 PP/03 Bugzilla 994: added openssl_options main configuration option.
1013 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1015 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1017 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1019 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1020 without a peer certificate, leading to a segfault because of an
1021 assumption that peers always have certificates. Be a little more
1022 paranoid. Problem reported by Martin Tscholak.
1024 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1025 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1026 NB: ClamAV planning to remove STREAM in "middle of 2010".
1027 CL also introduces -bmalware, various -d+acl logging additions and
1028 more caution in buffer sizes.
1030 PP/09 Implemented reverse_ip expansion operator.
1032 PP/10 Bugzilla 937: provide a "debug" ACL control.
1034 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1036 PP/12 Bugzilla 973: Implement --version.
1038 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1040 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1042 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1044 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1045 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1046 ignore trailing whitespace.
1048 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1050 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1053 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1054 Notification from Dr Andrew Aitchison.
1056 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1057 ExtendedDetectionInfo response format.
1058 Notification from John Horne.
1060 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1063 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1064 XSL and documented dependency on system catalogs, with examples of how
1067 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1070 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1071 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1072 configuration file which is writeable by the Exim user or group.
1074 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1075 of configuration files to cover files specified with the -C option if
1076 they are going to be used with root privileges, not just the default
1079 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1080 option (effectively making it always true).
1082 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1083 files to be used while preserving root privileges.
1085 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1086 that rogue child processes cannot use them.
1088 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1089 run-time user, instead of root.
1091 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1092 Exim run-time user without dropping privileges.
1094 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1095 result string, instead of calling string_vformat() twice with the same
1098 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1099 for other users. Others should always drop root privileges if they use
1100 -C on the command line, even for a whitelisted configure file.
1102 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1104 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1110 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1111 $data_path, and $header_path variables; fixed documentation bugs and
1114 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1115 exipick to access non-standard spools, including the "frozen" queue
1118 NM/01 Bugzilla 965: Support mysql stored procedures.
1119 Patch from Alain Williams
1121 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1123 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1124 Patch from Andreas Metzler
1126 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1127 Patch from Kirill Miazine
1129 NM/05 Bugzilla 671: Added umask to procmail example.
1131 JJ/03 installed exipick 20100323.0, fixing doc bug
1133 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1134 directory. Notification and patch from Dan Rosenberg.
1136 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1138 TK/02 Improve log output when DKIM signing operation fails.
1140 MH/01 Treat the transport option dkim_domain as a colon separated
1141 list, not as a single string, and sign the message with each element,
1142 omitting multiple occurences of the same signer.
1144 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1145 Bugzilla 985, 986. Patch by Simon Arlott
1147 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1148 Patch by Simon Arlott
1150 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1151 MBX locking. Notification from Dan Rosenberg.
1157 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1159 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1161 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1163 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1165 NM/04 Bugzilla 915: Flag broken perl installation during build.
1171 TK/01 Added patch by Johannes Berg that expands the main option
1172 "spamd_address" if it starts with a dollar sign.
1174 TK/02 Write list of recipients to X-Envelope-Sender header when building
1175 the mbox-format spool file for content scanning (suggested by Jakob
1178 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1179 (http://www.dcc-servers.net/) support via dccifd. Activated by
1180 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1182 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1183 by Mark Daniel Reidel <mr@df.eu>.
1185 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1186 When building exim an external PCRE library is now needed -
1187 PCRE is a system library on the majority of modern systems.
1188 See entry on PCRE_LIBS in EDITME file.
1190 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1191 conversation. Added nologin parameter to request.
1192 Patch contributed by Kirill Miazine.
1194 TF/01 Do not log submission mode rewrites if they do not change the address.
1196 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1198 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1199 log files in place. Contributed by Roberto Lima.
1201 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1203 TF/03 Bugzilla 615: When checking the local_parts router precondition
1204 after a local_part_suffix or local_part_prefix option, Exim now
1205 does not use the address's named list lookup cache, since this
1206 contains cached lookups for the whole local part.
1208 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1209 Robert Millan. Documentation is in experimental-spec.txt.
1211 TF/04 Bugzilla 668: Fix parallel build (make -j).
1213 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1215 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1216 Patch provided by Jan Srzednicki.
1218 TF/05 Leading white space used to be stripped from $spam_report which
1219 wrecked the formatting. Now it is preserved.
1221 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1222 that they are available at delivery time.
1224 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1226 TF/08 TLS error reporting now respects the incoming_interface and
1227 incoming_port log selectors.
1229 TF/09 Produce a more useful error message if an SMTP transport's hosts
1230 setting expands to an empty string.
1232 NM/06 Bugzilla 744: EXPN did not work under TLS.
1233 Patch provided by Phil Pennock.
1235 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1236 Patch provided by Richard Godbee.
1238 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1239 acl_smtp_notquit, added index entry.
1241 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1242 Patch provided by Eugene Bujak.
1244 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1245 accept(). Patch provided by Maxim Dounin.
1247 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1248 Patch provided by Phil Pennock.
1250 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1252 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1253 Patch provided by Brad "anomie" Jorsch.
1255 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1256 Patch provided by Dean Brooks.
1258 TK/05 Add native DKIM support (does not depend on external libraries).
1260 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1261 Patch provided by Graeme Fowler.
1263 NM/16 Bugzilla 851: Documentation example syntax fix.
1265 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1267 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1270 NM/19 Bugzilla 745: TLS version reporting.
1271 Patch provided by Phil Pennock.
1273 NM/20 Bugzilla 167: bool: condition support.
1274 Patch provided by Phil Pennock.
1276 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1277 clients. Patch provided by Phil Pennock.
1279 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1280 Patch provided by Brad "anomie" Jorsch.
1282 NM/23 Bugzilla 687: Fix misparses in eximstats.
1283 Patch provided by Heiko Schlittermann.
1285 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1286 Patch provided by Heiko Schlittermann.
1288 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1289 plus update to original patch.
1291 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1293 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1294 Patch provided by David Brownlee.
1296 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1298 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1300 NM/30 Bugzilla 888: TLS documentation bugfixes.
1302 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1304 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1305 Unlike the original bugzilla I have changed all shell scripts in src tree.
1307 NM/33 Bugzilla 898: Transport filter timeout fix.
1308 Patch by Todd Rinaldo.
1310 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1311 Patch by Serge Demonchaux.
1313 NM/35 Bugzilla 39: Base64 decode bug fixes.
1314 Patch by Jakob Hirsch.
1316 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1318 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1320 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1322 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1328 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1329 ALT-N's libdkim that I have put here:
1330 http://duncanthrax.net/exim-experimental/
1332 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1333 were called 'true' and it seems that C99 defines that as a reserved
1334 keyword to be used with 'bool' variable types. That means you could
1335 not include C99-style headers which use bools without triggering
1336 build errors in sieve.c.
1338 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1339 as mailq or other aliases. Changed the --help handling significantly
1340 to do whats expected. exim_usage() emits usage/help information.
1342 SC/01 Added the -bylocaldomain option to eximstats.
1344 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1346 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1348 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1354 PH/01 Another patch from the Sieve maintainer.
1356 PH/02 When an IPv6 address is converted to a string for single-key lookup
1357 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1358 dots are used instead of colons so that keys in lsearch files need not
1359 contain colons. This was done some time before quoting was made available
1360 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1361 (notated using the quote facility) so as to distinguish them from IPv4
1362 keys. This meant that lookups for IP addresses in host lists did not work
1363 for iplsearch lookups.
1365 This has been fixed by arranging for IPv6 addresses to be expressed with
1366 colons if the lookup type is iplsearch. This is not incompatible, because
1367 previously such lookups could never work.
1369 The situation is now rather anomolous, since one *can* have colons in
1370 ordinary lsearch keys. However, making the change in all cases is
1371 incompatible and would probably break a number of configurations.
1373 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1376 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1377 conversion specification without a maximum field width, thereby enabling
1378 a rogue spamd server to cause a buffer overflow. While nobody in their
1379 right mind would setup Exim to query an untrusted spamd server, an
1380 attacker that gains access to a server running spamd could potentially
1381 exploit this vulnerability to run arbitrary code as the Exim user.
1383 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1384 $primary_hostname instead of what libspf2 thinks the hosts name is.
1386 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1387 a directory entry by the name of the lookup key. Previously, if a
1388 symlink pointed to a non-existing file or a file in a directory that
1389 Exim lacked permissions to read, a lookup for a key matching that
1390 symlink would fail. Now it is enough that a matching directory entry
1391 exists, symlink or not. (Bugzilla 503.)
1393 PH/03 The body_linecount and body_zerocount variables are now exported in the
1396 PH/04 Added the $dnslist_matched variable.
1398 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1399 This means they are set thereafter only if the connection becomes
1402 PH/06 Added the client_condition to authenticators so that some can be skipped
1403 by clients under certain conditions.
1405 PH/07 The error message for a badly-placed control=no_multiline_responses left
1406 "_responses" off the end of the name.
1408 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1410 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1411 (without spaces) instead of just copying the configuration text.
1413 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1415 PH/11 Added $max_received_linelength.
1417 PH/12 Added +ignore_defer and +include_defer to host lists.
1419 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1420 way in which PCRE > 7.0 is built.
1422 PH/14 Implemented queue_only_load_latch.
1424 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1425 MAIL command. The effect was to mangle the value on 64-bit systems.
1427 PH/16 Another patch from the Sieve maintainer.
1429 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1431 PH/18 If a system quota error occurred while trying to create the file for
1432 a maildir delivery, the message "Mailbox is full" was not appended to the
1433 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1434 only to a quota excession during the actual writing of the file.
1436 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1437 characters?) which causes problems in log lines. The DN values are now
1438 passed through string_printing() before being added to log lines.
1440 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1441 and InterBase are left for another time.)
1443 PH/21 Added message_body_newlines option.
1445 PH/22 Guard against possible overflow in moan_check_errorcopy().
1447 PH/23 POSIX allows open() to be a macro; guard against that.
1449 PH/24 If the recipient of an error message contained an @ in the local part
1450 (suitably quoted, of course), incorrect values were put in $domain and
1451 $local_part during the evaluation of errors_copy.
1457 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1458 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1461 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1462 issue a MAIL command.
1464 PH/02 In an ACL statement such as
1466 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1468 if a client was not listed at all, or was listed with a value other than
1469 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1470 the condition was not true (as it should be), so access was not denied.
1471 The bug was that the ! inversion was incorrectly passed on to the second
1472 item. This has been fixed.
1474 PH/03 Added additional dnslists conditions == and =& which are different from
1475 = and & when the dns lookup returns more than one IP address.
1477 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1478 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1480 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1481 FSYNC, which compiles an option called disable_fsync that allows for
1482 bypassing fsync(). The documentation is heavily laced with warnings.
1484 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1486 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1487 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1488 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1489 including adding "make clean"; (3) Added -fPIC when compiling the test
1490 dynamically loaded module, to get rid of a warning.
1492 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1493 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1494 = 0s. The bug is otherwise harmless.
1496 PH/07 There was a bug in the dovecot authenticator such that the value of
1497 $auth1 could be overwritten, and so not correctly preserved, after a
1498 successful authentication. This usually meant that the value preserved by
1499 the server_setid option was incorrect.
1501 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1503 PH/09 Installed PCRE release 7.0.
1505 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1506 run for batched SMTP input. It is now run at the start of every message
1507 in the batch. While fixing this I discovered that the process information
1508 (output by running exiwhat) was not always getting set for -bs and -bS
1509 input. This is fixed, and it now also says "batched" for BSMTP.
1511 PH/11 Added control=no_pipelining.
1513 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1514 patch, slightly modified), and move the expansion of helo_data till after
1515 the connection is made in the smtp transport (so it can use these
1518 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1520 PH/14 Added log_selector = +pid.
1522 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1524 PH/16 Add ${if forany and ${if forall.
1526 PH/17 Added dsn_from option to vary the From: line in DSNs.
1528 PH/18 Flush SMTP output before performing a callout, unless control =
1529 no_callout_flush is set.
1531 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1532 was true (the default) a successful delivery failed to delete the retry
1533 item, thus causing premature timeout of the address. The bug is now
1536 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1538 PH/21 Long custom messages for fakedefer and fakereject are now split up
1539 into multiline reponses in the same way that messages for "deny" and
1540 other ACL rejections are.
1542 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1543 with slight modification.
1545 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1546 draft, changing the syntax and factoring some duplicate code".
1548 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1549 for deliveries of the second and subsequent messages over the same SMTP
1552 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1553 ${reduce, with only minor "tidies".
1555 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1557 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1558 expansion side effects.
1560 PH/27 When a message times out after an over-quota error from an Exim-imposed
1561 quota, the bounce message says "mailbox is full". This message was not
1562 being given when it was a system quota that was exceeded. It now should
1565 MH/03 Made $recipients available in local_scan(). local_scan() already has
1566 better access to the recipient list through recipients_list[], but
1567 $recipients can be useful in postmaster-provided expansion strings.
1569 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1570 in the case of a MAIL command with additional options following the
1571 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1572 were accidentally chopped off.
1574 PH/29 SMTP synchronization checks are implemented when a command is read -
1575 there is a check that no more input is waiting when there shouldn't be
1576 any. However, for some commands, a delay in an ACL can mean that it is
1577 some time before the response is written. In this time, more input might
1578 arrive, invalidly. So now there are extra checks after an ACL has run for
1579 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1580 pipelining has not been advertised.
1582 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1584 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1585 returned in the SMTP response when smtp_return_error_details was false.
1586 This has been fixed.
1588 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1589 instead of the C I/O that was originally supplied, because problems were
1590 reported on Solaris.
1592 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1593 Exim which did not show up earlier: it was assuming that a call to
1594 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1595 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1596 was a macro that became an assignment, so it seemed to work. This has
1597 changed to a proper function call with a void return, hence the compile
1598 error. Exim's code has been fixed.
1600 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1603 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1605 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1607 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1608 because it turns out to be too big or there is a timeout) there is no
1609 "Completed" line in the log. When some messages of this type were
1610 selected by exigrep, they were listed as "not completed". Others were
1611 picked up by some special patterns. I have improved the selection
1612 criteria to be more general.
1614 PH/38 The host_find_failed option in the manualroute router can now be set
1615 to "ignore", to completely ignore a host whose IP address cannot be
1616 found. If all hosts are ignored, the behaviour is controlled by the new
1617 host_all_ignored option.
1619 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1620 homing or because of multiple MX records with /mx) generated more than
1621 one IP address, and the following item turned out to be the local host,
1622 all the secondary addresses of the first item were incorrectly removed
1623 from the list, along with the local host and any following hosts (which
1624 is what is supposed to happen).
1626 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1627 whoever called Exim into the -H file. In the case of the daemon it was
1628 behaving confusingly. When first started, it used values for whoever
1629 started the daemon, but after a SIGHUP it used the Exim user (because it
1630 calls itself on a restart). I have changed the code so that it now always
1633 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1634 message are rejected with the same error (e.g. no authentication or bad
1635 sender address), and a DATA command is nevertheless sent (as can happen
1636 with PIPELINING or a stupid MUA), the error message that was given to the
1637 RCPT commands is included in the rejection of the DATA command. This is
1638 intended to be helpful for MUAs that show only the final error to their
1641 PH/42 Another patch from the Sieve maintainer.
1643 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1644 Eximstats - Fixed some broken HTML links and added missing column headers
1646 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1647 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1649 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1651 PH/43 Yet another patch from the Sieve maintainer.
1653 PH/44 I found a way to check for a TCP/IP connection going away before sending
1654 the response to the final '.' that terminates a message, but only in the
1655 case where the client has not sent further data following the '.'
1656 (unfortunately, this is allowed). However, in many cases there won't be
1657 any further data because there won't be any more messages to send. A call
1658 to select() can be used: if it shows that the input is "ready", there is
1659 either input waiting, or the socket has been closed. An attempt to read
1660 the next input character can distinguish the two cases. Previously, Exim
1661 would have sent an OK response which the client would never have see.
1662 This could lead to message repetition. This fix should cure that, at
1663 least in a lot of common cases.
1665 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1666 advertised in response to EHLO.
1672 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1673 fixed by 4.65/MH/01 (is this a record?) are fixed:
1675 (i) An empty string was always treated as zero by the numeric comparison
1676 operators. This behaviour has been restored.
1678 (ii) It is documented that the numeric comparison operators always treat
1679 their arguments as decimal numbers. This was broken in that numbers
1680 starting with 0 were being interpreted as octal.
1682 While fixing these problems I realized that there was another issue that
1683 hadn't been noticed. Values of message_size_limit (both the global option
1684 and the transport option) were treated as octal if they started with 0.
1685 The documentation was vague. These values are now always treated as
1686 decimal, and I will make that clear in the documentation.
1692 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1693 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1696 MH/01 Don't check that the operands of numeric comparison operators are
1697 integers when their expansion is in "skipping" mode (fixes bug
1698 introduced by 4.64-PH/07).
1700 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1701 child addresses, Exim now panics and dies. Previously, because the count
1702 is held in a short int, deliveries were likely to be lost. As such a
1703 large number of recipients for a single message is ridiculous
1704 (performance will be very, very poor), I have chosen to impose a limit
1705 rather than extend the field.
1711 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1712 leftover -K file (the existence of which was triggered by #402).
1713 While we were at it, introduced process PID as part of the -K
1714 filename. This should rule out race conditions when creating
1717 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1718 processing considerably. Previous code took too long for large mails,
1719 triggering a timeout which in turn triggers #401.
1721 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1722 in the DK code in transports.c. sendfile() is not really portable,
1723 hence the _LINUX specificness.
1725 TF/01 In the add_headers option to the mail command in an Exim filter,
1726 there was a bug that Exim would claim a syntax error in any
1727 header after the first one which had an odd number of characters
1730 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1731 callout verification, Exim cached a "reject" for the entire domain. This
1732 is correct for most verifications, but it is not correct for a recipient
1733 verification with use_sender or use_postmaster set, because in that case
1734 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1735 case of MAIL FROM:<> rejection from other early rejections (e.g.
1736 rejection of HELO). When verifying a recipient using a non-null MAIL
1737 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1738 Whatever the result of the callout, the value of the domain cache is
1739 left unchanged (for any other kind of callout, getting as far as trying
1740 RCPT means that the domain itself is ok).
1742 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1745 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1746 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1747 session, the daemon ignores ECONNECT errors and logs others; it now
1748 ignores EPIPE as well.
1750 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1751 (quoted-printable decoding).
1753 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1754 later the small subsequent patch to fix an introduced bug.
1756 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1758 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1760 PH/08 An error is now given if message_size_limit is specified negative.
1762 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1763 to be given (somewhat) arbitrary names.
1765 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1768 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1769 miscellaneous code fixes
1771 PH/10 Added the log_reject_target ACL modifier to specify where to log
1774 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1775 hostname. This is wrong, because it relates to the incoming message (and
1776 probably the interface on which it is arriving) and not to the outgoing
1777 callout (which could be using a different interface). This has been
1778 changed to use the value of the helo_data option from the smtp transport
1779 instead - this is what is used when a message is actually being sent. If
1780 there is no remote transport (possible with a router that sets up host
1781 addresses), $smtp_active_hostname is used.
1783 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1784 tweaks were necessary in order to get it to work (see also 21 below):
1785 (a) The code assumed that strncpy() returns a negative number on buffer
1786 overflow, which isn't the case. Replaced with Exim's string_format()
1788 (b) There were several signed/unsigned issues. I just did the minimum
1789 hacking in of casts. There is scope for a larger refactoring.
1790 (c) The code used strcasecmp() which is not a standard C function.
1791 Replaced with Exim's strcmpic() function.
1792 (d) The code set only $1; it now sets $auth1 as well.
1793 (e) A simple test gave the error "authentication client didn't specify
1794 service in request". It would seem that Dovecot has changed its
1795 interface. Fortunately there's a specification; I followed it and
1796 changed what the client sends and it appears to be working now.
1798 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1801 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1802 address A is aliased to B and C, where B exists and C does not. Without
1803 -v the output is "A verified" because verification stops after a
1804 successful redirection if more than one address is generated. However,
1805 with -v the child addresses are also verified. Exim was outputting "A
1806 failed to verify" and then showing the successful verification for C,
1807 with its parentage. It now outputs "B failed to verify", showing B's
1808 parentage before showing the successful verification of C.
1810 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1811 look up a TXT record in a specific list after matching in a combined
1814 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1815 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1816 they consult the DNS. I had assumed they would set it the way they
1817 wanted; and indeed my experiments on Linux seem to show that in some
1818 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1819 To be on the safe side, however, I have now made the interface to
1820 host_find_byname() similar to host_find_bydns(), with an argument
1821 containing the DNS resolver options. The host_find_byname() function now
1822 sets these options at its start, just as host_find_bydns() does. The smtp
1823 transport options dns_qualify_single and dns_search_parents are passed to
1824 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1825 of host_find_byname() use the default settings of RES_DEFNAMES
1826 (qualify_single) but not RES_DNSRCH (search_parents).
1828 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1829 spool_read_header() do less string testing, by means of a preliminary
1830 switch on the second character of optional "-foo" lines. (This is
1831 overdue, caused by the large number of possibilities that now exist.
1832 Originally there were few.) While I was there, I also converted the
1833 str(n)cmp tests so they don't re-test the leading "-" and the first
1834 character, in the hope this might squeeze out yet more improvement.
1836 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1837 flag allowing group syntax was set by the header_syntax check but not
1838 turned off, possible causing trouble later; (2) The flag was not being
1839 set at all for the header_verify test, causing "group"-style headers to
1840 be rejected. I have now set it in this case, and also caused header_
1841 verify to ignore an empty address taken from a group. While doing this, I
1842 came across some other cases where the code for allowing group syntax
1843 while scanning a header line wasn't quite right (mostly, not resetting
1844 the flag correctly in the right place). These bugs could have caused
1845 trouble for malformed header lines. I hope it is now all correct.
1847 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1848 with the "reply" argument non-NULL. The code, however (which originally
1849 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1850 but it didn't always do it. This confused somebody who was copying the
1851 code for some other use. I have removed all the tests.
1853 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1854 feature that was used to support insecure browsers during the U.S. crypto
1855 embargo. It requires special client support, and Exim is probably the
1856 only MTA that supported it -- and would never use it because real RSA is
1857 always available. This code has been removed, because it had the bad
1858 effect of slowing Exim down by computing (never used) parameters for the
1859 RSA_EXPORT functionality.
1861 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1862 authenticator to fail if there's a tab character in the incoming data
1863 (there should never be unless someone is messing about, as it's supposed
1864 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1865 if the connection is using TLS or if the remote IP is the same as the
1866 local IP, and the "valid-client-cert option" if a client certificate has
1869 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1870 authenticators. This can be used for authorization after authentication
1871 succeeds. (In the case of plaintext, it servers for both authentication
1874 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1875 if any retry times were supplied.
1877 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1878 connection, where there is no client IP address to check. In this
1879 situation, the verify now always succeeds.
1881 PH/25 Applied John Jetmore's -Mset patch.
1883 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1885 PH/27 In a string expansion for a processed (not raw) header when multiple
1886 headers of the same name were present, leading whitespace was being
1887 removed from all of them, but trailing whitespace was being removed only
1888 from the last one. Now trailing whitespace is removed from each header
1889 before concatenation. Completely empty headers in a concatenation (as
1890 before) are ignored.
1892 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1893 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1895 PH/29 [Removed. This was a change that I later backed out, and forgot to
1896 correct the ChangeLog entry (that I had efficiently created) before
1897 committing the later change.]
1899 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1900 address errors (4xx response to RCPT) over the same connection as other
1901 messages routed to the same hosts. Such deliveries are always "forced",
1902 so retry times are not inspected. This resulted in far too many retries
1903 for the affected addresses. The effect occurred only when there were more
1904 hosts than the hosts_max_try setting in the smtp transport when it had
1905 the 4xx errors. Those hosts that it had tried were not added to the list
1906 of hosts for which the message was waiting, so if all were tried, there
1907 was no problem. Two fixes have been applied:
1909 (i) If there are any address or message errors in an SMTP delivery, none
1910 of the hosts (tried or untried) are now added to the list of hosts
1911 for which the message is waiting, so the message should not be a
1912 candidate for sending over the same connection that was used for a
1913 successful delivery of some other message. This seems entirely
1914 reasonable: after all the message is NOT "waiting for some host".
1915 This is so "obvious" that I'm not sure why it wasn't done
1916 previously. Hope I haven't missed anything, but it can't do any
1917 harm, as the worst effect is to miss an optimization.
1919 (ii) If, despite (i), such a delivery is accidentally attempted, the
1920 routing retry time is respected, so at least it doesn't keep
1921 hammering the server.
1923 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1924 in ${readsocket because some servers need this prod.
1926 PH/32 Added some extra debug output when updating a wait-xxx database.
1928 PH/33 The hint "could be header name not terminated by colon", which has been
1929 given for certain expansion errors for a long time, was not being given
1930 for the ${if def:h_colon_omitted{... case.
1932 PH/34 The spec says: "With one important exception, whenever a domain list is
1933 being scanned, $domain contains the subject domain." There was at least
1934 one case where this was not true.
1936 PH/35 The error "getsockname() failed: connection reset by peer" was being
1937 written to the panic log as well as the main log, but it isn't really
1938 panic-worthy as it just means the connection died rather early on. I have
1939 removed the panic log writing for the ECONNRESET error when getsockname()
1942 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1943 runs only) independently of the message's sender address. This meant
1944 that, if the 4xx error was in fact related to the sender, a different
1945 message to the same recipient with a different sender could confuse
1946 things. In particualar, this can happen when sending to a greylisting
1947 server, but other circumstances could also provoke similar problems.
1948 I have changed the default so that the retry time for these errors is now
1949 based a combination of the sender and recipient addresses. This change
1950 can be overridden by setting address_retry_include_sender=false in the
1953 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1954 remote server are returned as part of bounce messages. This was not
1955 happening for LMTP over a pipe (the lmtp transport), but now it is the
1956 same for both kinds of LMTP.
1958 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1959 in header lines that were in fact CNAMEs. This is no longer the case.
1961 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1962 and queue runs started by the daemon processed all messages. This has
1963 been fixed so that -R and -S can now usefully be given with -q<time>.
1965 PH/40 Import PCRE release 6.7 (fixes some bugs).
1967 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1969 PH/42 Give an error if -q is specified more than once.
1971 PH/43 Renamed the variables $interface_address and $interface_port as
1972 $received_ip_address and $received_port, to make it clear that these
1973 values apply to message reception, and not to the outgoing interface when
1974 a message is delivered. (The old names remain recognized, of course.)
1976 PH/44 There was no timeout on the connect() call when using a Unix domain
1977 socket in the ${readsocket expansion. There now is.
1979 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1980 be meaningful with "accept".
1982 SC/01 Eximstats V1.43
1983 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1985 SC/02 Eximstats V1.44
1986 Use a glob alias rather than an array ref in the generated
1987 parser. This improves both readability and performance.
1989 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1990 Collect SpamAssassin and rejection statistics.
1991 Don't display local sender or destination tables unless
1992 there is data to show.
1993 Added average volumes into the top table text output.
1995 SC/04 Eximstats V1.46
1996 Collect data on the number of addresses (recipients)
1997 as well as the number of messages.
1999 SC/05 Eximstats V1.47
2000 Added 'Message too big' to the list of mail rejection
2001 reasons (thanks to Marco Gaiarin).
2003 SC/06 Eximstats V1.48
2004 Mainlog lines which have GMT offsets and are too short to
2005 have a flag are now skipped.
2007 SC/07 Eximstats V1.49 (Alain Williams)
2008 Added the -emptyok flag.
2010 SC/08 Eximstats V1.50
2011 Fixes for obtaining the IP address from reject messages.
2013 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2014 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2015 whitesspace changes from 4.64-PH/27
2017 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2020 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2021 are found, allow negative numbers in numeric criteria)
2023 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2025 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2026 to match changes made in 4.64-PH/43
2028 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2030 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2031 "domain" is actually an IP address, and force a failure. This locks out
2032 those revolvers/nameservers that support "A-for-A" lookups, in
2033 contravention of the specifications.
2035 PH/48 When a host name was looked up from an IP address, and the subsequent
2036 forward lookup of the name timed out, the host name was left in
2037 $sender_host_name, contrary to the specification.
2039 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2040 restricted to single-key lookups, Exim was not diagnosing an error if
2041 * or *@ was used with a query-style lookup.
2043 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2045 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2046 long ago, but noone interested enough thought of it. Let's just say that
2047 the "1.1" means that there are some new functions that weren't there at
2048 some point in the past.
2050 PH/51 Error processing for expansion failure of helo_data from an smtp
2051 transport during callout processing was broken.
2053 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2054 tested/used via the -bh/-bhc/-bs options.
2056 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2057 bug, fixed in subsequent PCRE releases).
2059 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2060 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2062 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2068 SC/01 Use a glob alias rather than an array ref in eximstats generated
2069 parser. This improves both readability and performance.
2071 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2072 Don't display local sender or destination tables in eximstats unless
2073 there is data to show.
2074 Added average volumes into the eximstats top table text output.
2076 SC/03 Collect data on the number of addresses (recipients) as well
2077 as the number of messages in eximstats.
2079 TF/01 Correct an error in the documentation for the redirect router. Exim
2080 does (usually) call initgroups() when daemonizing.
2082 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2083 with consistent privilege compared to when running as a daemon.
2085 TF/03 Note in the spec that $authenticated_id is not set for local
2086 submissions from trusted users.
2088 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2089 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2091 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2092 by adding some example configuration directives to the default
2093 configuration file. A little bit of work is required to uncomment the
2094 directives and define how usernames and passwords are checked, but
2095 there is now a framework to start from.
2097 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2098 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2099 without this. I don't know how relevant this is to other LDAP libraries.
2101 PH/02 Add the verb name to the "unknown ACL verb" error.
2103 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2105 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2107 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2108 directory not expanded when it should be if an expanded home directory
2109 was set for the address (which is overridden by the transport).
2111 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2114 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2115 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2116 because it is too late at that time, and has no effect.
2118 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2119 security issue with \' (bugzilla #107). I could not use the
2120 PQescapeStringConn() function, because it needs a PGconn value as one of
2123 PH/08 When testing addresses using -bt, indicate those final addresses that
2124 are duplicates that would not cause an additional delivery. At least one
2125 person was confused, thinking that -bt output corresponded to deliveries.
2126 (Suppressing duplicates isn't a good idea as you lose the information
2127 about possibly different redirections that led to the duplicates.)
2129 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2130 systems where poll() doesn't work, in particular OS X.
2132 PH/10 Added more information to debugging output for retry time not reached.
2134 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2135 operations in malware.c.
2137 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2140 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2141 syslog_duplication set false, log lines that would normally be written
2142 both the the main log and to the reject log were not written to syslog at
2145 PH/14 In the default configuration, change the use of "message" in ACL warn
2146 statements to "add_header".
2148 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2149 not followed by a command (e.g. "seen endif").
2151 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2152 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2155 PH/17 Added extra conditions to the default value of delay_warning_condition
2159 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2160 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2161 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2164 The Auto-Submitted: and various List- headers are standardised, whereas I
2165 don't think Precedence: ever was.
2167 PH/18 Refactored debugging code in route_finduser() to show more information,
2168 in particular, the error code if getpwnam() issues one.
2170 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2171 This is apparently needed in addition to the PH/07 change above to avoid
2172 any possible encoding problems.
2174 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2175 but not after initializing Perl.
2177 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2178 output them only if debugging. By default they are written stderr,
2179 apparently, which is not desirable.
2181 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2184 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2187 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2189 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2190 authenticated or an ident call has been made. Suppress the default
2191 values for $authenticated_id and $authenticated_sender (but permit -oMai
2192 and -oMas) when testing with -bh.
2194 PH/24 Re-jigged the order of the tests in the default configuration so that the
2195 tests for valid domains and recipients precede the DNS black list and CSA
2196 tests, on the grounds that those ones are more expensive.
2198 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2199 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2200 This bug exists in every version of Exim that I still have, right back to
2203 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2204 However, an attempt to turn on case-sensitivity in a regex key by
2205 including (?-i) didn't work because the subject string was already
2206 lowercased, and the effects were non-intuitive. It turns out that a
2207 one-line patch can be used to allow (?-i) to work as expected.
2213 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2214 other effects) broke the use of negated acl sub-conditions.
2216 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2219 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2220 "Deny" causes Exim to reject the incoming connection with a 554 error.
2221 Unfortunately, if there is a major crisis, such as a disk failure,
2222 tcp-wrappers gives "deny", whereas what one would like would be some
2223 kind of temporary error. A kludge has been added to help with this.
2224 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2225 554 error is used if errno is still zero or contains ENOENT (which occurs
2226 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2229 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2231 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2232 errors. Otherwise a message that provokes a temporary error (when other
2233 messages do not) can cause a whole host to time out.
2235 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2236 addresses were routed directly to files or pipes from a redirect router.
2237 File deliveries just didn't batch; pipe deliveries might have suffered
2240 PH/06 A failure to get a lock for a hints database would erroneously always say
2241 "Failed to get write lock", even when it was really a read lock.
2243 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2244 of 0600. This has been changed to use the value of the lockfile_mode
2245 option (which defaults to 0600).
2247 PH/08 Applied small patch from the Sieve maintainer.
2249 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2250 folder from quota calculations, a direct delivery into this folder messed
2251 up the contents of the maildirsize file. This was because the regex was
2252 used only to exclude .Trash (or whatever) when the size of the mailbox
2253 was calculated. There was no check that a delivery was happening into an
2254 excluded directory. This bug has been fixed by ignoring all quota
2255 processing for deliveries into excluded directories.
2257 PH/10 Added the maildirfolder_create_regex option to appendfile.
2263 PH/01 The code for finding all the local interface addresses on a FreeBSD
2264 system running IPv6 was broken. This may well have applied to all BSD
2265 systems, as well as to others that have similar system calls. The broken
2266 code found IPv4 interfaces correctly, but gave incorrect values for the
2267 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2268 that it would not match correctly against @[] and not recognize the IPv6
2271 PH/02 The ipliteral router was not recognizing addresses of the form user@
2272 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2274 PH/03 Added disable_ipv6.
2276 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2277 decoded form, because it is most often used to construct To: headers
2278 lines in autoreplies, and the decoded form may well be syntactically
2279 invalid. However, $reply_address has leading white space removed, and all
2280 newlines turned into spaces so that the autoreply transport does not
2283 PH/05 If group was specified without a user on a router, and no group or user
2284 was specified on a transport, the group from the router was ignored.
2286 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2287 for visible compile-time settings that can be used to change these
2288 numbers, for those that want even more. Backwards compatibility with old
2289 spool files has been maintained. However, going back to a previous Exim
2290 release will lost any variables that are in spool files.
2292 PH/07 Two small changes when running in the test harness: increase delay when
2293 passing a TCP/IP connection to a new process, in case the original
2294 process has to generate a bounce, and remove special handling of
2295 127.0.0.2 (sic), which is no longer necessary.
2297 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2298 be the same on different OS.
2300 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2303 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2304 whether --show-vars was specified or not
2306 JJ/02 exipick: Added support for new ACL variable spool format introduced
2309 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2310 syntactically invalid From: or Reply-to: line, and a filter used this to
2311 generate an autoreply, and therefore failed to obtain an address for the
2312 autoreply, Exim could try to deliver to a non-existent relative file
2313 name, causing unrelated and misleading errors. What now happens is that
2314 it logs this as a hard delivery error, but does not attempt to create a
2317 PH/11 The exinext utility has a -C option for testing purposes, but although
2318 the given file was scanned by exinext itself; it wasn't being passed on
2319 when Exim was called.
2321 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2322 an end-of-file indication when reading a command response.
2324 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2325 compiled. In many other places in Exim, IPv6 addresses are always
2326 recognized, so I have changed this. It also means that IPv4 domain
2327 literals of the form [IPV4:n.n.n.n] are now always recognized.
2329 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2330 used if the router is not running as root, for example, when verifying at
2331 ACL time, or when using -bh. The debugging output from this situation was
2332 non-existent - all you got was a failure to exec. I have made two
2335 (a) Failures to set uid/gid, the current directory, or a process leader
2336 in a subprocess such as that created by queryprogram now generate
2337 suitable debugging ouput when -d is set.
2339 (b) The queryprogram router detects when it is not running as root,
2340 outputs suitable debugging information if -d is set, and then runs
2341 the subprocess without attempting to change uid/gid.
2343 PH/15 Minor change to Makefile for building test_host (undocumented testing
2346 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2347 additional section of a DNS packet that returns MX or SRV records.
2348 Instead, it always explicitly searches for A/AAAA records. This avoids
2349 major problems that occur when a DNS server includes only records of one
2350 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2351 fixed another bug: if SRV records were looked up and the corresponding
2352 address records were *not* found in the additional section, the port
2353 values from the SRV records were lost.
2355 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2356 using the correct key (the original address) when searching the retry
2357 rules in order to find which one to use for generating the retry hint.
2359 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2360 adding the default one. Similarly, if it contains a Reply-To: header, the
2361 errors_reply_to option, if set, is not used.
2363 PH/19 When calculating a retry time, Exim used to measure the "time since
2364 failure" by looking at the "first failed" field in the retry record. Now
2365 it does not use this if it is later than than the arrival time of the
2366 message. Instead it uses the arrival time. This makes for better
2367 behaviour in cases where some deliveries succeed, thus re-setting the
2368 "first failed" field. An example is a quota failure for a huge message
2369 when small messages continue to be delivered. Without this change, the
2370 "time since failure" will always be short, possible causing more frequent
2371 delivery attempts for the huge message than are intended.
2372 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2374 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2375 $1, $2, $3) because the numerical variables can be reset during some
2376 expansion items (e.g. "match"), thereby losing the authentication data.
2378 PH/21 Make -bV show the size of off_t variables so that the test suite can
2379 decide whether to run tests for quotas > 2G.
2381 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2382 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2383 filecount value is greater than 2G or if a quota value is greater than 2G
2384 on a system where the size of off_t is not greater than 4, a panic error
2387 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2388 never match. The debug and -bh output now contains an explicit error
2389 message indicating a malformed IPv4 address or mask.
2391 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2392 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2393 PH/23 above applies.
2395 PH/25 Do not write to syslog when running in the test harness. The only
2396 occasion when this arises is a failure to open the main or panic logs
2397 (for which there is an explicit test).
2399 PH/26 Added the /no_tell option to "control=freeze".
2401 PH/27 If a host name lookup failed very early in a connection, for example, if
2402 the IP address matched host_lookup and the reverse lookup yielded a name
2403 that did not have a forward lookup, an error message of the form "no IP
2404 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2405 could be logged. Now it outputs the IP address instead of "NULL".
2407 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2408 allows the sender and the authenticated sender to be set when
2409 submitting a message from within Exim. Since child_open_exim() is
2410 documented for local_scan(), the new function should be too.
2412 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2413 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2414 results in an empty string is now treated as unset.
2416 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2418 PH/31 Added sender_verify_fail logging option.
2420 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2421 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2422 tidied the source and removed it altogether.
2424 PH/33 When a queue run was abandoned because the load average was too high, a
2425 log line was always written; now it is written only if the queue_run log
2426 selector is set. In addition, the log line for abandonment now contains
2427 information about the queue run such as the pid. This is always present
2428 in "start" and "stop" lines but was omitted from the "abandon" line.
2430 PH/34 Omit spaces between a header name and the colon in the error message that
2431 is given when verify = headers_syntax fails (if there are lots of them,
2432 the message gets confusing).
2434 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2435 names, as there are now some PTR records that contain slashes. This check
2436 is only to protect against broken name servers that fall over on strange
2437 characters, so the fact that it applies to all lookups doesn't matter.
2439 PH/36 Now that the new test suite is complete, we can remove some of the
2440 special code in Exim that was needed for the old test suite. For example,
2441 sorting DNS records because real resolvers return them in an arbitrary
2442 order. The new test suite's fake resolver always returns records in the
2445 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2446 bounces) except when queue_only is set, to avoid logging races between
2447 the different processes.
2449 PH/38 Panic-die if .include specifies a non-absolute path.
2451 PH/39 A tweak to the "H" retry rule from its user.
2453 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2454 a label. They prevented compilation on older perls.
2456 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2457 a warning to be raised on newish perls.
2459 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2460 on queue. Changes to match documented behaviour of showing count of
2461 messages matching specified criteria.
2463 PH/40 Changed the default ident timeout from 30s to 5s.
2465 PH/41 Added support for the use of login_cap features, on those BSD systems
2466 that have them, for controlling the resources used by pipe deliveries.
2468 PH/42 The content-scanning code uses fopen() to create files in which to put
2469 message data. Previously it was not paying any attention to the mode of
2470 the files. Exim runs with umask(0) because the rest of the code creates
2471 files with open(), and sets the required mode explicitly. Thus, these
2472 files were ending up world-writeable. This was not a big issue, because,
2473 being within the spool directory, they were not world-accessible. I have
2474 created a function called modefopen, which takes an additional mode
2475 argument. It sets umask(777), creates the file, chmods it to the required
2476 mode, then resets the umask. All the relevant calls to fopen() in the
2477 content scanning code have been changed to use this function.
2479 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2480 to 24 hours. This avoids potential overflow problems when processing G
2481 and H retry rules. I suspect nobody ever tinkers with this value.
2483 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2485 PH/45 When the plaintext authenticator is running as a client, the server's
2486 challenges are checked to ensure they are valid base64 strings. By
2487 default, the authentication attempt is cancelled if an invalid string is
2488 received. Setting client_ignore_invalid_base64 true ignores these errors.
2489 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2490 they are received. Thus, the responses can be made to depend on the
2491 challenges. If an invalid string is ignored, an empty string is placed in
2494 PH/46 Messages that are created by the autoreply transport now contains a
2495 References: header, in accordance with RFCs 2822 and 3834.
2497 PH/47 Added authenticated_sender_force to the smtp transport.
2499 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2501 PH/49 Installed latest patch from the Sieve maintainer.
2503 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2504 was also set, the appendfile transport was unnecessarily scanning a
2505 directory of message files (e.g. for maildir delivery) to find the count
2506 of files (along with the size), even though it did not need this
2507 information. It now does the scan only if it needs to find either the
2508 size of the count of files.
2510 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2512 PH/52 Two bugs concerned with error handling when the smtp transport is
2515 (i) Exim was not creating retry information for temporary errors given
2516 for individual recipients after the DATA command when the smtp transport
2517 was used in LMTP mode. This meant that they could be retried too
2518 frequently, and not timed out correctly.
2520 (ii) Exim was setting the flag that allows error details to be returned
2521 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2522 recipients that were returned after the DATA command.
2524 PH/53 This is related to PH/52, but is more general: for any failing address,
2525 when detailed error information was permitted to be returned to the
2526 sender, but the error was temporary, then after the final timeout, only
2527 "retry timeout exceeded" was returned. Now it returns the full error as
2528 well as "retry timeout exceeded".
2530 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2531 do this, and (what is worse) MTAs that accept it.
2533 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2534 will now be deprecated.
2536 PH/56 New os.c-cygwin from the Cygwin maintainer.
2538 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2539 formats (previously only available in exim formats via -bpr, -bpru,
2540 and -bpra. Now also available in native and exiqgrep formats)
2542 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2543 with very large, slow to parse queues
2545 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2547 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2549 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2550 responses to authentication challenges, though it was showing the
2551 challenges; (ii) I've removed the CR characters from the debug output for
2554 PH/58 Allow for the insertion of a newline as well as a space when a string
2555 is turned into more than one encoded-word during RFC 2047 encoding. The
2556 Sieve code now uses this.
2558 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2559 data_4xx, lost_connection, tls_required.
2561 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2562 message was being sent as an SMTP response.
2564 PH/61 Add -l and -k options to exicyclog.
2566 PH/62 When verifying, if an address was redirected to one new address, so that
2567 verification continued, and the new address failed or deferred after
2568 having set something in $address_data, the value of $address_data was not
2569 passed back to the ACL. This was different to the case when no
2570 redirection occurred. The value is now passed back in both cases.
2572 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2573 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2574 use login_cap.h, so on its own it isn't the distinguishing feature. The
2575 new name refers directly to the setclassresources() function.
2577 PH/65 Added configuration files for NetBSD3.
2579 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2581 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2582 is preferred over IPv4.
2584 PH/68 The bounce_return_message and bounce_return_body options were not being
2585 honoured for bounces generated during the reception of non-SMTP messages.
2586 In particular, this applied to messages rejected by the ACL. This bug has
2587 been fixed. However, if bounce_return_message is true and bounce_return_
2588 body is false, the headers that are returned for a non-SMTP message
2589 include only those that have been read before the error was detected.
2590 (In the case of an ACL rejection, they have all been read.)
2592 PH/69 The HTML version of the specification is now built in a directory called
2593 spec_html instead of spec.html, because the latter looks like a path with
2594 a MIME-type, and this confuses some software.
2596 PH/70 Catch two compiler warnings in sieve.c.
2598 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2599 function verify_get_ident() calls ip_connect() to connect a socket, but
2600 if the "connect()" function timed out, ip_connect() used to close the
2601 socket. However, verify_get_ident() also closes the socket later, and in
2602 between Exim writes to the log, which may get opened at this point. When
2603 the socket was closed in ip_connect(), the log could get the same file
2604 descriptor number as the socket. This naturally causes chaos. The fix is
2605 not to close the socket in ip_connect(); the socket should be closed by
2606 the function that creates it. There was only one place in the code where
2607 this was missing, in the iplookup router, which I don't think anybody now
2608 uses, but I've fixed it anyway.
2610 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2611 well as to direct DNS lookups. Otherwise the handling of names in host
2612 lists is inconsistent and therefore confusing.
2618 PH/01 Two changes to the default runtime configuration:
2620 (1) Move the checks for relay_from_hosts and authenticated clients from
2621 after to before the (commented out) DNS black list checks.
2623 (2) Add control=submission to the relay_from_hosts and authenticated
2624 clients checks, on the grounds that messages accepted by these
2625 statements are most likely to be submissions.
2627 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2629 (1) Generate an error if the third argument for the ${prvs expansion is
2632 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2635 (3) Reset the variables that are obtained from the first argument of
2636 ${prvscheck and used in the second argument before leaving the code,
2637 because their memory is reclaimed, so using them afterwards may do
2640 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2641 one (it's much easier than Tom thought :-).
2643 (5) Because of (4), we can now allow for the use of $prvscheck_result
2644 inside the third argument.
2646 PH/03 For some reason, the default setting of PATH when running a command from
2647 a pipe transport was just "/usr/bin". I have changed it to
2650 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2651 anything to be listed in the output from -bV.
2653 PH/05 When a filter generated an autoreply, the entire To: header line was
2654 quoted in the delivery log line, like this:
2656 => >A.N.Other <ano@some.domain> <original@ddress> ...
2658 This has been changed so that it extracts the operative address. There
2659 may be more than one such address. If so, they are comma-separated, like
2662 => >ano@some.domain,ona@other.domain <original@ddress> ...
2664 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2665 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2666 not being looked up in the rDNS to get a host name, Exim was showing the
2667 IP address twice in Received: lines, even though the IP addresses were
2668 identical. For example:
2670 Received: from [1.2.3.4] (helo=[1.2.3.4])
2672 However, if the real host name was known, it was omitting the HELO data
2673 if it matched the actual IP address. This has been tidied up so that it
2674 doesn't show the same IP address twice.
2676 PH/07 When both +timestamp and +memory debugging was on, the value given by
2677 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2678 called by the debug printing, thereby overwriting the timestamp buffer.
2679 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2681 PH/08 When the original message was included in an autoreply transport, it
2682 always said "this is a copy of the message, including all the headers",
2683 even if body_only or headers_only was set. It now gives an appropriate
2686 PH/09 Applied a patch from the Sieve maintainer which:
2688 o fixes some comments
2689 o adds the (disabled) notify extension core
2690 o adds some debug output for the result of if/elsif tests
2691 o points to the current vacation draft in the documentation
2692 and documents the missing references header update
2696 o fixes a bug in processing the envelope test (when testing
2697 multiple envelope elements, the last element determinted the
2700 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2701 Electronic Mail") by including:
2703 Auto-submitted: auto-generated
2705 in the messages that it generates (bounce messages and others, such as
2706 warnings). In the case of bounce messages for non-SMTP mesages, there was
2707 also a typo: it was using "Auto_submitted" (underscore instead of
2708 hyphen). Since every message generated by Exim is necessarily in response
2709 to another message, thes have all been changed to:
2711 Auto-Submitted: auto-replied
2713 in accordance with these statements in the RFC:
2715 The auto-replied keyword:
2717 - SHOULD be used on messages sent in direct response to another
2718 message by an automatic process,
2720 - MUST NOT be used on manually-generated messages,
2722 - MAY be used on Delivery Status Notifications (DSNs) and Message
2723 Disposition Notifications (MDNs),
2725 - MUST NOT be used on messages generated by automatic or periodic
2726 processes, except for messages which are automatic responses to
2729 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2730 to the default Received: header definition.
2732 PH/12 Added log selector acl_warn_skipped (default on).
2734 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2735 variables because (a) they are in the wrong storage pool and (b) even if
2736 they were copied, it wouldn't work properly because of the caching.
2738 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2739 checking when decoding. Apparently there are clients that generate
2740 overlong encoded strings. Why am I not surprised?
2742 PH/15 If the first argument of "${if match_address" was not empty, but did not
2743 contain an "@" character, Exim crashed. Now it writes a panic log message
2744 and treats the condition as false.
2746 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2748 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2749 extension "envelope-auth". The code is finished and in agreement with
2750 other implementations, but there is no documentation so far and in fact,
2751 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2752 not changing the active code.
2754 Print executed "if" and "elsif" statements when debugging is used. This
2755 helps a great deal to understand what a filter does.
2757 Document more things not specified clearly in RFC3028. I had all this
2758 sorted out, when out of a sudden new issues came to my mind. Oops."
2760 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2763 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2764 canonical form (as documented). However, after a host name lookup from
2765 the IP address, check_host() was doing a simple string comparison with
2766 addresses acquired from the DNS when checking that the found name did
2767 have the original IP as one of its addresses. Since any found IPv6
2768 addresses are likely to be in abbreviated form, the comparison could
2769 fail. Luckily, there already exists a function for doing the comparison
2770 by converting both addresses to binary, so now that is used instead of
2771 the text comparison.
2773 PH/20 There was another similar case to PH/19, when a complete host name was
2774 given in a host list; looking up its IP address could give an abbreviated
2775 form, whereas the current host's name might or might not be abbreviated.
2776 The same fix has been applied.
2782 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2783 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2786 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2787 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2789 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2791 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2792 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2793 Also, for "Auto-Submitted", treat anything other than "no" as
2794 non-personal, in accordance with RFC 3834. (Previously it treated
2795 anything starting "auto-" as non-personal.)
2797 TF/01 The control=submission/name=... option had a problem with syntax
2798 errors if the name included a slash character. The /name= option
2799 now slurps the rest of the string, so it can include any characters
2800 but it must come last in the list of options (after /sender_retain
2803 PH/05 Some modifications to the interface to the fake nameserver for the new
2811 TK/01 Added the "success_on_redirect" address verification option. See
2812 NewStuff for rationale and an example.
2814 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2816 PH/02 Patch to exigrep to allow it to work on syslog lines.
2818 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2819 fread() to skip over the body file's header line, because in Cygwin the
2820 header line is locked and is inaccessible.
2822 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2823 co-exist for some time) to make it clear that it is the Exim ID that is
2824 referenced, not the Message-ID: header line.
2826 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2827 string_format() function, because snprintf() does not exist on all
2830 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2831 ${stat: expansion item.
2833 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2834 protocol synchronization error", to keep the pedants happy.
2836 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2837 well as for IRIX systems, when gcc is being used. See the host.c source
2840 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2842 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2845 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2846 transport and to the smtp transport in LMTP mode.
2848 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2850 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2851 run a filter in a subprocess. This could lead to confusion in subsequent
2852 lookups in the parent process. There should also be a search_tidyup() at
2853 the end of the subprocess.
2855 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2856 only if the host matched helo_try_verify_hosts, which caused the
2857 verification to occur when the EHLO/HELO command was issued. The ACL just
2858 tested the remembered result. Now, if a previous verification attempt has
2859 not happened, "verify = helo" does it there and then.
2861 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2863 TK/03 Fix log output including CR from clamd.
2865 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2866 exist provoked a memory error which could cause a segfault.
2868 PH/15 Installed PCRE 6.2
2870 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2872 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2873 of the problem. Specifically, suggested +O2 rather than +O1 for the
2876 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2878 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2879 record, the port was not correctly logged when the outgoing_port log
2880 selector was set (it logged the transort's default port).
2882 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2883 fallback_hosts, and "hosts" in the smtp transport.
2885 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2886 host errors such as "Connection refused".
2888 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2889 authentication with radiusclient 0.4.9:
2891 - Error returned from rc_read_config was caught wrongly
2892 - Username/password not passed on to radius server due to wrong length.
2894 The presumption is that some radiusclient API changes for 4.51/PH/17
2895 were not taken care of correctly. The code is still untested by me (my
2896 Linux distribution still has 0.3.2 of radiusclient), but it was
2897 contributed by a Radius user.
2899 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2900 expanding the "port" option of the smtp transport.
2902 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2903 while reading a MIME header. Thanks to Tom Hughes for a patch.
2905 PH/24 Include config.h inside local_scan.h so that configuration settings are
2908 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2909 that in an ACL for RCPT (for example), you can examine exactly what was
2912 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2913 commands, but it was not correctly comparing the address with the actual
2914 client host address. Thus, it would show the EHLO address in Received:
2915 header lines when this was not necessary.
2917 PH/27 Added the % operator to ${eval:}.
2919 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2920 it should be ignoring failures (because with -C, for example, it has lost
2921 privilege). It wasn't ignoring creation failures other than "already
2924 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2927 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2928 because an input line was too long, either on its own, or by virtue of
2929 too many continuations, the temporary file was not being removed, and the
2930 return code was incorrect.
2932 PH/31 Missing "BOOL" in function definition in filtertest.c.
2934 PH/32 Applied Sieve patches from the maintainer.
2936 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2938 PH/33 Added "verify = not_blind".
2940 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2941 Local/Makefile (with some defaults set). These are used in built scripts
2942 such as exicyclog, but they have never been used in the exim_install
2943 script (though there are many overriding facilities there). I have
2944 arranged that the exim_install script now takes note of these two
2947 PH/35 Installed configuration files for Dragonfly.
2949 PH/36 When a locally submitted message by a trusted user did not contain a
2950 From: header, and the sender address was obtained from -f or from an SMTP
2951 MAIL command, and the trusted user did not use -F to supply a sender
2952 name, $originator_name was incorrectly used when constructing a From:
2953 header. Furthermore, $originator_name was used for submission mode
2954 messages from external hosts without From: headers in a similar way,
2955 which is clearly wrong.
2957 PH/37 Added control=suppress_local_fixups.
2959 PH/38 When log_selector = +received_sender was set, and the addition of the
2960 sender made the log line's construction buffer exactly full, or one byte
2961 less than full, an overflow happened when the terminating "\n" was
2964 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2965 when the result of a list match is failure because a DNS lookup failed.
2967 PH/40 RM_COMMAND is now used in the building process.
2969 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2970 the "build-* directories that it finds.
2972 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2973 address in a domain literal was a prefix of an interface address.
2975 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2976 when verifying a sender address, unless rewrite_headers is false.
2978 PH/44 Wrote a long comment about why errors_to addresses are verified as
2979 recipients, not senders.
2981 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2982 the ratelimit ACL was added.
2984 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2986 PH/46 Added extra information about PostgreSQL errors to the error string.
2988 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2989 avoiding the need to install special zones in a real server. This is
2990 backwards compatible; if it can't find the fake resolver, it drops back.
2991 Thus, both old and new test suites can be run.
2993 TF/02 Added util/ratelimit.pl
2995 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2996 clock is set back in time.
2998 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2999 Candler <B.Candler@pobox.com>.
3001 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3002 OK for addresses that are the result of redirections.
3004 PH/48 A number of further additions for the benefit of the new test suite,
3005 including a fake gethostbyname() that interfaces to the fake DNS resolver
3008 TF/06 The fix for widen_domains has also been applied to qualify_single and
3009 search_parents which are the other dnslookup options that can cause
3012 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3015 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3017 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3018 service" if it encountered an MX record with an empty target hostname.
3019 The message is now "an MX or SRV record indicated no SMTP service".
3021 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3022 if the DNS of the sending site is misconfigured. This is quite a
3023 common situation. This change restores the behaviour of treating a
3024 helo verification defer as a failure.
3026 PH/51 If self=fail was set on a router, the bounce message did not include the
3027 actual error message.
3033 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3035 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3036 command itself ended in error, the underlying message about the transport
3037 filter timeout was being overwritten with the pipe command error. Now the
3038 underlying error message should be appended to the second error message.
3040 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3042 PH/02 Reduce the amount of output that "make" produces by default. Full output
3043 can still be requested.
3045 PH/03 The warning log line about a condition test deferring for a "warn" verb
3046 was being output only once per connection, rather than after each
3047 occurrence (because it was using the same function as for successful
3048 "warn" verbs). This seems wrong, so I have changed it.
3050 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3051 it should not have, which might have caused a crash in the right
3052 circumstances, but probably never did.
3054 PH/04 Installed a modified version of Tony Finch's patch to make submission
3055 mode fix the return path as well as the Sender: header line, and to
3056 add a /name= option so that you can make the user's friendly name appear
3059 TF/03 Added the control = fakedefer ACL modifier.
3061 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3062 Mark Lowes for thorough testing.
3064 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3066 TK/03 Merged latest SRS patch from Miles Wilton.
3068 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3069 with the definition in sysexits.h (which is #included earlier).
3070 Fortunately, Exim does not actually use EX_OK. The code used to try to
3071 preserve the sysexits.h value, by assumimg that macro definitions were
3072 scanned for macro replacements. I have been disabused of this notion,
3073 so now the code just undefines EX_OK before #including unistd.h.
3075 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3076 in the smtp transport. When a block could not be written in a single
3077 write() function, the timeout was being re-applied to each part-write.
3078 This seems wrong - if the receiver was accepting one byte at a time it
3079 would take for ever. The timeout is now adjusted when this happens. It
3080 doesn't have to be particularly precise.
3082 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3083 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3085 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3086 without a domain if the check to <postmaster@domain> fails.
3088 SC/01 Eximstats: added -xls and the ability to specify output files
3089 (patch written by Frank Heydlauf).
3091 SC/02 Eximstats: use FileHandles for outputing results.
3093 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3095 SC/04 Eximstats: fixed display of large numbers with -nvr option
3097 SC/05 Eximstats: fixed merging of reports with empty tables.
3099 SC/06 Eximstats: added the -include_original_destination flag
3101 SC/07 Eximstats: removed tabs and trailing whitespace.
3103 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3105 TK/06 MBOX spool code: Add real "From " MBOX separator line
3106 so the .eml file is really in mbox format (even though
3107 most programs do not really care). Patch from Alex Miller.
3109 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3110 The latter is generated from $received_to and is only set if the
3111 message has one envelope recipient. SA can use these headers,
3112 obviously out-of-the-box. Patch from Alex Miller.
3114 PH/08 The ${def test on a variable was returning false if the variable's
3115 value was "0", contrary to what the specification has always said!
3116 The result should be true unless the variable is empty.
3118 PH/09 The syntax error of a character other than { following "${if
3119 def:variable_name" (after optional whitespace) was not being diagnosed.
3120 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3121 accidental colon was present, for example, could give incorrect results.
3123 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3124 result is used (not including appendfile, where other changes are about
3127 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3128 This involved changing a lot of size variables from int to off_t. It
3129 should work with maildirs and everything.
3131 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3132 spamd dying while we are connected to it.
3134 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3137 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3138 function for BDB 4.3.
3140 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3142 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3143 the sources for PCRE have been re-arranged and more files are now
3146 PH/15 The code I had for printing potentially long long variables in PH/11
3147 above was not the best (it lost precision). The length of off_t variables
3148 is now inspected at build time, and an appropriate printing format (%ld
3149 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3150 to be "long long int" or "long int". This is needed for the internal
3151 formatting function string_vformat().
3153 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3154 the configuration file to be ":syslog", then the script "guesses" where
3155 the logs files are, rather than using the compiled in default. In our
3156 case the guess is not the same as the compiled default, so the script
3157 suddenly stopped working when I started to use syslog. The patch checks
3158 to see if log_file_path is "". If so, it attempts to read it from exim
3159 with no configuration file to get the compiled in version, before it
3160 falls back to the previous guessing code."
3162 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3163 implementing BATV in an Exim configuration. See NewStuff for the gory
3166 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3167 Makefile that are specific to HP-UX.
3169 PH/18 If the "use_postmaster" option was set for a recipient callout together
3170 with the "random" option, the postmaster address was used as the MAIL
3171 FROM address for the random test, but not for the subsequent recipient
3172 test. It is now used for both.
3174 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3175 patch removes a few documentation additions to RFC 3028, because the
3176 latest draft now contains them. It adds the new en;ascii-case comparator
3177 and a new error check for 8bit text in MIME parts. Comparator and
3178 require names are now matched exactly. I enabled the subaddress
3179 extension, but it is not well tested yet (read: it works for me)."
3181 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3182 rework some of the code of TK/09 above to avoid the hardwired use of
3183 "%lld" and "long long". Replaced the call to snprintf() with a call to
3186 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3187 records point to non-existent hosts", "retry timeout exceeded", and
3188 "retry time not reached for any host after a long failure period".
3190 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3191 experimental DomainKeys support:
3193 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3194 (2) On an error such as an illegally used "control", the wrong name for
3195 the control was given.
3197 These problems did NOT occur unless DomainKeys support was compiled.
3199 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3201 PH/24 Added ${if match_ip condition.
3203 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3204 databases so that it will be absolutely obvious if a crash occurs in the
3205 DB library. This is a regular occurrence (often caused by mis-matched
3208 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3209 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3210 picked up on a user's system that detects such things. There doesn't seem
3211 to be a gcc warning option for this - only an attribute that has to be
3212 put on the function's prototype. It seems that in Fedora Core 4 they have
3213 set this on a number of new functions. No doubt there will be more in due
3216 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3217 specify a transport. However, if an address that was verified by such a
3218 router was the subject of a callout, Exim crashed because it tried to
3219 read the rcpt_include_affixes from the non-existent transport. Now it
3220 just assumes that the setting of that option is false. This bug was
3221 introduced by 4.51/PH/31.
3223 PH/28 Changed -d+all to exclude +memory, because that information is very
3224 rarely of interest, but it makes the output a lot bigger. People tend to
3225 do -d+all out of habit.
3227 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3228 code in os-type was giving problems when libc.so lives in lib64, like on
3231 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3232 aren't the modern standard, and it seems that some systems' include files
3233 don't always have them. Exim was already checking for some of the newer
3234 ones like T_AAAA, and defining it itself. I've added checks for all the
3235 record types that Exim uses.
3237 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3238 not automatically generating a new one, as it is supposed to. This
3239 prevented TLS from working. If the file did exist, but contained invalid
3240 data, a new version was generated, as expected. It was only the case of a
3241 non-existent file that was broken.
3243 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3244 with a change in libdomainkeys > 0.64.
3246 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3247 from DNS. If the selector record carries the flag, it now has
3248 precedence over the domain-wide flag.
3250 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3252 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3253 the use of a port name that isn't defined in /etc/services) occurred, the
3254 message was deferred as in a normal delivery, and thus remained on the
3255 spool, instead of being failed because of the mua_wrapper setting. This
3256 is now fixed, and I tidied up some of the mua_wrapper messages at the
3259 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3260 the messages in a hash of arrays rather than using individual hashes.
3261 This is a bit cleaner and results in dramatic memory savings, albeit
3262 at a slight CPU cost.
3264 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3265 as requested by Marc Sherman.
3267 SC/10 Eximstats: added histograms for user specified patterns as requested
3270 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3272 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3273 fopen() in the content-scanning modules that did not already have it.
3279 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3280 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3282 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3284 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3286 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3287 to test Sieve filters that use "vacation".
3289 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3290 that changes the way the GnuTLS parameters are stored in the cache file.
3291 The new format can be generated externally. For backward compatibility,
3292 if the data in the cache doesn't make sense, Exim assumes it has read an
3293 old-format file, and it generates new data and writes a new file. This
3294 means that you can't go back to an older release without removing the
3297 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3298 work if there are any delivery delays because "one_time" forces the
3299 parent to be marked "delivered", so its unseen clone is never tried
3300 again. For this reason, Exim now forbids the simultaneous setting of
3303 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3304 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3305 entry if you want to know the details. The fix, however, neglected to
3306 consider the case when local delivery batching is involved. The test for
3307 "previously delivered" was not happening when checking to see if an
3308 address could be batched with a previous (undelivered) one; under
3309 certain circumstances this could lead to multiple deliveries to the same
3312 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3313 in its include files, and this causes problems building Exim.
3315 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3316 header_syntax) but Exim was just ignoring anything given after a slash.
3317 In particular, this caused confusion with an attempt to use "verify =
3318 reverse_host_lookup/defer_ok". An error is now given when options are
3319 supplied for verify items that do not have them. (Maybe reverse_host_
3320 lookup should have a defer_ok option, but that's a different point.)
3322 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3323 defined by RFC 821) to 2048, because there were problems with some AUTH
3324 commands, and RFC 1869 says the size should be increased for extended
3325 SMTP commands that take arguments.
3327 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3330 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3331 "unknown" error; now it says that the functionality isn't in the binary.
3333 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3334 an address' error message when a string expansion fails (syntax or
3335 whatever). Otherwise the password may appear in the log. Following change
3336 PH/42 below, there is no longer a chance of it appearing in a bounce
3339 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3341 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3342 of its addresses was ever tried. (Bugzilla bug #2.)
3344 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3345 the result incorrectly in the debug output. (It correctly added a newline
3346 to what was transported.)
3348 TF/01 Added $received_time.
3350 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3351 commented out examples of how to interface to a virus scanner and to
3352 SpamAssassin. Also added commented examples of av_scanner and
3353 spamd_address settings.
3355 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3356 and controls are allowed in which ACLs. There were a couple of minor
3357 errors. Some of the entries in the conditions table (which is a table of
3358 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3359 negation of where the condition IS allowed.
3361 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3363 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3364 header file does not have a version number, so I've had to invent a new
3365 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3366 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3367 radiusclient), but it was contributed by a Radius user.
3369 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3370 files or directories, for OpenSSL.
3372 PH/19 When an Exim process that is running as root has to create an Exim log
3373 file, it does so in a subprocess that runs as exim:exim so as to get the
3374 ownership right at creation (otherwise, other Exim processes might see
3375 the file with the wrong ownership). There was no test for failure of this
3376 fork() call, which would lead to the process getting stuck as it waited
3377 for a non-existent subprocess. Forks do occasionally fail when resources
3378 run out. I reviewed all the other calls to fork(); they all seem to check
3381 PH/20 When checking for unexpected SMTP input at connect time (before writing
3382 the banner), Exim was not dealing correctly with a non-positive return
3383 from the read() function. If the client had disconnected by this time,
3384 the result was a log entry for a synchronization error with an empty
3385 string after "input=" when read() returned zero. If read() returned -1
3386 (an event I could not check), uninitialized data bytes were printed.
3387 There were reports of junk text (parts of files, etc) appearing after
3390 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3392 PH/22 Added support for macro redefinition, and (re)definition in between
3393 driver and ACL definitions.
3395 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3396 forgetting to use the resulting value; it was using the unexpanded value.
3398 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3399 hadn't been configured. The fix is from Juergen Kreileder, who
3400 understands it better than I do:
3402 "Here's what I see happening with three configured cyrus_sasl
3403 authenticators configured (plain, login, cram-md5):
3405 On startup auth_cyrus_sasl_init() gets called for each of these.
3406 This means three calls to sasl_listmech() without a specified mech_list.
3407 => SASL tests which mechs of all available mechs actually work
3408 => three warnings about OTP not working
3409 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3411 With the patch, sasl_listmech() also gets called three times. But now
3412 SASL's mech_list option is set to the server_mech specified in the the
3413 authenticator. Or in other words, the answer from sasl_listmech()
3414 gets limited to just the mech you're testing for (which is different
3416 => the return list contains just 'plain' or 'login', 'cram-md5' or
3417 nothing depending on the value of ob->server_mech.
3419 I've just tested the patch: Authentication still works fine,
3420 unavailable mechs specified in the exim configuration are still
3421 caught, and the auth.log warnings about OTP are gone."
3423 PH/25 When debugging is enabled, the contents of the command line are added
3424 to the debugging output, even when log_selector=+arguments is not
3427 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3428 answer is "GNU", and only if the return is "GNU/something" is the answer
3431 PH/27 $acl_verify_message is now set immediately after the failure of a
3432 verification in an ACL, and so is available in subsequent modifiers. In
3433 particular, the message can be preserved by coding like this:
3435 warn !verify = sender
3436 set acl_m0 = $acl_verify_message
3438 Previously, $acl_verify_message was set only while expanding "message"
3439 and "log_message" when a very denied access.
3441 PH/28 Modified OS/os.c-Linux with
3443 -#ifndef OS_LOAD_AVERAGE
3444 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3446 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3447 nomenclature these days.)
3449 PH/29 Installed patch from the Sieve maintainer that adds the options
3450 sieve_useraddress and sieve_subaddress to the redirect router.
3452 PH/30 In these circumstances:
3453 . Two addresses routed to the same list of hosts;
3454 . First host does not offer TLS;
3455 . First host accepts first address;
3456 . First host gives temporary error to second address;
3457 . Second host offers TLS and a TLS session is established;
3458 . Second host accepts second address.
3459 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3460 and peerdn, if requested) that were in fact used only for the second
3463 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3464 attention to any local part prefix or suffix that was matched by the
3465 router that accepted the address. It now behaves in the same way as it
3466 does for delivery: the affixes are removed from the local part unless
3467 rcpt_include_affixes is set on the transport.
3469 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3470 timeout during the DATA phase of an incoming message.
3472 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3473 applied a patch sanctioned by the Sieve maintainer.
3475 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3476 the uid or gid is negative. A case of a negative gid caused this to be
3477 noticed. The fix allows for either to be negative.
3479 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3480 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3483 PH/36 The change PH/12 above was broken. Fixed it.
3485 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3486 the grounds that routing the same address twice would always produce the
3487 same answer. This might have been true once, but it is certainly no
3488 longer true now. Routing a child address may depend on the previous
3489 routing that produced that child. Some complicated redirection strategies
3490 went wrong when messages had multiple recipients, and made Exim's
3491 behaviour dependent on the order in which the addresses were given.
3493 I have moved the duplicate checking until after the routing is complete.
3494 Exim scans the addresses that are assigned to local and remote
3495 transports, and removes any duplicates. This means that more work will be
3496 done, as duplicates will always all be routed, but duplicates are
3497 presumably rare, so I don't expect this is of any significance.
3499 For deliveries to pipes, files, and autoreplies, the duplicate checking
3500 still happens during the routing process, since they are not going to be
3503 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3504 It corrects a timeout issue with spamd. This is Ian's comment: "The
3505 background is that sometimes spamd either never reads data from a
3506 connection it has accepted, or it never writes response data. The exiscan
3507 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3508 blindly assumes that writes won't block so it may never time out."
3510 PH/39 Allow G after quota size as well as K and M.
3512 PH/40 The value set for $authenticated_id in an authenticator may not contain
3513 binary zeroes or newlines because the value is written to log lines and
3514 to spool files. There was no check on this. Now the value is run through
3515 the string_printing() function so that such characters are converted to
3516 printable escape sequences.
3518 PH/41 $message_linecount is a new variable that contains the total number of
3519 lines in the message. Compare $body_linecount, which is the count for the
3522 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3523 bounce and delay warning messages, except in certain special cases, which
3526 (a) An SMTP error message from a remote host;
3527 (b) A message specified in a :fail: redirection;
3528 (c) A message specified in a "fail" command in a system filter;
3529 (d) A message specified in a FAIL return from the queryprogram router;
3530 (e) A message specified by the cannot_route_message router option.
3532 In these cases only, Exim does include the error details in bounce and
3533 warning messages. There are also a few cases where bland messages such
3534 as "unrouteable address" or "local delivery error" are given.
3536 PH/43 $value is now also set for the "else" part of a ${run expansion.
3538 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3539 being worked on, but at least Exim now implements the latest version to
3542 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3543 process to complete was treated as a delivery failure, a timeout while
3544 writing the message to the pipe was logged, but erroneously treated as a
3545 successful delivery. Such timeouts include transport filter timeouts. For
3546 consistency with the overall process timeout, these timeouts are now
3547 treated as errors, giving rise to delivery failures by default. However,
3548 there is now a new Boolean option for the pipe transport called
3549 timeout_defer, which, if set TRUE, converts the failures into defers for
3550 both kinds of timeout. A transport filter timeout is now identified in
3553 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3554 systems where "make" and "gmake" are different, calling "gmake" at top
3555 level broke things. I've arranged for the value of $(MAKE) to be passed
3556 from the Makefile to this script so that it can call the same version of
3560 A note about Exim versions 4.44 and 4.50
3561 ----------------------------------------
3563 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3564 changes of various kinds. As a consequence, a big documentation update was
3565 needed. This delayed the release for rather longer than seemed good, especially
3566 in the light of a couple of (minor) security issues. Therefore, the changes
3567 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3568 release. So 4.44 and 4.50 are in effect two different branches that both start
3571 I have left the 4.50 change log unchanged; it contains all the changes since
3572 4.43. The change log for 4.44 is below; many of its items are identical to
3573 those for 4.50. This seems to be the most sensible way to preserve the
3574 historical information.
3580 1. Minor wording change to the doc/README.SIEVE file.
3582 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3583 computation of the current number of files was incorrect.
3585 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3586 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3589 4. Give more explanation in the error message when the command for a transport
3590 filter fails to execute.
3592 5. There are several places where Exim runs a non-Exim command in a
3593 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3594 was being done only for the command run by the queryprogram router. It is
3595 now done for all such subprocesses. The other cases are: ${run, transport
3596 filters, and the commands run by the lmtp and pipe transports.
3598 6. Added CONFIGURE_GROUP build-time option.
3600 7. Some older OS have a limit of 256 on the maximum number of file
3601 descriptors. Exim was using setrlimit() to set 1000 as a large value
3602 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3603 systems. I've change it so that if it can't get 1000, it tries for 256.
3605 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3606 was an oversight, and furthermore, ever since the addition of extra
3607 controls (e.g. 4.43/32), the checks on when to allow different forms of
3608 "control" were broken. There should now be diagnostics for all cases when a
3609 control that does not make sense is encountered.
3611 9. Added the /retain_sender option to "control=submission".
3613 10. $recipients is now available in the predata ACL (oversight).
3615 11. Tidy the search cache before the fork to do a delivery from a message
3616 received from the command line. Otherwise the child will trigger a lookup
3617 failure and thereby defer the delivery if it tries to use (for example) a
3618 cached ldap connection that the parent has called unbind on.
3620 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3621 of $address_data from the recipient verification was clobbered by the
3622 sender verification.
3624 13. The value of address_data from a sender verification is now available in
3625 $sender_address_data in subsequent conditions in the ACL statement.
3627 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3629 15. Added a new option "connect=<time>" to callout options, to set a different
3632 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3633 was its contents. (It was OK if the option was not defined at all.)
3635 17. A "Completed" log line is now written for messages that are removed from
3636 the spool by the -Mrm option.
3638 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3639 information about exactly what failed.
3641 19. Added -dd to debug only the daemon process.
3643 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3644 handles timeouts, both on the server side and network timeouts. Renamed the
3645 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3647 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3648 It is now set to "smtps".
3650 22. $host_address is now set to the target address during the checking of
3651 ignore_target_hosts.
3653 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3654 being passed; this would have caused $sender_host_name to have been used if
3655 matching the list had actually called for a host name (not very likely,
3656 since this list is usually IP addresses). A host name is now passed as
3659 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3660 code that specifies a non-restarting handler (typically sigaction() in
3661 modern systems) in an attempt to fix a rare and obscure crash bug.
3663 25. Narrowed the window for a race in the daemon that could cause it to ignore
3664 SIGCHLD signals. This is not a major problem, because they are used only to
3665 wake it up if nothing else does.
3667 26. A malformed maildirsize file could cause Exim to calculate negative values
3668 for the mailbox size or file count. Odd effects could occur as a result.
3669 The maildirsize information is now recalculated if the size or filecount
3672 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3673 support for a long time. Removed HAVE_SYS_VFS_H.
3675 28. Installed the latest version of exipick from John Jetmore.
3677 29. In an address list, if the pattern was not a regular expression, an empty
3678 subject address (from a bounce message) matched only if the pattern was an
3679 empty string. Non-empty patterns were not even tested. This was the wrong
3680 because it is perfectly reasonable to use an empty address as part of a
3681 database query. An empty address is now tested by patterns that are
3682 lookups. However, all the other forms of pattern expect the subject to
3683 contain a local part and a domain, and therefore, for them, an empty
3684 address still always fails if the pattern is not itself empty.
3686 30. Exim went into a mad DNS loop when attempting to do a callout where the
3687 host was specified on an smtp transport, and looking it up yielded more
3688 than one IP address.
3690 31. Re-factored the code for checking spool and log partition space into a
3691 function that finds that data and another that does the check. The former
3692 is then used to implement four new variables: $spool_space, $log_space,
3693 $spool_inodes, and $log_inodes.
3695 32. The RFC2047 encoding function was originally intended for short strings
3696 such as real names; it was not keeping to the 75-character limit for
3697 encoded words that the RFC imposes. It now respects the limit, and
3698 generates multiple encoded words if necessary. To be on the safe side, I
3699 have increased the buffer size for the ${rfc2047: expansion operator from
3702 33. It is now permitted to omit both strings after an "if" condition; if the
3703 condition is true, the result is "true". As before, when the second string
3704 is omitted, a false condition yields an empty string. This makes it less
3705 cumbersome to write custom ACL and router conditions.
3707 34. Failure to deliver a bounce message always caused it to be frozen, even if
3708 there was an errors_to setting on the router. The errors_to setting is now
3711 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3712 canonical form (fully expanded) before being placed in
3713 $sender_host_address.
3715 36. The table in the code that translates DNS record types into text (T_A to
3716 "A" for instance) was missing entries for NS and CNAME. It is just possible
3717 that this could have caused confusion if both these types were looked up
3718 for the same domain, because the text type is used as part of Exim's
3719 per-process caching. But the chance of anyone hitting this buglet seems
3722 37. The dnsdb lookup has been extended in a number of ways.
3724 (1) There is a new type, "zns", which walks up the domain tree until it
3725 finds some nameserver records. It should be used with care.
3727 (2) There is a new type, "mxh", which is like "mx" except that it returns
3728 just the host names, not the priorities.
3730 (3) It is now possible to give a list of domains (or IP addresses) to be
3731 looked up. The behaviour when one of the lookups defers can be
3732 controlled by a keyword.
3734 (4) It is now possible to specify the separator character for use when
3735 multiple records are returned.
3737 38. The dnslists ACL condition has been extended: it is now possible to supply
3738 a list of IP addresses and/or domains to be looked up in a particular DNS
3741 39. Added log_selector=+queue_time_overall.
3743 40. When running the queue in the test harness, wait just a tad after forking a
3744 delivery process, to get repeatability of debugging output.
3746 41. Include certificate and key file names in error message when GnuTLS fails
3747 to set them up, because the GnuTLS error message doesn't include the name
3748 of the failing file when there is a problem reading it.
3750 42. Allow both -bf and -bF in the same test run.
3752 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3754 44. The "Exiscan patch" is now merged into the mainline Exim source.
3756 45. Sometimes the final signoff response after QUIT could fail to get
3757 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3758 < 0 before doing a fflush(). This bug looks as though it goes back to the
3759 introduction of TLS in release 3.20, but "sometimes" must have been rare
3760 because the tests only now provoked it.
3762 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3763 (this can affect the format of dates).
3765 47. exim_tidydb, when checking for the continued existence of a message for
3766 which it has found a message-specific retry record, was not finding
3767 messages that were in split spool directories. Consequently, it was
3768 deleting retry records that should have stayed in existence.
3770 48. Steve fixed some bugs in eximstats.
3772 49. The SPA authentication driver was not abandoning authentication and moving
3773 on to the next authenticator when an expansion was forced to fail,
3774 contradicting the general specification for all authenticators. Instead it
3775 was generating a temporary error. It now behaves as specified.
3777 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3778 (the order specifies the preference for clients). The order is now AES256,
3779 AES128, 3DES, ARCFOUR128.
3781 51. Small patch to Sieve code - explicitly set From: when generating an
3784 52. Exim crashed if a remote delivery caused a very long error message to be
3785 recorded - for instance if somebody sent an entire SpamAssassin report back
3786 as a large number of 550 error lines. This bug was coincidentally fixed by
3787 increasing the size of one of Exim's internal buffers (big_buffer) that
3788 happened as part of the Exiscan merge. However, to be on the safe side, I
3789 have made the code more robust (and fixed the comments that describe what
3792 53. Now that there can be additional text after "Completed" in log lines (if
3793 the queue_time_overall log selector is set), a one-byte patch to exigrep
3794 was needed to allow it to recognize "Completed" as not the last thing in
3797 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3798 patch that reportedly fixes this has been added. I am not expert enough to
3799 create a test for it. This is what the patch creator wrote:
3801 "I found a little strange behaviour of ldap code when working with
3802 Windows 2003 AD Domain, where users was placed in more than one
3803 Organization Units. When I tried to give exim partial DN, the exit code
3804 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3805 But simultaneously result of request was absolutely normal ldap result,
3806 so I produce this patch..."
3808 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3809 REFERENCE, so I have modified the code to exclude the patch when that macro
3812 55. Some experimental protocols are using DNS PTR records for new purposes. The
3813 keys for these records are domain names, not reversed IP addresses. The
3814 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3815 leaves it alone. Component reversal etc. now happens only for IP addresses.
3818 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3820 57. Double the size of the debug message buffer (to 2048) so that more of very
3821 long debug lines gets shown.
3823 58. The exicyclog utility now does better if the number of log files to keep
3824 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3826 59. Two changes related to the smtp_active_hostname option:
3828 (1) $smtp_active_hostname is now available as a variable.
3829 (2) The default for smtp_banner uses $smtp_active_hostname instead
3830 of $primary_hostname.
3832 60. The host_aton() function is supposed to be passed a string that is known
3833 to be a valid IP address. However, in the case of IPv6 addresses, it was
3834 not checking this. This is a hostage to fortune. Exim now panics and dies
3835 if the condition is not met. A case was found where this could be provoked
3836 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3837 components; fortuitously, this particular loophole had already been fixed
3838 by change 4.50/55 above.
3840 If there are any other similar loopholes, the new check in host_aton()
3841 itself should stop them being exploited. The report I received stated that
3842 data on the command line could provoke the exploit when Exim was running as
3843 exim, but did not say which command line option was involved. All I could
3844 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3845 running as the user.
3848 61. There was a buffer overflow vulnerability in the SPA authentication code
3849 (which came originally from the Samba project). I have added a test to the
3850 spa_base64_to_bits() function which I hope fixes it.
3853 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3854 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3856 63. The daemon start-up calls getloadavg() while still root for those OS that
3857 need the first call to be done as root, but it missed one case: when
3858 deliver_queue_load_max is set with deliver_drop_privilege. This is
3859 necessary for the benefit of the queue runner, because there is no re-exec
3860 when deliver_drop_privilege is set.
3862 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3863 This has been fixed.
3865 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3866 was in use, was not putting the data itself into the right store pool;
3867 consequently, it could be overwritten for a subsequent message in the same
3868 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3871 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3873 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3874 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3875 the function were treating the return as a boolean value, which happened to
3876 work because 0=false and not-0=true, but is not correct code.
3878 68. The host_aton() function was not handling scoped IPv6 addresses (those
3879 with, for example, "%eth0" on the end) correctly.
3881 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3882 negated items (that is, ~something) in unsigned ints. Some compilers
3883 apparently mutter when there is no cast.
3885 70. If an address verification called from an ACL failed, and did not produce a
3886 user-specific message (i.e. there was only a "system" message), nothing was
3887 put in $acl_verify_message. In this situation, it now puts the system
3890 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3891 session; change 4.31/43 added the unwanted input to the log line - except
3892 that it did not do this in the start of session case. It now does.
3894 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3895 This is wrong and can cause the other end to generate a synchronization
3896 error if it is another Exim or anything else that does the synchronization
3897 check. A QUIT command is no longer sent after a timeout.
3899 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3900 during host lookups.
3902 74. The defer_ok option of callout verification was not working if it was used
3903 when verifying addresses in header lines, that is, for this case:
3905 verify = header_sender/callout=defer_ok
3907 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3908 those file descriptors could be used for SMTP connections. If anything
3909 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3910 could be sent to the SMTP client, causing chaos. The daemon now opens
3911 stdin, stdout, and stderr to /dev/null when it puts itself into the
3914 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3915 log by default. The user can override this with suitable Perl magic.
3917 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3918 the log message when discard triggers, was not working for the DATA ACL or
3919 for the non-SMTP ACL.
3921 78. Error message wording change in sieve.c.
3923 79. If smtp_accept_max_per_host was set, the number of connections could be
3924 restricted to fewer than expected, because the daemon was trying to set up
3925 a new connection before checking whether the processes handling previous
3926 connections had finished. The check for completed processes is now done
3927 earlier. On busy systems, this bug wouldn't be noticed because something
3928 else would have woken the daemon, and it would have reaped the completed
3931 80. If a message was submitted locally by a user whose login name contained one
3932 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3933 It caused a spool format error. I have fixed the spool reading code. A
3934 related problem was that the "from" clause in the Received: line became
3935 illegal because of the space(s). It is now covered by ${quote_local_part.
3937 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3940 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3942 83. Patch from Sieve maintainer to fix unterminated string problem in
3943 "vacation" handling.
3945 84. Some minor changes to the Linux configuration files to help with other
3946 OS variants using glibc.
3948 85. One more patch for Sieve to update vacation handling to latest spec.
3951 ----------------------------------------------------
3952 See the note above about the 4.44 and 4.50 releases.
3953 ----------------------------------------------------
3959 1. Change 4.43/35 introduced a bug that caused file counts to be
3960 incorrectly computed when quota_filecount was set in an appendfile
3963 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3964 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3967 3. Give more explanation in the error message when the command for a transport
3968 filter fails to execute.
3970 4. There are several places where Exim runs a non-Exim command in a
3971 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3972 was being done only for the command run by the queryprogram router. It is
3973 now done for all such subprocesses. The other cases are: ${run, transport
3974 filters, and the commands run by the lmtp and pipe transports.
3976 5. Some older OS have a limit of 256 on the maximum number of file
3977 descriptors. Exim was using setrlimit() to set 1000 as a large value
3978 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3979 systems. I've change it so that if it can't get 1000, it tries for 256.
3981 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3982 was an oversight, and furthermore, ever since the addition of extra
3983 controls (e.g. 4.43/32), the checks on when to allow different forms of
3984 "control" were broken. There should now be diagnostics for all cases when a
3985 control that does not make sense is encountered.
3987 7. $recipients is now available in the predata ACL (oversight).
3989 8. Tidy the search cache before the fork to do a delivery from a message
3990 received from the command line. Otherwise the child will trigger a lookup
3991 failure and thereby defer the delivery if it tries to use (for example) a
3992 cached ldap connection that the parent has called unbind on.
3994 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3995 of $address_data from the recipient verification was clobbered by the
3996 sender verification.
3998 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3999 was its contents. (It was OK if the option was not defined at all.)
4001 11. A "Completed" log line is now written for messages that are removed from
4002 the spool by the -Mrm option.
4004 12. $host_address is now set to the target address during the checking of
4005 ignore_target_hosts.
4007 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4008 being passed; this would have caused $sender_host_name to have been used if
4009 matching the list had actually called for a host name (not very likely,
4010 since this list is usually IP addresses). A host name is now passed as
4013 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4014 code that specifies a non-restarting handler (typically sigaction() in
4015 modern systems) in an attempt to fix a rare and obscure crash bug.
4017 15. Narrowed the window for a race in the daemon that could cause it to ignore
4018 SIGCHLD signals. This is not a major problem, because they are used only to
4019 wake it up if nothing else does.
4021 16. A malformed maildirsize file could cause Exim to calculate negative values
4022 for the mailbox size or file count. Odd effects could occur as a result.
4023 The maildirsize information is now recalculated if the size or filecount
4026 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4027 support for a long time. Removed HAVE_SYS_VFS_H.
4029 18. Updated exipick to current release from John Jetmore.
4031 19. Allow an empty sender to be matched against a lookup in an address list.
4032 Previously the only cases considered were a regular expression, or an
4035 20. Exim went into a mad DNS lookup loop when doing a callout where the
4036 host was specified on the transport, if the DNS lookup yielded more than
4039 21. The RFC2047 encoding function was originally intended for short strings
4040 such as real names; it was not keeping to the 75-character limit for
4041 encoded words that the RFC imposes. It now respects the limit, and
4042 generates multiple encoded words if necessary. To be on the safe side, I
4043 have increased the buffer size for the ${rfc2047: expansion operator from
4046 22. Failure to deliver a bounce message always caused it to be frozen, even if
4047 there was an errors_to setting on the router. The errors_to setting is now
4050 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4051 canonical form (fully expanded) before being placed in
4052 $sender_host_address.
4054 24. Updated eximstats to version 1.33
4056 25. Include certificate and key file names in error message when GnuTLS fails
4057 to set them up, because the GnuTLS error message doesn't include the name
4058 of the failing file when there is a problem reading it.
4060 26. Expand error message when OpenSSL has problems setting up cert/key files.
4063 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4064 (this can affect the format of dates).
4066 28. exim_tidydb, when checking for the continued existence of a message for
4067 which it has found a message-specific retry record, was not finding
4068 messages that were in split spool directories. Consequently, it was
4069 deleting retry records that should have stayed in existence.
4071 29. eximstats updated to version 1.35
4072 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4073 1.35 - bugfix such that pie charts by volume are generated correctly
4075 30. The SPA authentication driver was not abandoning authentication and moving
4076 on to the next authenticator when an expansion was forced to fail,
4077 contradicting the general specification for all authenticators. Instead it
4078 was generating a temporary error. It now behaves as specified.
4080 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4081 (the order specifies the preference for clients). The order is now AES256,
4082 AES128, 3DES, ARCFOUR128.
4084 31. Small patch to Sieve code - explicitly set From: when generating an
4087 32. Exim crashed if a remote delivery caused a very long error message to be
4088 recorded - for instance if somebody sent an entire SpamAssassin report back
4089 as a large number of 550 error lines. This bug was coincidentally fixed by
4090 increasing the size of one of Exim's internal buffers (big_buffer) that
4091 happened as part of the Exiscan merge. However, to be on the safe side, I
4092 have made the code more robust (and fixed the comments that describe what
4095 33. Some experimental protocols are using DNS PTR records for new purposes. The
4096 keys for these records are domain names, not reversed IP addresses. The
4097 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4098 leaves it alone. Component reversal etc. now happens only for IP addresses.
4101 34. The host_aton() function is supposed to be passed a string that is known
4102 to be a valid IP address. However, in the case of IPv6 addresses, it was
4103 not checking this. This is a hostage to fortune. Exim now panics and dies
4104 if the condition is not met. A case was found where this could be provoked
4105 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4106 components; fortuitously, this particular loophole had already been fixed
4107 by change 4.50/55 or 4.44/33 above.
4109 If there are any other similar loopholes, the new check in host_aton()
4110 itself should stop them being exploited. The report I received stated that
4111 data on the command line could provoke the exploit when Exim was running as
4112 exim, but did not say which command line option was involved. All I could
4113 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4114 running as the user.
4117 35. There was a buffer overflow vulnerability in the SPA authentication code
4118 (which came originally from the Samba project). I have added a test to the
4119 spa_base64_to_bits() function which I hope fixes it.
4122 36. The daemon start-up calls getloadavg() while still root for those OS that
4123 need the first call to be done as root, but it missed one case: when
4124 deliver_queue_load_max is set with deliver_drop_privilege. This is
4125 necessary for the benefit of the queue runner, because there is no re-exec
4126 when deliver_drop_privilege is set.
4128 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4129 was in use, was not putting the data itself into the right store pool;
4130 consequently, it could be overwritten for a subsequent message in the same
4131 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4134 38. Sometimes the final signoff response after QUIT could fail to get
4135 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4136 < 0 before doing a fflush(). This bug looks as though it goes back to the
4137 introduction of TLS in release 3.20, but "sometimes" must have been rare
4138 because the tests only now provoked it.
4144 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4145 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4146 Now it returns an integer. A number of calls were still expecting a T/F
4147 return. Fortuitously, in all cases, the tests worked in OK situations,
4148 which is the norm. However, things would have gone wrong on any write
4149 failures on the smtp file descriptor. This function is used when sending
4150 messages over SMTP and also when doing verify callouts.
4152 2. When Exim is called to do synchronous delivery of a locally submitted
4153 message (the -odf or -odi options), it no longer closes stderr before doing
4156 3. Implemented the mua_wrapper option.
4158 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4160 5. Implemented the functions header_remove(), header_testname(),
4161 header_add_at_position(), and receive_remove_recipient(), and exported them
4164 6. If an ACL "warn" statement specified the addition of headers, Exim already
4165 inserted X-ACL-Warn: at the start if there was no header name. However, it
4166 was not making this test for the second and subsequent header lines if
4167 there were newlines in the string. This meant that an invalid header could
4168 be inserted if Exim was badly configured.
4170 7. Allow an ACL "warn" statement to add header lines at the start or after all
4171 the Received: headers, as well as at the end.
4173 8. Added the rcpt_4xx retry error code.
4175 9. Added postmaster_mailfrom=xxx to callout verification option.
4177 10. Added mailfrom=xxxx to the callout verification option, for verify=
4180 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4181 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4183 12. Inserted some casts to stop certain compilers warning when using pointer
4184 differences as field lengths or precisions in printf-type calls (mostly
4185 affecting debugging statements).
4187 13. Added optional readline() support for -be (dynamically loaded).
4189 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4190 same clock tick as a message's arrival, so that its received time was the
4191 same as the "first fail" time on the retry record, and that message
4192 remained on the queue past the ultimate address timeout, every queue runner
4193 would try a delivery (because it was past the ultimate address timeout) but
4194 after another failure, the ultimate address timeout, which should have then
4195 bounced the address, did not kick in. This was a "< instead of <=" error;
4196 in most cases the first failure would have been in the next clock tick
4197 after the received time, and all would be well.
4199 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4200 being recognized when the domain list was tested by the match_domain
4201 condition in an expansion string.
4203 16. Added the ${str2b64: operator.
4205 17. Exim was always calling setrlimit() to set a large limit for the number of
4206 processes, without checking whether the existing limit was already
4207 adequate. (It did check for the limit on file descriptors.) Furthermore,
4208 errors from getrlimit() and setrlimit() were being ignored. Now they are
4209 logged to the main and panic logs, but Exim does carry on, to try to do its
4210 job under whatever limits there are.
4212 18. Imported PCRE 5.0.
4214 19. Trivial typo in log message " temporarily refused connection" (the leading
4217 20. If the log selector return_path_on_delivery was set and an address was
4218 redirected to /dev/null, the delivery process crashed because it assumed
4219 that a return path would always be set for a "successful" delivery. In this
4220 case, the whole delivery is bypassed as an optimization, and therefore no
4223 21. Internal re-arrangement: the function for sending a challenge and reading
4224 a response while authentication was assuming a zero-terminated challenge
4225 string. It's now changed to take a pointer and a length, to allow for
4226 binary data in such strings.
4228 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4230 23. Exim was not respecting finduser_retries when seeking the login of the
4231 uid under which it was called; it was always trying 10 times. (The default
4232 setting of finduser_retries is zero.) Also, it was sleeping after the final
4233 failure, which is pointless.
4235 24. Implemented tls_on_connect_ports.
4237 25. Implemented acl_smtp_predata.
4239 26. If the domain in control=submission is set empty, Exim assumes that the
4240 authenticated id is a complete email address when it generates From: or
4241 Sender: header lines.
4243 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4244 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4245 chown and chgrp in /bin and hostname in /usr/bin.
4247 28. Exim was keeping the "process log" file open after each use, just as it
4248 does for the main log. This opens the possibility of it remaining open for
4249 long periods when the USR1 signal hits a daemon. Occasional processlog
4250 errors were reported, that could have been caused by this. Anyway, it seems
4251 much more sensible not to leave this file open at all, so that is what now
4254 29. The long-running daemon process does not normally write to the log once it
4255 has entered its main loop, and it closes the log before doing so. This is
4256 so that log files can straightforwardly be renamed and moved. However,
4257 there are a couple of unusual error situations where the daemon does write
4258 log entries, and I had neglected to close the log afterwards.
4260 30. The text of an SMTP error response that was received during a remote
4261 delivery was being truncated at 512 bytes. This is too short for some of
4262 the long messages that one sometimes sees. I've increased the limit to
4265 31. It is now possible to make retry rules that apply only when a message has a
4266 specific sender, in particular, an empty sender.
4268 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4269 it possible to be selective about when SMTP synchronization is enforced.
4271 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4273 32. Implemented hosts_connection_nolog.
4275 33. Added an ACL for QUIT.
4277 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4280 35. Added mailbox_size and mailbox_filecount to appendfile.
4282 36. Added control = no_multiline_responses to ACLs.
4284 37. There was a bug in the logic of the code that waits for the clock to tick
4285 in the case where the clock went backwards by a substantial amount such
4286 that the microsecond fraction of "now" was more than the microsecond
4287 fraction of "then" (but the whole seconds number was less).
4289 38. Added support for the libradius Radius client library this is found on
4290 FreeBSD (previously only the radiusclient library was supported).
4296 1. When certain lookups returned multiple values in the form name=value, the
4297 quoting of the values was not always being done properly. Specifically:
4298 (a) If the value started with a double quote, but contained no whitespace,
4300 (b) If the value contained whitespace other than a space character (i.e.
4301 tabs or newlines or carriage returns) it was not quoted.
4302 This fix has been applied to the mysql and pgsql lookups by writing a
4303 separate quoting function and calling it from the lookup code. The fix
4304 should probably also be applied to nisplus, ibase and oracle lookups, but
4305 since I cannot test any of those, I have not disturbed their existing code.
4307 2. A hit in the callout cache for a specific address caused a log line with no
4308 reason for rejecting RCPT. Now it says "Previous (cached) callout
4309 verification failure".
4311 3. There was an off-by-one bug in the queryprogram router. An over-long
4312 return line was truncated at 256 instead of 255 characters, thereby
4313 overflowing its buffer with the terminating zero. As well as fixing this, I
4314 have increased the buffer size to 1024 (and made a note to document this).
4316 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4317 when Exim is waiting for an SMTP response from a remote server, Exim
4318 restarts its select() call on the socket, thereby resetting its timeout.
4319 This is not a problem when such interrupts are rare. Somebody set up a cron
4320 job to run exiwhat every 2 minutes, which is less than the normal select()
4321 timeout (5 or 10 minutes). This meant that the select() timeout never
4322 kicked in because it was always reset. I have fixed this by comparing the
4323 time when an interrupt arrives with the time at the start of the first call
4324 to select(). If more time than the timeout has elapsed, the interrupt is
4325 treated as a timeout.
4327 5. Some internal re-factoring in preparation for the addition of Sieve
4328 extensions (by MH). In particular, the "personal" test is moved to a
4329 separate function, and given an option for scanning Cc: and Bcc: (which is
4330 not set for Exim filters).
4332 6. When Exim created an email address using the login of the caller as the
4333 local part (e.g. when creating a From: or Sender: header line), it was not
4334 quoting the local part when it contained special characters such as @.
4336 7. Installed new OpenBSD configuration files.
4338 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4339 try to make them clearer.
4341 9. Callout options, other than the timeout value, were being ignored when
4342 verifying sender addresses in header lines. For example, when using
4344 verify = header_sender/callout=no_cache
4346 the cache was (incorrectly) being used.
4348 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4349 only the Cygwin environment.
4351 11. When return_path_on_delivery was set as a log selector, if different remote
4352 addresses in the same message used different return paths and parallel
4353 remote delivery occurred, the wrong values would sometimes be logged.
4354 (Whenever a remote delivery process finished, the return path value from
4355 the most recently started remote delivery process was logged.)
4357 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4358 lines when AUTH and/or TLS are in use. This is the "received protocol"
4359 field. Exim used to use "asmtp" for authenticated SMTP, without any
4360 indication (in the protocol name) for TLS use. Now it follows the RFC and
4361 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4362 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4363 names appear in log lines as well as in Received: header lines.
4365 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4366 extensions, and comparison tests, and to fix some bugs.
4368 14. Changes to the "personal" filter test:
4370 (1) The test was buggy in that it was just doing the equivalent of
4371 "contains" tests on header lines. For example, if a user's address was
4372 anne@some.where, the "personal" test would incorrectly be true for
4374 To: susanne@some.where
4376 This test is now done by extracting each address from the header in turn,
4377 and checking the entire address. Other tests that are part of "personal"
4378 are now done using regular expressions (for example, to check local parts
4379 of addresses in From: header lines).
4381 (2) The list of non-personal local parts in From: addresses has been
4382 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4383 taken from the Sieve specification recommendations.
4385 (3) If the message contains any header line starting with "List-" it is
4386 treated as non-personal.
4388 (4) The test for "circular" in the Subject: header line has been removed
4389 because it now seems ill-conceived.
4391 15. Minor typos in src/EDITME comments corrected.
4393 16. Installed latest exipick from John Jetmore.
4395 17. If headers_add on a router specified a text string that was too long for
4396 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4397 of string_sprintf() is now avoided.
4399 18. $message_body_size was not set (it was always zero) when running the DATA
4400 ACL and the local_scan() function.
4402 19. For the "mail" command in an Exim filter, no default was being set for
4403 the once_repeat time, causing a random time value to be used if "once" was
4404 specified. (If the value happened to be <= 0, no repeat happened.) The
4405 default is now 0s, meaning "never repeat". The "vacation" command was OK
4406 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4407 (I found it when inspecting the code).
4409 20. There is now an overall timeout for performing a callout verification. It
4410 defaults to 4 times the callout timeout, which applies to individual SMTP
4411 commands during the callout. The overall timeout applies when there is more
4412 than one host that can be tried. The timeout is checked before trying the
4413 next host. This prevents very long delays if there are a large number of
4414 hosts and all are timing out (e.g. when the network connections are timing
4415 out). The value of the overall timeout can be changed by specifying an
4416 additional sub-option for "callout", called "maxwait". For example:
4418 verify = sender/callout=5s,maxwait=20s
4420 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4421 to the end before writing, but this should make it even safer).
4423 22. Exim was forgetting that it had advertised PIPELINING for the second and
4424 subsequent messages on an SMTP connection. It was also not resetting its
4425 memory on STARTTLS and an internal HELO.
4427 23. When Exim logs an SMTP synchronization error within a session, it now
4428 records whether PIPELINING has been advertised or not.
4430 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4431 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4432 rather than long int.
4434 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4436 26. Added the never_mail option to autoreply.
4442 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4443 crash if the getsockname() call failed; this can happen if a connection is
4444 closed very soon after it is established. The problem was simply in the
4445 order in which certain operations were done, causing Exim to try to write
4446 to the SMTP stream before it had set up the file descriptor. The bug has
4447 been fixed by making things happen in the correct order.
4453 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4454 before the connection was closed, thus losing the rejection response.
4456 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4457 some early Solaris releases, but causes trouble in current releases where
4458 socklen_t is defined.
4460 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4463 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4466 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4467 that releasing the top of it at the end releases what was used for sub-
4468 expansions (unless the block got too big). However, discard this block if
4469 the first thing is a variable or header, so that we can use its block when
4470 it is dynamic (useful for very large $message_headers, for example).
4472 6. Lookups now cache *every* query, not just the most recent. A new, separate
4473 store pool is used for this. It can be recovered when all lookup caches are
4474 flushed. Lookups now release memory at the end of their result strings.
4475 This has involved some general refactoring of the lookup sources.
4477 7. Some code has been added to the store_xxx() functions to reduce the amount
4478 of flapping under certain conditions.
4480 8. log_incoming_interface used to affect only the <= reception log lines. Now
4481 it causes the local interface and port to be added to several more SMTP log
4482 lines, for example "SMTP connection from", and rejection lines.
4484 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4486 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4488 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4489 could be overwritten at the end of the current message (or the start of a
4490 new message if it was set in a HELO ACL). The value is now preserved for
4491 the duration of the SMTP connection.
4493 12. If a transport had a headers_rewrite setting, and a matching header line
4494 contained an unqualified address, that address was qualified, even if it
4495 did not match any rewriting rules. The underlying bug was that the values
4496 of the flags that permit the existence of unqualified sender and recipient
4497 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4498 non-local messages, and by -bnq for local messages) were not being
4499 preserved with the message after it was received.
4501 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4502 "next input=" as part of the text comprising the host identity instead of
4503 the correct text. The code was using the same buffer for two different
4504 strings. However, depending on which order the printing function evaluated
4505 its arguments, the bug did not always show up. Under Linux, for example, my
4506 test suite worked just fine.
4508 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4509 doesn't work with some older versions of Perl. It has been changed to "my",
4510 which in any case is probably the better facility to use.
4512 15. A really picky compiler found some instances of statements for creating
4513 error messages that either had too many or two few arguments for the format
4516 16. The size of the buffer for calls to the DNS resolver has been increased
4517 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4518 for addresses that have a lot of PTR records. This alleviates a problem; it
4519 does not fully solve it.
4521 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4522 buffer now truncates the list and logs the incident, which is the same
4523 action as happens when Exim is looking up a host name and its aliases.
4524 Previously in this situation something unpredictable would happen;
4525 sometimes it was "internal error: store_reset failed".
4527 18. If a server dropped the connection unexpectedly when an Exim client was
4528 using GnuTLS and trying to read a response, the client delivery process
4529 crashed while trying to generate an error log message.
4531 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4532 string, for example:
4534 warn message = H1: something\nH2: something
4536 the text was added as a single header line from Exim's point of view
4537 though it ended up OK in the delivered message. However, searching for the
4538 second and subsequent header lines using $h_h2: did not work. This has been
4539 fixed. Similarly, if a system filter added multiple headers in this way,
4540 the routers could not see them.
4542 20. Expanded the error message when iplsearch is called with an invalid key to
4543 suggest using net-iplsearch in a host list.
4545 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4546 ACLs are no longer actually imposed (and a message to that effect is
4549 22. If a "gecos" field in a passwd entry contained escaped characters, in
4550 particular, if it contained a \" sequence, Exim got it wrong when building
4551 a From: or a Sender: header from that name. A second bug also caused
4552 incorrect handling when an unquoted " was present following a character
4553 that needed quoting.
4555 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4556 was not being matched caselessly.
4558 24. Arranged for all hyphens in the exim.8 source to be escaped with
4561 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4562 properly. Recipient callout cache records were still being keyed to include
4563 the sender, even when use_sender was set false. This led to far more
4564 callouts that were necessary. The sender is no longer included in the key
4565 when use_sender is false.
4567 26. Added "control = submission" modifier to ACLs.
4569 27. Added the ${base62d: operator to decode base 62 numbers.
4571 28. dnsdb lookups can now access SRV records.
4573 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4574 the configuration file.
4576 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4577 (-v) mode. This makes the output for a verbose queue run more intelligible.
4579 31. Added a use_postmaster feature to recipient callouts.
4581 32. Added the $body_zerocount variable, containing the number of binary zero
4582 bytes in the message body.
4584 33. The time of last modification of the "new" subdirectory is now used as the
4585 "mailbox time last read" when there is a quota error for a maildir
4588 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4590 35. Added +ignore_unknown as a special item in host lists.
4592 36. Code for decoding IPv6 addresses in host lists is now included, even if
4593 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4594 address was recognized as an IP address, but was then not correctly decoded
4595 into binary, causing unexpected and incorrect effects when compared with
4602 1. Very minor rewording of debugging text in manualroute to say "list of
4603 hosts" instead of "hostlist".
4605 2. If verify=header_syntax was set, and a header line with an unqualified
4606 address (no domain) and a large number of spaces between the end of the
4607 name and the colon was received, the reception process suffered a buffer
4608 overflow, and (when I tested it) crashed. This was caused by some obsolete
4609 code that should have been removed. The fix is to remove it!
4611 3. When running in the test harness, delay a bit after writing a bounce
4612 message to get a bit more predictability in the log output.
4614 4. Added a call to search_tidyup() just before forking a reception process. In
4615 theory, someone could use a lookup in the expansion of smtp_accept_max_
4616 per_host which, without the tidyup, could leave open a database connection.
4618 5. Added the variables $recipient_data and $sender_data which get set from a
4619 lookup success in an ACL "recipients" or "senders" condition, or a router
4620 "senders" option, similar to $domain_data and $local_part_data.
4622 6. Moved the writing of debug_print from before to after the "senders" test
4625 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4626 problems for message scanning, either using a data ACL, or using
4627 local_scan() because the Received: header was not generated till after they
4628 were called (in order to set the time as the time of reception completion).
4629 I have revised the way this works. The header is now generated after the
4630 body is received, but before the ACL or local_scan() are called. After they
4631 are run, the timestamp in the header is updated.
4637 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4638 before starting a queue runner without re-exec. This happened only when
4639 deliver_drop_privilege was set or when the Exim user was set to root. The
4640 effect of the bug was that timeouts during subsequent deliveries caused
4641 crashes instead of being properly handled. The handler is now left at its
4642 default (and expected) setting.
4644 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4645 message, again when deliver_drop_privilege is set or Exim is run as root.
4646 The bug described in (1) was not present in this case, but the tidying up
4647 of the other signals was missing. I have made the two cases consistent.
4649 3. The ignore_target_hosts setting on a manualroute router was being ignored
4650 for hosts that were looked up using the /MX notation.
4652 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4655 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4656 operated on the sender address. After changing the $sender_address to <>
4657 for the sender address verify, Exim was re-instated it as the original
4658 (before rewriting) address, but remembering that it had rewritten it, so it
4659 wasn't rewriting it again. This bug also had the effect of breaking the
4660 sender address verification caching when the sender address was rewritten.
4662 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4663 This has been changed so that if the ip literal address matches
4664 ignore_target_hosts, the router declines.
4666 7. Added expansion conditions match_domain, match_address, and match_local_
4667 part (NOT match_host).
4669 8. The placeholder for the Received: header didn't have a length field set.
4671 9. Added code to Exim itself and to exim_lock to test for a specific race
4672 condition that could lead to file corruption when using MBX delivery. The
4673 issue is with the lockfile that is created in /tmp. If this file is removed
4674 after a process has opened it but before that process has acquired a lock,
4675 there is the potential for a second process to recreate the file and also
4676 acquire a lock. This could lead to two Exim processes writing to the file
4677 at the same time. The added code performs the same test as UW imapd; it
4678 checks after acquiring the lock that its file descriptor still refers to
4679 the same named file.
4681 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4682 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4685 11. Added the smtp_active_hostname option. If used, this will typically be made
4686 to depend on the incoming interface address. Because $interface_address is
4687 not set up until the daemon has forked a reception process, error responses
4688 that can happen earlier (such as "too many connections") no longer contain
4691 12. If an expansion in a condition on a "warn" statement fails because a lookup
4692 defers, the "warn" statement is abandoned, and the next ACL statement is
4693 processed. Previously this caused the whole ACL to be aborted.
4695 13. Added the iplsearch lookup type.
4697 14. Added ident_timeout as a log selector.
4699 15. Added tls_certificate_verified as a log selector.
4701 16. Added a global option tls_require_ciphers (compare the smtp transport
4702 option of the same name). This controls incoming TLS connections.
4704 17. I finally figured out how to make tls_require_ciphers do a similar thing
4705 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4706 before starting the TLS session.
4708 18. Tabs are now shown as \t in -bP output.
4710 19. If the log selector return_path_on_delivery was set, Exim crashed when
4711 bouncing a message because it had too many Received: header lines.
4713 20. If two routers both had headers_remove settings, and the first one included
4714 a superfluous trailing colon, the final name in the first list and the
4715 first name in the second list were incorrectly joined into one item (with a
4716 colon in the middle).
4722 1. Added -C and -D options to the exinext utility, mainly to make it easier
4723 to include in the automated testing, but these could be helpful when
4724 multiple configurations are in use.
4726 2. The exinext utility was not formatting the output nicely when there was
4727 an alternate port involved in the retry record key, nor when there was a
4728 message id as well (for retries that were specific to a specific message
4729 and a specific host). It was also confused by IPv6 addresses, because of
4730 the additional colons they contain. I have fixed the IPv4 problem, and
4731 patched it up to do a reasonable job for IPv6.
4733 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4734 delivery, the log line now contains "pipelined" if PIPELINING was used.
4736 4. An SMTP transport process used to panic and die if the bind() call to set
4737 an explicit outgoing interface failed. This has been changed; it is now
4738 treated in the same way as a connect() failure.
4740 5. A reference to $sender_host_name in the part of a conditional expansion
4741 that was being skipped was still causing a DNS lookup. This no longer
4744 6. The def: expansion condition was not recognizing references to header lines
4745 that used bh_ and bheader_.
4747 7. Added the _cache feature to named lists.
4749 8. The code for checking quota_filecount in the appendfile transport was
4750 allowing one more file than it should have been.
4752 9. For compatibility with Sendmail, the command line option
4760 and sets the incoming protocol and host name (for trusted callers). The
4761 host name and its colon can be omitted when only the protocol is to be set.
4762 Note the Exim already has two private options, -pd and -ps, that refer to
4763 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4764 "s", but I don't think that's a major issue.
4766 10. A number of refactoring changes to the code, none of which should affect
4769 (a) The number of logging options was getting close to filling up the
4770 32-bit word that was used as a bit map. I have split them into two classes:
4771 those that are passed in the argument to log_write(), and those that are
4772 only ever tested independently outside of that function. These are now in
4773 separate 32-bit words, so there is plenty of room for expansion again.
4774 There is no change in the user interface or the logging behaviour.
4776 (b) When building, for example, log lines, the code previously used a
4777 macro that called string_cat() twice, in order to add two strings. This is
4778 not really sufficiently general. Furthermore, there was one instance where
4779 it was actually wrong because one of the argument was used twice, and in
4780 one call a function was used. (As it happened, calling the function twice
4781 did not affect the overall behaviour.) The macro has been replaced by a
4782 function that can join an arbitrary number of extra strings onto a growing
4785 (c) The code for expansion conditions now uses a table and a binary chop
4786 instead of a serial search (which was left over from when there were very
4787 few conditions). Also, it now recognizes conditions like "pam" even when
4788 the relevant support is not compiled in: a suitably worded error message is
4789 given if an attempt is made to use such a condition.
4791 11. Added ${time_interval:xxxxx}.
4793 12. A bug was causing one of the ddress fields not to be passed back correctly
4794 from remote delivery subprocesses. The field in question was not being
4795 subsequently used, so this caused to problems in practice.
4797 13. Added new log selectors queue_time and deliver_time.
4799 14. Might have fixed a bug in maildirsizefile handling that threw up
4800 "unexpected character" debug warnings, and recalculated the data
4801 unnecessarily. In any case, I expanded the warning message to give more
4804 15. Added the message "Restricted characters in address" to the statements in
4805 the default ACL that block characters like @ and % in local parts.
4807 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4808 Three changes have been made:
4810 (a) There was a serious bug; a negative response to MAIL caused the whole
4811 recipient domain to be cached as invalid, thereby blocking all messages
4812 to all local parts at the same domain, from all senders. This bug has
4813 been fixed. The domain is no longer cached after a negative response to
4814 MAIL if the sender used is not empty.
4816 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4819 (c) A new callout option, "use_sender" has been added for people who want
4820 the modified behaviour.
4826 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4829 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4830 indeed breaks things for older releases.
4832 3. Added additional logging to the case where there is a problem reading data
4833 from a filter that is running in a subprocess using a pipe, in order to
4834 try to track down a specific problem.
4836 4. Testing facility fudge: when running in the test harness and attempting
4837 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4838 getting "No route to host". Convert this to a timeout.
4840 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4843 6. Some OS don't have socklen_t but use size_t instead. This affects the
4844 fifth argument of getsockopt() amongst other things. This is now
4845 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4846 can be set for individual OS. I have set it for SunOS5, OSF1, and
4847 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4848 some earlier ones do not.
4850 7. Change 4.30/15 was not doing the test caselessly.
4852 8. The standard form for an IPv6 address literal was being rejected by address
4853 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4854 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4855 this, as well as the form without the "IPv6" on the front (but only when
4856 address literals are enabled, of course).
4858 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4860 10. Exim crashed if a message with an empty sender address specified by -f
4861 encountered a router with an errors_to setting. This could be provoked only
4862 by a command such as
4866 where an empty string was supplied; "<>" did not hit this bug.
4868 11. Installed PCRE release 4.5.
4870 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4871 remained set. It is now erased.
4873 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4874 times from message ids (which are base 36 rather than the normal 62).
4876 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4877 were being counted as actual protocol errors, and logged if the log
4878 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4879 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4880 and DATA following a set of rejected RCPTs do not count as protocol errors.
4881 In other words, Exim assumes they were pipelined, though this may not
4882 actually be the case. Of course, in all cases the client gets an
4883 appropriate error code.
4885 15. If a lookup fails in an ACL condition, a message about the failure may
4886 be available; it is used if testing the ACL cannot continue, because most
4887 such messages specify what the cause of the deferral is. However, some
4888 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4889 that caused an old message to be retained and used if a later statement
4890 caused a defer, replacing the real cause of the deferral.
4892 16. If an IP address had so many PTR records that the DNS lookup buffer
4893 was not large enough to hold them, Exim could crash while trying to process
4894 the truncated data. It now detects and logs this case.
4896 17. Further to 4.21/58, another change has been made: if (and only if) the
4897 first line of a message (the first header line) ends with CRLF, a bare LF
4898 in a subsequent header line has a space inserted after it, so as not to
4899 terminate the header.
4901 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4902 unnecessarily, used atoi() instead of strtol(), and didn't check the
4903 termination when getting file sizes from file names by regex.
4905 19. Completely re-implemented the support for maildirsize files, in the light
4906 of a number of problems with the previous contributed implementation
4907 (4.30/29). In particular:
4909 . If the quota is zero, the maildirsize file is maintained, but no quota is
4912 . If the maildir directory does not exist, it is created before any attempt
4913 to write a maildirsize file.
4915 . The quota value in the file is just a cache; if the quota is changed in
4916 the transport, the new value overrides.
4918 . A regular expression is available for excluding directories from the
4921 20. The autoreply transport checks the characters in options that define the
4922 message's headers; it allows continued headers, but it was checking with
4923 isspace() after an embedded newline instead of explicitly looking for a
4926 21. If all the "regular" hosts to which an address was routed had passed their
4927 expiry times, and had not reached their retry times, the address was
4928 bounced, even if fallback hosts were defined. Now Exim should go on to try
4931 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4932 equivalent code in the SMTP transport. Some hosts send humungous responses
4933 to HELO/EHLO, more than 1024 it seems.
4935 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4936 gives compiler warnings in some environments. I've now done it "properly",
4939 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4940 (because of problems with the built-in one) was declared to return uschar *
4941 instead of char *, causing compiler failure.
4943 25. Fixed a file descriptor leak when processing alias/forward files.
4945 26. Fixed a minor format string issue in dbfn.c.
4947 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4949 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4950 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4953 29. When a sender address is verified, it is cached, to save repeating the test
4954 when there is more than one recipient in a message. However, when the
4955 verification involves a callout, it is possible for different callout
4956 options to be set for different recipients. It is too complicated to keep
4957 track of this in the cache, so now Exim always runs a verification when a
4958 callout is required, relying on the callout cache for the optimization.
4959 The overhead is duplication of the address routing, but this should not be
4962 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4963 to be verified with callout=postmaster, and the main callout worked but the
4964 postmaster check failed, the verification correctly failed. However, if a
4965 subsequent RCPT command asked for sender verification *without* the
4966 postmaster check, incorrect caching caused this verification also to fail,
4969 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4970 it was not caching the DNS options (qualify_single, search_parents) that
4971 were used when the lookup failed. A subsequent lookup with different
4972 options therefore always gave the same answer, though there were cases
4973 where it should not have. (Example: a "domains = !$mx_any" option on a
4974 dnslookup router: the "domains" option is always processed without any
4975 widening, but the router might have qualify_single set.) Now Exim uses the
4976 cached value only when the same options are set.
4978 32. Added John Jetmore's "exipick" utility to the distribution.
4980 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4981 than a timeout (e.g. a certificate is required, and is not provided), an
4982 Exim server now closes the connection immediately. Previously it waited for
4983 the client to close - but if the client is SSL, it seems that they each
4984 wait for each other, leading to a delay before one of them times out.
4986 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4987 maintained 0.8.x compatibility because I don't think many are using it, and
4988 it is clearly obsolete.
4990 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4993 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4994 client certificate was expired. A simple patch fixes this, though I don't
4995 understand the full logic of why the verify callback is called multiple
4998 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4999 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5000 which causes problems with some clients (such as the Certicom SSL Plus
5001 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5002 disables the coutermeasure allowing Eudora to connect."
5004 38. Exim was not checking that a write() to a log file succeeded. This could
5005 lead to Bad Things if a log got too big, in particular if it hit a file
5006 size limit. Exim now panics and dies if it cannot write to a log file, just
5007 as it does if it cannot open a log file.
5009 39. Modified OS/Makefile-Linux so that it now contains
5011 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5013 The two -D definitions ensure that Exim is compiled with large file
5014 support, which makes it possible to handle log files that are bigger than
5017 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5018 instance) a domain was checked against a named list that involved a lookup,
5019 causing $domain_data to be set, then another domain was checked against the
5020 same list, then the first domain was re-checked, the value of $domain_data
5021 after the final check could be wrong. In particular, if the second check
5022 failed, it could be set empty. This bug probably also applied to
5025 41. The strip_trailing_dot option was not being applied to the address given
5026 with the -f command-line option.
5028 42. The code for reading a message's header from the spool was incrementing
5029 $received_count, but never initializing it. This meant that the value was
5030 incorrect (doubled) while delivering a message in the same process in which
5031 it was received. In the most common configuration of Exim, this never
5032 happens - a fresh exec is done - but it can happen when
5033 deliver_drop_privilege is set.
5035 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5036 it now includes up to 150 characters of the unexpected data in the log
5039 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5040 and building data strings. The size of both of these buffers was 10 000
5041 bytes - far larger than anybody would *ever* want, thought I. Needless to
5042 say, somebody hit the limit. I have increased the maximum line length to
5043 20 000 and the maximum data length of concatenated lines to 100 000. I have
5044 also fixed two bugs, because there was no checking on these buffers. Tsk,
5045 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5046 buffer is too small.
5048 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5049 lsearch lookups. Now it does.
5051 46. When parsing a route_list item in a manualroute router, a fixed-length
5052 buffer was used for the list of hosts. I made this 1024 bytes long,
5053 thinking that nobody would ever have a list of hosts that long. Wrong.
5054 Somebody had a whole pile of complicated expansion conditions, and the
5055 string was silently truncated, leading to an expansion error. It turns out
5056 that it is easier to change to an unlimited length (owing to other changes
5057 that have happened since this code was originally written) than to build
5058 structure for giving a limitation error. The length of the item that
5059 expands into the list of hosts is now unlimited.
5061 47. The lsearch lookup could not handle data where the length of text line was
5062 more than 4095 characters. Such lines were truncated, leading to shortened
5063 data being returned. It should now handle lines of any length.
5065 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5066 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5069 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5070 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5071 no longer happens in comment lines. A list of replacements is now placed
5072 at the head of all of the source files, except those whose only change is
5073 to replace PERL_COMMAND in the very first #! line.
5075 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5076 messages on the queue, with a bottom-up merge sort, using code contributed
5077 by Michael Haardt. This should make operations like -bp somewhat faster on
5078 large queues. It won't affect queue runners, except when queue_run_in_order
5081 51. Installed eximstats 1.31 in the distribution.
5083 52. Added support for SRV lookups to the dnslookup router.
5085 53. If an ACL referred to $message_body or $message_body_end, the value was not
5086 reset for any messages that followed in the same SMTP session.
5088 54. The store-handling optimization for building very long strings was not
5089 differentiating between the different store pools. I don't think this
5090 actually made any difference in practice, but I've tidied it.
5092 55. While running the routers to verify a sender address, $sender_address
5093 was still set to the sender address. This is wrong, because when routing to
5094 send a bounce to the sender, it would be empty. Therefore, I have changed
5095 it so that, while verifying a sender address, $sender_address is set to <>.
5096 (There is no change to what happens when verifying a recipient address.)
5098 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5099 target A or AAAA records (if not already returned) without resetting the
5100 qualify_single or search_parents options of the DNS resolver. These are
5101 inappropriate in this case because the targets of MX and SRV records must
5102 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5103 target that, when qualified, matched something in the local domain. These
5104 two options are now turned off when doing these lookups.
5106 57. It seems that at least some releases of Reiserfs (which does not have the
5107 concept of a fixed number of inodes) returns zero and not -1 for the
5108 number of available inodes. This interacted badly with check_spool_inodes,
5109 which assumed that -1 was the "no such thing" setting. What I have done is
5110 to check that the total number of inodes is greater than zero before doing
5111 the test of how many are available.
5113 58. When a "warn" ACL statement has a log_message modifier, the message is
5114 remembered, and not repeated. This is to avoid a lot of repetition when a
5115 message has many recipients that cause the same warning to be written.
5116 Howewer, Exim was preserving the list of already written lines for an
5117 entire SMTP session, which doesn't seem right. The memory is now reset if a
5118 new message is started.
5120 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5121 debugging output unless log_rewrite was also set.
5123 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5124 of (int)(handle) when we know that handle contains (void *)(-1).
5126 61. The Exim daemon panic-logs an error return when it closes the incoming
5127 connection. However "connection reset by peer" seems to be common, and
5128 isn't really an error worthy of noting specially, so that particular error
5131 62. When Exim is trying to find all the local interfaces, it used to panic and
5132 die if the ioctl to get the interface flags failed. However, it seems that
5133 on at least one OS (Solaris 9) it is possible to have an interface that is
5134 included in the list of interfaces, but for which you get a failure error
5135 for this call. This happens when the interface is not "plumbed" into a
5136 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5137 failure of the "get flags" call assumes that the interface is down.
5139 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5140 makes life easier for people who are doing arithmetic on fields extracted
5141 from dates, where you often get leading zeros that should not be
5142 interpreted as octal.
5144 64. Added qualify_domain to the redirect router, to override the global
5147 65. If a pathologically long header line contained very many addresses (the
5148 report of this problem mentioned 10 000) and each of them was rewritten,
5149 Exim could use up a very large amount of memory. (It kept on making new
5150 copies of the header line as it rewrote, and never released the old ones.)
5151 At the expense of a bit more processing, the header rewriting function has
5152 been changed so that it no longer eats memory in this way.
5154 66. The generation of the Received: header has been moved from the time that a
5155 message starts to be received, to the time that it finishes. The timestamp
5156 in the Received: header should now be very close to that of the <= log
5157 line. There are two side-effects of this change:
5159 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5160 logged header lines no longer include the local Received: line, because
5161 it has not yet been created. The same applies to a copy of the message
5162 that is returned to a non-SMTP sender when a message is rejected.
5164 (b) When a filter file is tested using -bf, no additional Received: header
5165 is added to the test message. After some thought, I decided that this
5168 This change does not affect the value of $received_for. It is still set
5169 after address rewriting, but before local_scan() is called.
5171 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5173 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5174 gave an unhelpful panic error message, and a defer error. I have managed to
5175 change this behaviour so that it now rejects any supplied certificate,
5176 which seems right, as the list of acceptable certificates is empty.
5178 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5179 gave an unhelpful defer error. I have not managed to make this reject any
5180 supplied certificates, but the error message it gives is "no certificate
5181 supplied", which is not helpful.
5183 70. exigrep's output now also includes lines that are not associated with any
5184 message, but which match the given pattern. Implemented by a patch from
5185 Martin Sluka, which also tidied up the Perl a bit.
5187 71. Recipient callout verification, like sender verification, was using <> in
5188 the MAIL FROM command. This isn't really the right thing, since the actual
5189 sender may affect whether the remote host accepts the recipient or not. I
5190 have changed it to use the actual sender in the callout; this means that
5191 the cache record is now keyed on a recipient/sender pair, not just the
5192 recipient address. There doesn't seem to be a real danger of callout loops,
5193 since a callout by the remote host to check the sender would use <>.
5194 [SEE ABOVE: changed after hitting problems.]
5196 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5197 temporary errors. However, in the case of such a code being given after
5198 the end of a data transmission (i.e. after ".") Exim was failing to write
5199 a retry record for the message. (Yes, there was some broken host that was
5200 actually sending 8xx at this point.)
5202 73. An unknown lookup type in a host list could cause Exim to panic-die when
5203 the list was checked. (An example that provoked this was putting <; in the
5204 middle of a list instead of at the start.) If this happened during a DATA
5205 ACL check, a -D file could be left lying around. This kind of configuration
5206 error no longer causes Exim to die; instead it causes a defer errror. The
5207 incident is still logged to the main and panic logs.
5209 74. Buglet left over from Exim 3 conversion. The message "too many messages
5210 in one connection" was written to the rejectlog but not the mainlog, except
5211 when address rewriting (yes!) was being logged.
5213 75. Added write_rejectlog option.
5215 76. When a system filter was run not as root (that is, when system_filter_user
5216 was set), the values of the $n variables were not being returned to the
5217 main process; thus, they were not subsequently available in the $sn
5220 77. Added +return_path_on_delivery log selector.
5222 78. A connection timeout was being treated differently from recipients deferred
5223 when testing hosts_max_try with a message that was older than the host's
5224 retry timeout. (The host should not be counted, thus allowing all hosts to
5225 be tried at least once before bouncing.) This may have been the cause of an
5226 occasionally reported bug whereby a message would remain on the queue
5227 longer than the retry timeout, but would be bounced if a delivery was
5228 forced. I say "may" because I never totally pinned down the problem;
5229 setting up timeout/retry tests is difficult. See also the next item.
5231 79. The ultimate address timeout was not being applied to errors that involved
5232 a combination of host plus message (for example, a timeout on a MAIL
5233 command). When an address resolved to a number of possible hosts, and they
5234 were not all tried for each delivery (e.g. because of hosts_max_try), a
5235 message could remain on the queue longer than the retry timeout.
5237 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5240 81. Fixed an obscure SMTP outgoing bug which required at least the following
5241 conditions: (a) there was another message waiting for the same server;
5242 (b) the server returned 5xx to all RCPT commands in the first message so
5243 that the message was not completed; (c) the server dropped the connection
5244 or gave a negative response to the RSET that Exim sends to abort the
5245 transaction. The observed case was a dropped connection after DATA that had
5246 been sent in pipelining mode. That is, the server had advertised PIPELINING
5247 but was not implementing it correctly. The effect of the bug was incorrect
5248 behaviour, such as trying another host, and this could lead to a crash.
5254 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5255 and daemon.c were passed as pointers to ints; they should have been
5256 pointers to socklen_t variables (which are typically unsigned ints).
5258 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5261 3. Fixed a really odd bug that affected only the testing scheme; patching a
5262 certain fixed string in the binary changed the value of another string that
5263 happened to be identical to the end of the original first string.
5265 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5266 name", it returns that address as the IP address. On some operating
5267 systems (e.g. Solaris), it also passes back the IP address string as the
5268 "host name". However, on others (e.g. Linux), it passes back an empty
5269 string. Exim wasn't checking for this, and was changing the host name to an
5270 empty string, assuming it had been canonicized.
5272 5. Although rare, it is permitted to have more than one PTR record for a given
5273 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5274 all the names associated with an address, because they do in Solaris.
5275 However, it seems that they do not in Linux for data that comes from the
5276 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5277 I found this out when I moved to a new Linux workstation and tried to run
5278 the Exim test suite.
5280 To get round this problem I have changed the code so that it now does its
5281 own call to the DNS to look up PTR records when searching for a host name.
5282 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5283 addresses that are only in /etc/hosts are still found.
5285 This behaviour is, however, controlled by an option called host_lookup_
5286 order, which defaults to "bydns:byaddr". If people want to use the other
5287 order, or indeed, just use one or the other means of lookup, they can
5288 specify it in this variable.
5290 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5291 some operating systems, this comes back from gethostbyaddr() as an empty
5292 string, and this is what Exim used to test for. However, it seems that in
5293 other systems, "." is yielded. Exim now tests for this case too.
5295 7. The values of check_spool_space and check_log_space are now held internally
5296 as a number of kilobytes instead of an absolute number of bytes. If a
5297 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5298 kilobyte. This means that much larger values can be stored.
5300 8. Exim monitor: an attempt to get the action menu when not actually pointing
5301 at a message produces an empty menu entitled "No message selected". This
5302 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5303 no entries in it ("Shell widget menu has zero width and/or height"). So I
5304 have added a single, blank menu entry in this case.
5306 9. Added ${quote_local_part.
5308 10. MIME decoding is now applied to the contents of Subject: header lines when
5311 11. Now that a reference to $sender_host_address automatically causes a reverse
5312 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5313 host lookup before query-style lookups in lists that might use this
5314 variable. This has therefore been abolished, and the "net-" prefix is no
5315 longer necessary for query-style lookups.
5317 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5318 have been a typo for LFLAGS, so it has been changed.
5320 13. The install script calls Exim with "-C /dev/null" in order to find the
5321 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5322 to be output. Howeve, since Exim outputs its version number before the
5323 error, it didn't break the script. It just looked ugly. I fixed this by
5324 always allowing "-C /dev/null" if the caller is root.
5326 14. Ignore overlarge ACL variable number when reading spool file - insurance
5327 against a later release with more variables having written the file.
5329 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5330 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5331 this, as well as the form without the "IPv6" on the front.
5333 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5334 OS/Makefile-Darwin file.
5336 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5337 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5339 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5340 variable. However, some OS use ints for inodes. I've added cast to long int
5341 to get rid of the compiler warning.
5343 19. I had forgotten to lock out "/../" in configuration file names when
5344 ALT_CONFIG_PREFIX was set.
5346 20. Routers used for verification do not need to specify transports. However,
5347 if such a router generated a host list, and callout was configured, Exim
5348 crashed, because it could not find a port number from the (non-existent)
5349 transport. It now assumes port 25 in this circumstance.
5351 21. Added the -t option to exigrep.
5353 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5354 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5355 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5356 them is set, LOOKUP_LSEARCH is forced.
5358 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5360 24. Added sender and host information to the "rejected by local_scan()" log
5361 line; previously there was no indication of these.
5363 25. Added .include_if_exists.
5365 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5366 Linux. It turns out that not all file systems support this. Apparently some
5367 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5368 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5369 not supported on the file descriptor, is now ignored when Exim is trying to
5370 sync a directory. This applies only to Linux.
5372 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5374 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5375 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5378 29. Added support for maildirsize files from supplied patch (modified a bit).
5380 30. The use of :fail: followed by an empty string could lead Exim to respond to
5381 sender verification failures with (e.g.):
5383 550 Verification failed for <xxx>
5384 550 Sender verify failed
5386 where the first response line was missing the '-' that indicates it is not
5387 the final line of the response.
5389 31. The loop for finding the name of the user that called Exim had a hardwired
5390 limit of 10; it now uses the value of finduser_retries, which is used for
5391 all other user lookups.
5393 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5396 33. Exim was neglecting to zero errno before one call of strtol() when
5397 expanding a string and expecting an integer value. On some systems this
5398 resulted in spurious "integer overflow" errors. Also, it was casting the
5399 result into an int without checking.
5401 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5402 did not work. The code looks as if it has *never* worked, though it appears
5403 to have been documented since at least releast 1.62. I have made it work.
5405 35. The "timeout_DNS" error in retry rules, also documented since at least
5406 1.62, also never worked. As it isn't clear exactly what this means, and
5407 clearly it isn't a major issue, I have abolished the feature by treating it
5408 as "timeout", and writing a warning to the main and panic logs.
5410 36. The display of retry rules for -brt wasn't always showing the error code
5413 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5414 timeout_connect_A, timeout_connect_MX.
5416 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5417 to the empty sender.
5419 39. The daemon was not analysing the content of -oX till after it had closed
5420 stderr and disconnected from the controlling terminal. This meant that any
5421 syntax errors were only noted on the panic log, and the return code from
5422 the command was 0. By re-arranging the code a little, I've made the
5423 decoding happen first, so such errors now appear on stderr, and the return
5424 code is 1. However, the actual setting up of the sockets still happens in
5425 the disconnected process, so errors there are still only recorded on the
5428 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5429 connections (as happens on some IP stacks) was logged at start up time as
5430 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5431 it from "IPv6 and IPv4", which means that two separate sockets are being
5434 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5435 says whether AF_INET or AF_INET6 was passed as an argument.
5437 42. Exiwhat output was messed up when time zones were included in log
5440 43. Exiwhat now gives more information about the daemon's listening ports,
5441 and whether -tls-on-connect was used.
5443 44. The "port" option of the smtp transport is now expanded.
5445 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5446 silently ignored. Now an error message is written to the main and panic
5449 46. There's a new ACL modifier called "logwrite" which writes to a log file
5450 as soon as it is encountered.
5452 47. Added $local_user_uid and $local_user_gid at routing time.
5454 48. Exim crashed when trying to verify a sender address that was being
5457 49. Exim was recognizing only a space character after ".include". It now also
5458 recognizes a tab character.
5460 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5461 extracting the relevant information from the specification. The man page no
5462 longer contains scrambled data for the -d option, and I've added a section
5463 at the front about calling Exim under different names.
5465 51. Added "extra_headers" argument to the "mail" command in filter files.
5467 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5470 53. Installed eximstats 1.29.
5472 54. Added transport_filter_timeout as a generic transport option.
5474 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5475 Cc: header lines. This was required by RFC 822, but it not required by RFC
5478 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5479 incoming messages that did not have them. Now it does so only if the
5480 message originates locally, that is, if there is no associated remote host
5481 address. When Resent- header lines are present, this applies to the Resent-
5482 lines rather than the non-Resent- lines.
5484 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5485 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5487 58. Messages for configuration errors now include the name of the main
5488 configuration file - useful now that there may be more than one file in a
5489 list (.included file names were always shown).
5491 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5492 for those rare installations that do not start the daemon as root or run it
5493 setuid root. I've cut out the call to initgroups() if the daemon is not
5496 60. The Exim user and group can now be bound into the binary as text strings
5497 that are looked up at the start of Exim's processing.
5499 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5501 62. Added $mailstore_basename variable.
5503 63. Installed patch to sieve.c from Michael Haardt.
5505 64. When Exim failed to open the panic log after failing to open the main log,
5506 the original message it was trying to log was written to stderr and debug
5507 output, but if they were not available (the usual case in production), it
5508 was lost. Now it is written to syslog before the two lines that record the
5509 failures to open the logs.
5511 65. Users' Exim filters run in subprocesses under the user's uid. It is
5512 possible for a "deliver" command or an alias in a "personal" command to
5513 provoke an address rewrite. If logging of address rewriting is configured,
5514 this fails because the process is not running as root or exim. There may be
5515 a better way of dealing with this, but for the moment (because 4.30 needs
5516 to be released), I have disabled address rewrite logging when running a
5517 filter in a non-root, non-exim process.
5523 1. The buildconfig auxiliary program wasn't quoting the value set for
5524 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5525 not defined. This bug was masked in 4.22 by the effect that was fixed in
5528 2. Some messages that were rejected after a message id was allocated were
5529 shown as "incomplete" by exigrep. It no longer does this for messages that
5530 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5532 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5533 have allow_domain_literals set, the ID did not get logged in the <= line.
5534 Domain literals are now always recognized in Message-ID: header lines.
5536 4. The first argument for a ${extract expansion item is the key name or field
5537 number. Leading and trailing spaces in this item were not being ignored,
5538 causing some misleading effects.
5540 5. When deliver_drop_privilege was set, single queue runner processes started
5541 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5542 same command in the process it spins off) were not dropping privilege.
5544 6. When the daemon running as "exim" started a queue runner, it always
5545 re-executed Exim in the spun-off process. This is a waste of effort when
5546 deliver_drop_privilege is set. The new process now just calls the
5547 queue-runner function directly.
5553 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5556 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5557 ignored. Though the use of -oP was forcing the writing of a pid file, it
5558 was always written to the default place.
5560 3. If the message "no IP address found for host xxxx" is generated during
5561 incoming verification, it is now followed by identification of the incoming
5562 connection (so you can more easily find what provoked it).
5564 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5566 5. Added some features to "harden" Exim a bit more against certain attacks:
5568 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5569 be put in Local/Makefile. This is like the never_users runtime option,
5570 but it cannot be overridden. The default setting is "root".
5572 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5573 prefix string with which any file named in a -C command line option
5576 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5577 is retained for -C and -D only if the caller of Exim is root. Without
5578 it, the exim user may also use -C and -D and retain privilege.
5580 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5581 command line option is disabled.
5583 6. Macro names set by the -D option must start with an upper case letter, just
5584 like macro names defined in the configuration file.
5586 7. Added "dereference=" facility to LDAP.
5588 8. Two instances of the typo "uknown" in the source files are fixed.
5590 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5591 the Configure-Makefile script screwed up while processing it.
5593 10. Incorporated PCRE 4.4.
5595 11. The SMTP synchronization check was not operating right at the start of an
5596 SMTP session. For example, it could not catch a HELO sent before the client
5597 waited for the greeting. There is now a check for outstanding input at the
5598 point when the greeting is written. Because of the duplex, asynchronous
5599 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5600 way, but not yet received, when the check is performed.
5602 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5603 on TCP/IP sockets, because this apparently causes some broken clients to
5606 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5607 unchanged) from the Cygwin maintainer.
5609 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5610 was supported instead of testing for mbx. Effectively a typo.
5612 15. The spa authenticator server code was not checking that the input it
5613 received was valid base64.
5615 16. The debug output line for the "set" modifier in ACLs was not showing the
5616 name of the variable that was being set.
5618 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5620 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5621 lookup on its own. Something else was needed to trigger the lookup. For
5622 example, a match in host_lookup or the need for a host name in a host list.
5623 Now, if $sender_host_name is referenced and the host name has not yet been
5624 looked up, a lookup is performed. If the lookup fails, the variable remains
5625 empty, and $host_lookup_failed is set to "1".
5627 19. Added "eqi" as a case-independent comparison operator.
5629 20. The saslauthd authentication condition could segfault if neither service
5630 nor realm was specified.
5632 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5633 error message that was logged was misleading, and incoming SMTP
5634 connections were dropped. The message is now more accurate, and temporary
5635 errors are given to SMTP connections.
5637 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5638 (except RSET) with a 421 error, until QUIT is received. However, it was
5639 failing to send a response to QUIT.
5641 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5642 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5643 ACL is now run after the helo_try_verify_hosts code.
5645 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5646 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5647 case-independent, so other case variants are also recognized). Apparently
5648 some systems use these upper case variants.
5650 25. If more than two messages were waiting for the same host, and a transport
5651 filter was specified for the transport, Exim sent two messages over the
5652 same TCP/IP connection, and then failed with "socket operation on non-
5653 socket" when it tried to send the third.
5655 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5657 27. The extern definition of crypt16() in expand.c was not being excluded when
5658 the OS had its own crypt16() function.
5660 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5661 as a preferred synonym for return_size_limit, both as an option and as an
5664 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5666 30. Changed the default configuration ACL to relax the local part checking rule
5667 for addresses that are not in any local domains. For these addresses,
5668 slashes and pipe symbols are allowed within local parts, but the sequence
5669 /../ is explicitly forbidden.
5671 31. SPA server authentication was not clearing the challenge buffer before
5674 32. log_message in a "warn" ACL statement was writing to the reject log as
5675 well as to the main log, which contradicts the documentation and doesn't
5676 seem right (because no rejection is happening). So I have stopped it.
5678 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5679 However, I am unable to do any testing of this.
5681 34. Fixed an infelicity in the appendfile transport. When checking directories
5682 for a mailbox, to see if any needed to be created, it was accidentally
5683 using path names with one or more superfluous leading slashes; tracing
5684 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5686 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5687 contents are added to the log line that is written for every discarded
5688 recipient. (Previously a log_message setting was ignored.)
5690 36. The ${quote: operator now quotes the string if it is empty.
5692 37. The install script runs exim in order to find its version number. If for
5693 some reason other than non-existence or emptiness, which it checks, it
5694 could not run './exim', it was installing it with an empty version number,
5695 i.e. as "exim-". This error state is now caught, and the installation is
5698 38. An argument was missing from the function that creates an error message
5699 when Exim fails to connect to the socket for saslauthd authentication.
5700 This could cause Exim to crash, or give a corrupted message.
5702 39. Added isip, isip4, and isip6 to ${if conditions.
5704 40. The ACL variables $acl_xx are now saved with the message, and can be
5705 accessed later in routers, transports, and filters.
5707 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5708 strings in the file are not string-expanded.
5710 42. If a MAIL command specified a SIZE value that was too large to fit into an
5711 int variable, the check against message_size_limit failed. Such values are
5712 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5713 day this will have to be increased, but I don't think I want to be around
5714 when emails are that large.
5721 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5722 iconv() is not standard in FreeBSD.
5724 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5725 IPv6 enabled. The observed symptom was a segmentation fault on return from
5726 the function os_common_find_running_interfaces() in src/os.c.
5728 3. In the check_special_case() function in daemon.c I had used "errno" as an
5729 argument name, which causes warnings on some systems. This was basically a
5730 typo, since it was named "eno" in the comments!
5732 4. The code that waits for the clock to tick (at a resolution of some fraction
5733 of a second) so as to ensure message-id uniqueness was always waiting for
5734 at least one whole tick, when it could have waited for less. [This is
5735 almost certainly not relevant at current processor speeds, where it is
5736 unlikely to ever wait at all. But we try to future-proof.]
5738 5. The function that sleeps for a time interval that includes fractions of a
5739 second contained a race. It did not block SIGALRM between setting the
5740 timer, and suspending (a couple of lines later). If the interval was short
5741 and the sigsuspend() was delayed until after it had expired, the suspension
5742 never ended. On busy systems this could lead to processes getting stuck for
5745 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5746 process, before it forks any delivery processes. The open lookup caching
5747 mechanism meant that the open file or database connection was passed into
5748 the delivery process. The problem was that delivery processes always tidy
5749 up cached lookup data. This could cause a problem for the next delivery
5750 process started by the queue runner, because the external queue runner
5751 process does not know about the closure. So the next delivery process
5752 still has data in the lookup cache. In the case of a file lookup, there was
5753 no problem because closing a file descriptor in a subprocess doesn't affect
5754 the parent. However, if the lookup was caching a connection to a database,
5755 the connection was closed, and the second delivery process was likely to
5756 see errors such as "PGSQL: query failed: server closed the connection
5757 unexpectedly". The problem has been fixed by closing all cached lookups
5758 in a queue runner before running a delivery process.
5760 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5761 seem to have the "const" qualifier which it has on other OS. I've
5764 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5765 *of the same type* (client or server) with the same public name that an
5766 error should be diagnosed.
5768 9. When Exim looked up a host name for an IP address, but failed to find the
5769 original IP address when looking up the host name (a safety check), it
5770 output the message "<ip address> does not match any IP for NULL", which was
5771 confusing, to say the least. The bug was that the host name should have
5772 appeared instead of "NULL".
5774 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5775 user that is built into the binary, and the -C or -D options is used, root
5776 privilege is dropped before the configuration file is read. In addition,
5777 logging is switched to stderr instead of the normal log files. If the
5778 configuration then re-defines the Exim user, the unprivileged environment
5779 is probably not what is expected, so Exim logs a panic warning message (but
5782 However, if deliver_drop_privilege is set, the unprivileged state may well
5783 be exactly what is intended, so the warning has been cut out in that case,
5784 and Exim is allowed to try to write to its normal log files.
5790 1. smtp_return_error_details was not giving details for temporary sender
5791 or receiver verification errors.
5793 2. Diagnose a configuration error if two authenticators have the same public
5796 3. Exim used not to create the message log file for a message until the first
5797 delivery attempt. This could be confusing when incoming messages were held
5798 for policy or load reasons. The message log file is now created at the time
5799 the message is received, and an initial "Received" line is written to it.
5801 4. The automatically generated man page for command line options had a minor
5802 bug that caused no ill effects; however, a more serious problem was that
5803 the procedure for building the man page automatically didn't always
5804 operate. Consequently, release 4.20 contains an out-of-date version. This
5805 shouldn't happen again.
5807 5. When building Exim with embedded Perl support, the script that builds the
5808 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5809 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5811 6. The freeze_tell option was not being used for messages that were frozen on
5812 arrival, either by an ACL or by local_scan().
5814 7. Added the smtp_incomplete_transaction log selector.
5816 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5817 was accepting AUTH without a new EHLO.
5819 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5820 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5823 10. Logging of TCP/IP connections (when configured) now happens in the main
5824 daemon process instead of the child process, so that the TCP/IP connection
5825 count is more accurate (but it can never be perfect).
5827 11. The use of "drop" in a nested ACL was not being handled correctly in the
5828 outer ACL. Now, if condition failure induced by the nested "drop" causes
5829 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5830 or "require"), the connection is dropped.
5832 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5833 that yield "discard" can now be used with an "accept" or a "discard" verb,
5834 but an error is generated for any others (because I can't see a useful way
5835 to define what should happen).
5837 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5838 are now processed in the same way as lines in the Exim configuration file.
5839 In particular, continuation lines are supported.
5841 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5843 15. Added -ti meaning -t -i.
5845 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5846 domains, and warn by logging if others are found.
5848 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5849 returned from GIFCONF when Exim is trying to find the list of interfaces on
5850 a host. The code in os.c has been modified to copy each ifreq to an aligned
5851 structure in all cases.
5853 Also, in some cases, the returned ifreq's were being copied to a 'struct
5854 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5855 means the last couple of bytes of an IPv6 address could be chopped if the
5856 ifreq contained only a normal sockaddr (14 bytes storage).
5858 18. Named domain lists were not supported in the hosts_treat_as_local option.
5859 An entry such as +xxxx was not recognized, and was treated as a literal
5862 19. Ensure that header lines added by a DATA ACL are included in the reject log
5863 if the ACL subsequently rejects the message.
5865 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5866 MD5 (which is deprecated).
5868 21. When testing a filter file using -bf, Exim was writing a message when it
5869 took the sender from a "From " line in the message, but it was not doing so
5870 when it took $return_path from a Return-Path: header line. It now does.
5872 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5873 with a valid header line field name (a series of printing characters
5874 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5876 23. Changed "disc" in the source to "disk" to conform to the documentation and
5877 the book and for uniformity.
5879 24. Ignore Sendmail's -Ooption=value command line item.
5881 25. When execve() failed while trying to run a command in a pipe transport,
5882 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5883 could be confused with a return value of 69 from the command itself. This
5884 has been changed to 127, the value the shell returns if it is asked to run
5885 a non-existent command. The wording for the related log line suggests a
5886 non-existent command as the problem.
5888 26. If received_header_text expands to an empty string, do not add a Received:
5889 header line to the message. (Well, it adds a token one on the spool, but
5890 marks it "old" so that it doesn't get used or transmitted.)
5892 27. Installed eximstats 1.28 (addition of -nt option).
5894 28. There was no check for failure on the call to getsockname() in the daemon
5895 code. This can fail if there is a shortage of resources on the system, with
5896 ENOMEM, for example. A temporary error is now given on failure.
5898 29. Contrary to the C standard, it seems that in some environments, the
5899 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5900 program. Exim now does this explicitly; it affects the formatting of
5901 timestamps using strftime().
5903 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5904 complaints. I've now initialized all the variables, to avoid this.
5906 32. Header lines added by a system filter were not being "seen" during
5907 transport-time rewrites.
5909 33. The info_callback() function passed to OpenSSL is set up with type void
5910 (*)(SSL *, int, int), as described somewhere. However, when calling the
5911 function (actually a macro) that sets it up, the type void(*)() is
5912 expected. I've put in a cast to prevent warnings from picky compilers.
5914 34. If a DNS black list lookup found a CNAME record, but there were no A
5915 records associated with the domain it pointed at, Exim crashed.
5917 35. If a DNS black list lookup returned more than one A record, Exim ignored
5918 all but the first. It now scans all returned addresses if a particular IP
5919 value is being sought. In this situation, the contents of the
5920 $dnslist_value variable are a list of all the addresses, separated by a
5923 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5924 to accept a host name and all its aliases if the forward lookup for any of
5925 them yielded the IP address of the incoming connection. Now it accepts only
5926 those names whose forward lookup yields the correct IP address. Any other
5927 names are discarded. This closes a loophole whereby a rogue DNS
5928 administrator could create reverse DNS records to break through a
5929 wildcarded host restriction in an ACL.
5931 37. If a user filter or a system filter that ran in a subprocess used any of
5932 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5933 the wrong values were passed to the pipe command ($thisaddress had the
5934 value of $0, $0 had the value of $1, etc). This bug was introduced by
5935 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5937 38. Improved the line breaking for long SMTP error messages from ACLs.
5938 Previously, if there was no break point between 40 and 75 characters, Exim
5939 left the rest of the message alone. Two changes have been made: (a) I've
5940 reduced the minimum length to 35 characters; (b) if it can't find a break
5941 point between 35 and 75 characters, it looks ahead and uses the first one
5942 that it finds. This may give the occasional overlong line, but at least the
5943 remaining text gets split now.
5945 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5946 file descriptors might be low, and that setting 1000 would always raise it.
5947 It turns out that in some environments, the limit is already over 1000 and
5948 that lowering it causes trouble. So now Exim takes care not to decrease it.
5950 40. When delivering a message, the value of $return_path is set to $sender_
5951 address at the start of routing (routers may change the value). By an
5952 oversight, this default was not being set up when an address was tested by
5953 -bt or -bv, which affected the outcome if any router or filter referred to
5956 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5957 log only when "message" or "log_message" are set. However, if one of the
5958 conditions was an address verification, or a call to a nested ACL, the
5959 messages generated by the underlying test were being passed through. This
5960 no longer happens. The underlying message is available in $acl_verify_
5961 message for both "message" and "log_message" expansions, so it can be
5962 passed through if needed.
5964 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5965 new expansion $bh_ to give the encoded byte string without charset
5966 translation. Translation happens only if iconv() is available; HAVE_ICONV
5967 indicates this at build time. HEADERS_CHARSET gives the charset to
5968 translate to; headers_charset can change it in the configuration, and
5969 "headers charset" can change it in an individual filter file.
5971 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5972 that creates RFC 2047 encoded "words" labels them as that charset instead
5973 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5974 expansion operator; (ii) when Exim creates a From: line for a local
5975 message; (iii) when a header line is rewritten to include a "phrase" part.
5977 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5978 buggy, causing it to skip the first lines of messages whose message ID
5979 ended in 'D'. This would not have bitten before Exim release 4.14, because
5980 message IDs were unlikely to end in 'D' before then. The effect was to have
5981 incorrect size information for certain domains.
5983 45. #include "config.h" was missing at the start of the crypt16.c module. This
5984 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5987 46. If there was a timeout during a "random" callout check, Exim treated it as
5988 a failure of the random address, and carried on sending RSET and the real
5989 address. If the delay was just some slowness somewhere, the response to the
5990 original RCPT would be taken as a response to RSET and so on, causing
5991 mayhem of various kinds.
5993 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5994 when I implemented it. It didn't allow for the fact that some option values
5995 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5996 the right test for positive values.
5998 48. Domain names in DNS records are case-independent. Exim always looks them up
5999 in lower case. Some resolvers return domain names in exactly the case they
6000 appear in the zone file, that is, they may contain uppercase letters. Not
6001 all resolvers do this - some return always lower case. Exim was treating a
6002 change of case by a resolver as a change of domain, similar to a widening
6003 of a domain abbreviation. This triggered its re-routing code and so it was
6004 trying to route what was effectively the same domain again. This normally
6005 caused routing to fail (because the router wouldn't handle the domain
6006 twice). Now Exim checks for this case specially, and just changes the
6007 casing of the domain that it ultimately uses when it transmits the message
6010 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6013 50. If a filter generated a file delivery with a non-absolute name (possible if
6014 no home directory exists for the router), the forbid_file option was not
6017 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6018 the existing equality matching.
6020 52. Exim was using ints instead of ino_t variables in some places where it was
6021 dealing with inode numbers.
6023 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6024 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6025 called TMPDIR, and if it finds it is different, it changes its value.
6027 54. The smtp_printf() function is now made available to local_scan() so
6028 additional output lines can be written before returning. There is also an
6029 smtp_fflush() function to enable the detection of a dropped connection.
6030 The variables smtp_input and smtp_batched_input are exported to
6033 55. Changed the default runtime configuration: the message "Unknown user"
6034 has been removed from the ACL, and instead placed on the localuser router,
6035 using the cannot_route_message feature. This means that any verification
6036 failures that generate their own messages won't get overridden. Similarly,
6037 the "Unrouteable address" message that was in the ACL for unverifiable
6038 relay addresses has also been removed.
6040 56. Added hosts_avoid_esmtp to the smtp transport.
6042 57. The exicyclog script was not checking for the esoteric option
6043 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6044 will work only if exicyclog is run under the appropriate euid.
6046 58. Following a discussion on the list, the rules by which Exim recognises line
6047 endings on incoming messages have been changed. The -dropcr and drop_cr
6048 options are now no-ops, retained only for backwards compatibility. The
6049 following line terminators are recognized: LF CRLF CR. However, special
6050 processing applies to CR:
6052 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6053 nor a local message in the state where . is a terminator.
6055 (ii) If a bare CR is encountered in a header line, an extra space is added
6056 after the line terminator so as not to end the header. The reasoning
6057 behind this is that bare CRs in header lines are most likely either
6058 to be mistakes, or people trying to play silly games.
6060 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6061 was being incorrectly given as 18 bytes larger than it should have been.
6062 This is a VOB (very old bug).
6064 60. This may never have affected anything current, but just in case it has:
6065 When the local host is found other than at the start of a list of hosts,
6066 the local host, those with the same MX, and any that follow, are discarded.
6067 When the list in question was part of a longer list of hosts, the following
6068 hosts (not currently being processed) were also being discarded. This no
6069 longer happens. I'm not sure if this situation could ever has previously
6072 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6075 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6076 header. This is something that is recommended in a new Internet Draft, and
6077 is something that is documented as being done by Sendmail. There are two
6078 possible values. For messages generated by the autoreply transport, Exim
6081 Auto-Submitted: auto-replied
6083 whereas for all other generated messages (e.g. bounces) it adds
6085 Auto-Submitted: auto-generated
6087 63. The "personal" condition in filters now includes a test for the
6088 Auto-Submitted: header. If it contains the string "auto-" the message it
6089 not considered personal.
6091 64. Added rcpt_include_affixes as a generic transport option.
6093 65. Added queue_only_override (default true).
6095 66. Added the syslog_duplication option.
6097 67. If what should have been the first header line of a message consisted of
6098 a space followed by a colon, Exim was mis-interpreting it as a header line.
6099 It isn't of course - it is syntactically invalid and should therefore be
6100 treated as the start of the message body. The misbehaviour could have
6101 caused a number of strange effects, including loss of data in subsequent
6102 header lines, and spool format errors.
6104 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6105 client host had authenticated. This control can now be exercised by an ACL
6106 for more flexibility.
6108 69. By default, callouts do not happen when testing with -bh. There is now a
6109 variant, -bhc, which does actually run the callout code, including
6110 consulting and updating the callout cache.
6112 70. Added support for saslauthd authentication, courtesy of Alexander
6115 71. If statvfs() failed on the spool or log directories while checking their
6116 size for availability, Exim confusingly gave the error "space shortage".
6117 Furthermore, in debugging mode it crashed with a floating point exception.
6118 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6119 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6120 serious problem, Exim now writes to the main and panic logs when this
6121 happens, with details of the failure. It then refuses to accept the
6122 incoming message, giving the message "spool directory problem" or "log
6123 directory problem" with a 421 code for SMTP messages.
6125 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6126 0, 1, and 2 exist, because some OS complain for execs without them (see
6127 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6128 descriptors. However, the code omitted to check that the open succeeded,
6129 causing mysterious errors if for some reason the permissions on /dev/null
6130 got screwed. Now Exim writes a message to the main and panic logs, and
6131 bombs out if it can't open /dev/null.
6133 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6134 interact so that it is all more flexible. It is supposed to remain
6135 backwards compatible. Also added extra_local_interfaces.
6137 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6138 to bomb out with an assertion failure - to the client this appears as a
6139 connection drop. This problem occurs in the part of the code that was taken
6140 from the Samba project. Fortunately, the assertion is in a very simple
6141 function, so I have fixed this by reproducing the function inline in the
6142 one place where it is called, and arranging for authentication to fail
6143 instead of killing the process with assert().
6145 75. The SPA client code was not working when the server requested OEM rather
6146 than Unicode encoding.
6148 76. Added code to make require_files with a specific uid setting more usable in
6149 the case where statting the file as root fails - usually a non-root-mounted
6150 NFS file system. When this happens and the failure is EACCES, Exim now
6151 forks a subprocess and does the per-uid checking as the relevant uid.
6153 77. Added process_log_path.
6155 78. If log_file_path was not explicitly set, a setting of check_log_space or
6156 check_log_inodes was ignored.
6158 79. If a space check for the spool or log partitions fails, the incident is now
6159 logged. Of course, in the latter case the data may get lost...
6161 80. Added the %p formatting code to string_format() so that it can be used to
6162 print addresses in debug_print(). Adjusted all the address printing in the
6163 debugging in store.c to use %p rather than %d.
6165 81. There was a concern that a line of code in smtp_in.c could overflow a
6166 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6167 initially expressed, the concern was not well-founded, because trailing
6168 spaces are removed early. However, if the trailing spaces were followed by
6169 a NULL, they did not get removed, so the overflow was possible. Two fixes
6172 (a) I re-wrote the offending code in a cleaner fashion.
6173 (b) If an incoming SMTP command contains a NULL character, it is rejected
6176 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6177 runs initgroups(), so that if the Exim user is in any additional groups,
6178 they will be used during message reception.
6184 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob