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).
68 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
69 checkers that were complaining about end of non-void function with no
72 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
73 This was a regression intruduced in 4.83 by another bugfix.
75 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
77 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
78 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
84 TF/01 Correctly close the server side of TLS when forking for delivery.
86 When a message was received over SMTP with TLS, Exim failed to clear up
87 the incoming connection properly after forking off the child process to
88 deliver the message. In some situations the subsequent outgoing
89 delivery connection happened to have the same fd number as the incoming
90 connection previously had. Exim would try to use TLS and fail, logging
91 a "Bad file descriptor" error.
93 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
94 utilities have not been installed.
96 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
97 temporary space as the ACL may create new global variables.
99 TL/01 LDAP support uses per connection or global context settings, depending
100 upon the detected version of the libraries at build time.
102 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
103 to extract and use the src ip:port in logging and expansions as if it
104 were a direct connection from the outside internet. PPv2 support was
105 updated based on HAProxy spec change in May 2014.
107 JH/02 Add ${listextract {number}{list}{success}{fail}}.
109 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
110 Properly escape header and check for NULL return.
112 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
115 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
117 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
118 characters in header names, implemented as a verify condition.
119 Contributed by Michael Fischer v. Mollard.
121 TL/05 Rename SPF condition results err_perm and err_temp to standardized
122 results permerror and temperror. Previous values are deprecated but
123 still accepted. In a future release, err_perm and err_temp will be
124 completely removed, which will be a backward incompatibility if the
125 ACL tests for either of these two old results. Patch contributed by
126 user bes-internal on the mailing list.
128 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
130 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
131 selectors, in both main and reject logs.
133 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
136 JH/07 Add malware type "sock" for talking to simple daemon.
138 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
140 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
141 routers/transports under cutthrough routing.
143 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
144 numbers. Touch up "bool" conditional to keep the same definition.
146 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
148 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
150 JH/12 Expand items in router/transport headers_add or headers_remove lists
151 individually rather than the list as a whole. Bug 1452.
153 Required for reasonable handling of multiple headers_ options when
154 they may be empty; requires that headers_remove items with embedded
155 colons must have them doubled (or the list-separator changed).
157 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
158 view the policy declared in the DMARC record. Currently, $dmarc_status
159 is a combined value of both the record presence and the result of the
162 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
164 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
165 dnslookup router and the smtp transport (applying to the forward
168 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
169 of ldap servers used for a specific lookup. Patch provided by Heiko
172 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
173 New variable $lookup_dnssec_authenticated for observability.
175 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
176 Patch submitted by Lars Timman.
178 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
180 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
181 Requires trusted mode and valid format message id, aborts otherwise.
182 Patch contributed by Heiko Schlichting.
184 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
185 certextract with support for various fields. Bug 1358.
187 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
188 is requested by default, modifiable by smtp transport option
191 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
192 operate on certificate variables to give certificate fingerprints
193 Also new ${sha256:cert_variable}.
195 JH/23 The PRDR feature is moved from being Experimental into the mainline.
197 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
198 Christian Aistleitner.
200 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
202 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
203 file. Patch from Wolfgang Breyha.
205 JH/25 Expand the coverage of the delivery $host and $host_address to
206 client authenticators run in verify callout. Bug 1476.
208 JH/26 Port service names are now accepted for tls_on_connect_ports, to
209 align with daemon_smtp_ports. Bug 72.
211 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
212 support and error reporting did not work properly.
214 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
215 and is readable. Patch from Andrew Colin Kissa.
217 TL/14 Enhance documentation of ${run expansion and how it parses the
218 commandline after expansion, particularly in the case when an
219 unquoted variable expansion results in an empty value.
221 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
223 PP/02 Fix internal collision of T_APL on systems which support RFC3123
224 by renaming away from it. Addresses GH issue 15, reported by
227 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
229 TL/15 SECURITY: prevent double expansion in math comparison functions
230 (can expand unsanitized data). Not remotely exploitable.
237 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
239 PP/02 Make -n do something, by making it not do something.
240 When combined with -bP, the name of an option is not output.
242 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
245 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
246 $sender_host_name and config options to manage this, and basic check
249 PP/05 DSCP support for outbound connections and control modifier for inbound.
251 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
252 (Only plugin which currently uses this is kerberos4, which nobody should
253 be using, but we should make it available and other future plugins might
254 conceivably use it, even though it would break NAT; stuff *should* be
255 using channel bindings instead).
257 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
258 name; added for Sendmail compatibility; requires admin caller.
259 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
260 just ignore it); requires trusted caller.
261 Also parse but ignore: -Ac -Am -X<logfile>
264 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
266 TL/02 Add +smtp_confirmation as a default logging option.
268 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
269 Patch by Magnus Holmgren from 2007-02-20.
271 TL/04 Bugzilla 1281 - Spec typo.
272 Bugzilla 1283 - Spec typo.
273 Bugzilla 1290 - Spec grammar fixes.
275 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
277 TL/06 Add Experimental DMARC support using libopendmarc libraries.
279 TL/07 Fix an out of order global option causing a segfault. Reported to dev
280 mailing list by by Dmitry Isaikin.
282 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
284 JH/02 Support "G" suffix to numbers in ${if comparisons.
286 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
288 NM/01 Bugzilla 1197 - Spec typo
289 Bugzilla 1196 - Spec examples corrections
291 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
293 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
294 gnutls_enable_pkcs11, but renamed to more accurately indicate its
297 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
298 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
300 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
301 "acl {{name}{arg}...}", and optional args on acl condition
304 JH/05 Permit multiple router/transport headers_add/remove lines.
306 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
308 JH/07 Avoid using a waiting database for a single-message-only transport.
309 Performance patch from Paul Fisher. Bugzilla 1262.
311 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
314 JH/09 Add $headers_added variable, with content from use of ACL modifier
315 add_header (but not yet added to the message). Bugzilla 199.
317 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
318 Pulled from Bugzilla 817 by Wolfgang Breyha.
320 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
322 (nb: this is the same fix as in Exim 4.80.1)
324 JH/11 Add A= logging on delivery lines, and a client_set_id option on
327 JH/12 Add optional authenticated_sender logging to A= and a log_selector
330 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
332 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
333 advertise SMTP AUTH mechanism to us, instead of a generic
334 protocol violation error. Also, make Exim more robust to bad
335 data from the Dovecot auth socket.
337 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
339 When a queue runner is handling a message, Exim first routes the
340 recipient addresses, during which it prunes them based on the retry
341 hints database. After that it attempts to deliver the message to
342 any remaining recipients. It then updates the hints database using
345 So if a recipient address works intermittently, it can get repeatedly
346 deferred at routing time. The retry hints record remains fresh so the
347 address never reaches the final cutoff time.
349 This is a fairly common occurrence when a user is bumping up against
350 their storage quota. Exim had some logic in its local delivery code
351 to deal with this. However it did not apply to per-recipient defers
352 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
354 This change adds a proper retry rule check during routing so that the
355 final cutoff time is checked against the message's age. We only do
356 this check if there is an address retry record and there is not a
357 domain retry record; this implies that previous attempts to handle
358 the address had the retry_use_local_parts option turned on. We use
359 this as an approximation for the destination being like a local
360 delivery, as in LMTP.
362 I suspect this new check makes the old local delivery cutoff check
363 redundant, but I have not verified this so I left the code in place.
365 TF/02 Correct gecos expansion when From: is a prefix of the username.
367 Test 0254 submits a message to Exim with the header
371 When I ran the test suite under the user fanf2, Exim expanded
372 the header to contain my full name, whereas it should have added
373 a Resent-Sender: header. It erroneously treats any prefix of the
374 username as equal to the username.
376 This change corrects that bug.
378 GF/01 DCC debug and logging tidyup
379 Error conditions log to paniclog rather than rejectlog.
380 Debug lines prefixed by "DCC: " to remove any ambiguity.
382 TF/03 Avoid unnecessary rebuilds of lookup-related code.
384 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
385 Bug spotted by Jeremy Harris; was flawed since initial commit.
386 Would have resulted in OCSP responses post-SNI triggering an Exim
387 NULL dereference and crash.
389 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
391 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
392 Bug detection, analysis and fix by Samuel Thibault.
393 Bugzilla 1331, Debian bug #698092.
395 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
397 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
398 Server implementation by Todd Lyons, client by JH.
399 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
400 config variable "prdr_enable" controls whether the server
401 advertises the facility. If the client requests PRDR a new
402 acl_data_smtp_prdr ACL is called once for each recipient, after
403 the body content is received and before the acl_smtp_data ACL.
404 The client is controlled by bolth of: a hosts_try_prdr option
405 on the smtp transport, and the server advertisement.
406 Default client logging of deliveries and rejections involving
407 PRDR are flagged with the string "PRDR".
409 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
410 fclose(). Diagnosis by Todd Lyons.
412 PP/17 Update configure.default to handle IPv6 localhost better.
413 Patch by Alain Williams (plus minor tweaks).
416 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
417 This is now consistent with GnuTLS, and is now documented: the
418 previous undocumented portable approach to treating the option as
419 unset was to force an expansion failure. That still works, and
420 an empty string is now equivalent.
422 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
423 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
424 not performing validation itself.
426 PP/20 Added force_command boolean option to pipe transport.
427 Patch from Nick Koston, of cPanel Inc.
429 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
432 TF/04 Added udpsend ACL modifer and hexquote expansion operator
434 PP/21 Fix eximon continuous updating with timestamped log-files.
435 Broken in a format-string cleanup in 4.80, missed when I repaired the
436 other false fix of the same issue.
437 Report and fix from Heiko Schlichting.
440 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
441 Report from Prashanth Katuri.
443 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
444 It's SecureTransport, so affects any MacOS clients which use the
445 system-integrated TLS libraries, including email clients.
447 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
448 using a MIME ACL for non-SMTP local injection.
449 Report and assistance in diagnosis by Warren Baker.
451 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
453 JH/16 Fix comparisons for 64b. Bugzilla 1385.
455 TL/09 Add expansion variable $authenticated_fail_id to keep track of
456 last id that failed so it may be referenced in subsequent ACL's.
458 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
461 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
462 ldap library initialization, allowing self-signed CA's to be
463 used. Also properly sets require_cert option later in code by
464 using NULL (global ldap config) instead of ldap handle (per
465 session). Bug diagnosis and testing by alxgomz.
467 TL/12 Enhanced documentation in the ratelimit.pl script provided in
468 the src/util/ subdirectory.
470 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
471 renamed to Transport Post Delivery Action by Jeremy Harris, as
474 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
475 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
476 redis_servers = needs to be configured which will be used by the redis
477 lookup. Patch from Warren Baker, of The Packet Hub.
479 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
481 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
482 hostname or reverse DNS when processing a host list. Used suggestions
483 from multiple comments on this bug.
485 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
487 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
488 Missed a few lines, added it to make the runtest require no keyboard
491 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
492 contains upper case chars. Make router use caseful_local_part.
494 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
495 support when GnuTLS has been built with p11-kit.
501 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
503 This, or similar/improved, will also be change PP/11 of 4.82.
509 PP/01 Handle short writes when writing local log-files.
510 In practice, only affects FreeBSD (8 onwards).
511 Bugzilla 1053, with thanks to Dmitry Isaikin.
513 NM/01 Bugzilla 949 - Documentation tweak
515 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
518 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
520 PP/02 Implemented gsasl authenticator.
522 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
524 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
525 `pkg-config foo` for cflags/libs.
527 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
528 with rest of GSASL and with heimdal_gssapi.
530 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
531 `pkg-config foo` for cflags/libs for the TLS implementation.
533 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
534 properties get this fed in as external SSF. A number of robustness
535 and debugging improvements to the cyrus_sasl authenticator.
537 PP/08 cyrus_sasl server now expands the server_realm option.
539 PP/09 Bugzilla 1214 - Log authentication information in reject log.
540 Patch by Jeremy Harris.
542 PP/10 Added dbmjz lookup type.
544 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
546 PP/12 MAIL args handles TAB as well as SP, for better interop with
547 non-compliant senders.
548 Analysis and variant patch by Todd Lyons.
550 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
551 Bug report from Lars Müller <lars@samba.org> (via SUSE),
552 Patch from Dirk Mueller <dmueller@suse.com>
554 PP/13 tls_peerdn now print-escaped for spool files.
555 Observed some $tls_peerdn in wild which contained \n, which resulted
556 in spool file corruption.
558 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
559 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
560 or write after TLS renegotiation, which otherwise led to messages
563 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
564 as a tracking header (ie: a signed header comes before the signature).
565 Patch from Wolfgang Breyha.
567 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
568 comma-sep list; embedded commas doubled.
570 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
572 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
574 Report and patch from Dmitry Banschikov.
576 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
577 Removed SSL_clear() after SSL_new() which led to protocol negotiation
578 failures. We appear to now support TLS1.1+ with Exim.
580 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
581 lets Exim select keys and certificates based upon TLS SNI from client.
582 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
583 before an outbound SMTP session. New log_selector, +tls_sni.
585 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
586 NULL dereference. Report and patch from Alun Jones.
588 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
589 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
590 Not seeing resolver debug output on NetBSD, but suspect this is a
591 resolver implementation change.
593 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
594 Left warnings. Added "eximon gdb" invocation mode.
596 PP/21 Defaulting "accept_8bitmime" to true, not false.
598 PP/22 Added -bw for inetd wait mode support.
600 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
601 locate the relevant includes and libraries. Made this the default.
603 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
604 Bugzilla 1246, report and most of solution from Tomasz Kusy.
606 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
607 This may cause build issues on older platforms.
609 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
610 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
611 gnutls_require_mac & gnutls_require_protocols (no longer supported).
612 Added SNI support via GnuTLS too.
613 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
615 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
617 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
618 Applied second patch from Janne, implementing suggestion to default
619 multiple-strings-in-record handling to match SPF spec.
621 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
623 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
624 read-only, out of scope).
625 Patch from Wolfgang Breyha, report from Stuart Northfield.
627 PP/29 Fix three issues highlighted by clang analyser static analysis.
628 Only crash-plausible issue would require the Cambridge-specific
629 iplookup router and a misconfiguration.
630 Report from Marcin Mirosław.
632 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
634 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
635 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
636 As part of this, removing so much warning spew let me fix some minor
637 real issues in debug logging.
639 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
640 assignment on my part. Fixed.
642 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
643 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
644 Janne Snabb (who went above and beyond: thank you).
646 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
647 string otherwise requires a connection and a bunch more work and it's
648 relatively easy to get wrong. Should also expose TLS library linkage
651 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
652 64-bit ${eval} (JH/03).
654 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
655 GNU libc to support some of the 64-bit stuff, should not lead to
656 conflicts. Defined before os.h is pulled in, so if a given platform
657 needs to override this, it can.
659 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
660 protection layer was required, which is not implemented.
661 Bugzilla 1254, patch from Wolfgang Breyha.
663 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
664 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
665 tls_dhparam take prime identifiers. Also unbreak combination of
666 OpenSSL+DH_params+TLSSNI.
668 PP/39 Disable SSLv2 by default in OpenSSL support.
674 PP/01 Solaris build fix for Oracle's LDAP libraries.
675 Bugzilla 1109, patch from Stephen Usher.
677 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
679 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
682 TF/02 Fix a couple more cases where we did not log the error message
683 when unlink() failed. See also change 4.74-TF/03.
685 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
686 lock up or crash if it happened to be inside a call to libc when it
687 got a SIGUSR1 from exiwhat.
689 The SIGUSR1 handler appends the current process status to the process
690 log which is later printed by exiwhat. It used to use the general
691 purpose logging code to do this, but several functions it calls are
692 not safe for signals.
694 The new output code in the SIGUSR1 handler is specific to the process
695 log, and simple enough that it's easy to inspect for signal safety.
696 Removing some special cases also simplifies the general logging code.
697 Removing the spurious timestamps from the process log simplifies
700 TF/04 Improved ratelimit ACL condition.
702 The /noupdate option has been deprecated in favour of /readonly which
703 has clearer semantics. The /leaky, /strict, and /readonly update modes
704 are mutually exclusive. The update mode is no longer included in the
705 database key; it just determines when the database is updated. (This
706 means that when you upgrde Exim will forget old rate measurements.)
708 Exim now checks that the per_* options are used with an update mode that
709 makes sense for the current ACL. For example, when Exim is processing a
710 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
711 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
712 must specify per_mail/readonly. If you omit the update mode it defaults to
713 /leaky where that makes sense (as before) or /readonly where required.
715 The /noupdate option is now undocumented but still supported for
716 backwards compatibility. It is equivalent to /readonly except that in
717 ACLs where /readonly is required you may specify /leaky/noupdate or
718 /strict/noupdate which are treated the same as /readonly.
720 A useful new feature is the /count= option. This is a generalization
721 of the per_byte option, so that you can measure the throughput of other
722 aggregate values. For example, the per_byte option is now equivalent
723 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
725 The per_rcpt option has been generalized using the /count= mechanism
726 (though it's more complicated than the per_byte equivalence). When it is
727 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
728 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
729 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
730 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
731 Note that using per_rcpt with a non-readonly update mode in more than
732 one ACL will cause the recipients to be double-counted. (The per_mail
733 and per_byte options don't have this problem.)
735 The handling of very low rates has changed slightly. If the computed rate
736 is less than the event's count (usually one) then this event is the first
737 after a long gap. In this case the rate is set to the same as this event's
738 count, so that the first message of a spam run is counted properly.
740 The major new feature is a mechanism for counting the rate of unique
741 events. The new per_addr option counts the number of different
742 recipients that someone has sent messages to in the last time period. It
743 behaves like per_rcpt if all the recipient addresses are different, but
744 duplicate recipient addresses do not increase the measured rate. Like
745 the /count= option this is a general mechanism, so the per_addr option
746 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
747 example, measure the rate that a client uses different sender addresses
748 with the options per_mail/unique=$sender_address. There are further
749 details in the main documentation.
751 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
753 TF/06 Removed a few PCRE remnants.
755 TF/07 Automatically extract Exim's version number from tags in the git
756 repository when doing development or release builds.
758 PP/02 Raise smtp_cmd_buffer_size to 16kB.
759 Bugzilla 879. Patch from Paul Fisher.
761 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
762 Heavily based on revision 40f9a89a from Simon Arlott's tree.
765 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
767 PP/05 Variable $av_failed, true if the AV scanner deferred.
768 Bugzilla 1078. Patch from John Horne.
770 PP/06 Stop make process more reliably on build failure.
771 Bugzilla 1087. Patch from Heiko Schlittermann.
773 PP/07 Make maildir_use_size_file an _expandable_ boolean.
774 Bugzilla 1089. Patch from Heiko Schlittermann.
776 PP/08 Handle ${run} returning more data than OS pipe buffer size.
777 Bugzilla 1131. Patch from Holger Weiß.
779 PP/09 Handle IPv6 addresses with SPF.
780 Bugzilla 860. Patch from Wolfgang Breyha.
782 PP/10 GnuTLS: support TLS 1.2 & 1.1.
784 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
787 PP/11 match_* no longer expand right-hand-side by default.
788 New compile-time build option, EXPAND_LISTMATCH_RHS.
789 New expansion conditions, "inlist", "inlisti".
791 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
793 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
795 PP/14 fix log_write() format string regression from TF/03.
796 Bugzilla 1152. Patch from Dmitry Isaikin.
802 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
804 PP/02 Harmonised TLS library version reporting; only show if debugging.
805 Layout now matches that introduced for other libraries in 4.74 PP/03.
807 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
809 PP/04 New "dns_use_edns0" global option.
811 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
814 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
815 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
817 TK/01 Updated PolarSSL code to 0.14.2.
818 Bugzilla 1097. Patch from Andreas Metzler.
820 PP/07 Catch divide-by-zero in ${eval:...}.
823 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
826 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
827 format-string attack -- SECURITY: remote arbitrary code execution.
829 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
830 time unintentionally subject to list matching rules, letting the header
831 cause arbitrary Exim lookups (of items which can occur in lists, *not*
832 arbitrary string expansion). This allowed for information disclosure.
834 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
835 INT_MIN/-1 -- value coerced to INT_MAX.
841 NM/01 Workround for PCRE version dependency in version reporting
844 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
845 This fixes portability to compilers other than gcc, notably
846 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
848 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
849 makefiles for portability to HP-UX and POSIX correctness.
851 PP/01 Permit LOOKUP_foo enabling on the make command-line.
852 Also via indented variable definition in the Makefile.
853 (Debugging by Oliver Heesakkers).
855 PP/02 Restore caching of spamd results with expanded spamd_address.
856 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
858 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
859 Improves build reliability. Fix from: Frank Elsner
861 NM/02 Fix wide character breakage in the rfc2047 coding
862 Fixes bug 1064. Patch from Andrey N. Oktyabrski
864 NM/03 Allow underscore in dnslist lookups
865 Fixes bug 1026. Patch from Graeme Fowler
867 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
868 Code patches from Adam Ciarcinski of NetBSD.
870 NM/04 Fixed exiqgrep to cope with mailq missing size issue
873 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
874 is logged, to avoid truncation. Patch from John Horne.
876 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
877 Patch from Jakob Hirsch.
879 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
880 SQL string expansion failure details.
881 Patch from Andrey Oktyabrski.
883 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
884 Patch from Simon Arlott.
886 PP/09 New lookups functionality failed to compile on old gcc which rejects
887 extern declarations in function scope.
888 Patch from Oliver Fleischmann
890 PP/10 Use sig_atomic_t for flags set from signal handlers.
891 Check getgroups() return and improve debugging.
892 Fixed developed for diagnosis in bug 927 (which turned out to be
895 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
896 Patch from Mark Zealey.
898 PP/12 Bugzilla 1056: Improved spamd server selection.
899 Patch from Mark Zealey.
901 PP/13 Bugzilla 1086: Deal with maildir quota file races.
902 Based on patch from Heiko Schlittermann.
904 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
905 Patch from Uwe Doering, sign-off by Michael Haardt.
907 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
908 variable declaration deep within a block. Bug and patch from
911 PP/15 lookups-Makefile IRIX compatibilty coercion.
913 PP/16 Make DISABLE_DKIM build knob functional.
915 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
916 Patch by Simon Arlott
918 TF/03 Fix valgrind.h portability to C89 compilers that do not support
919 variable argument macros. Our copy now differs from upstream.
925 TF/01 Failure to get a lock on a hints database can have serious
926 consequences so log it to the panic log.
928 TF/02 Log LMTP confirmation messages in the same way as SMTP,
929 controlled using the smtp_confirmation log selector.
931 TF/03 Include the error message when we fail to unlink a spool file.
933 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
934 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
935 for maintaining out-of-tree patches for some time.
937 PP/01 Bugzilla 139: Documentation and portability issues.
938 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
939 Handle per-OS dynamic-module compilation flags.
941 PP/02 Let /dev/null have normal permissions.
942 The 4.73 fixes were a little too stringent and complained about the
943 permissions on /dev/null. Exempt it from some checks.
944 Reported by Andreas M. Kirchwitz.
946 PP/03 Report version information for many libraries, including
947 Exim version information for dynamically loaded libraries. Created
948 version.h, now support a version extension string for distributors
949 who patch heavily. Dynamic module ABI change.
951 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
952 privilege escalation vulnerability whereby the Exim run-time user
953 can cause root to append content of the attacker's choosing to
956 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
959 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
960 If dropping privileges for untrusted macros, we disabled normal logging
961 on the basis that it would fail; for the Exim run-time user, this is not
962 the case, and it resulted in successful deliveries going unlogged.
963 Fixed. Reported by Andreas Metzler.
969 PP/01 Date: & Message-Id: revert to normally being appended to a message,
970 only prepend for the Resent-* case. Fixes regression introduced in
971 Exim 4.70 by NM/22 for Bugzilla 607.
973 PP/02 Include check_rfc2047_length in configure.default because we're seeing
974 increasing numbers of administrators be bitten by this.
976 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
978 PP/03 Bugzilla 994: added openssl_options main configuration option.
980 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
982 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
984 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
986 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
987 without a peer certificate, leading to a segfault because of an
988 assumption that peers always have certificates. Be a little more
989 paranoid. Problem reported by Martin Tscholak.
991 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
992 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
993 NB: ClamAV planning to remove STREAM in "middle of 2010".
994 CL also introduces -bmalware, various -d+acl logging additions and
995 more caution in buffer sizes.
997 PP/09 Implemented reverse_ip expansion operator.
999 PP/10 Bugzilla 937: provide a "debug" ACL control.
1001 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1003 PP/12 Bugzilla 973: Implement --version.
1005 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1007 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1009 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1011 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1012 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1013 ignore trailing whitespace.
1015 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1017 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1020 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1021 Notification from Dr Andrew Aitchison.
1023 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1024 ExtendedDetectionInfo response format.
1025 Notification from John Horne.
1027 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1030 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1031 XSL and documented dependency on system catalogs, with examples of how
1034 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1037 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1038 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1039 configuration file which is writeable by the Exim user or group.
1041 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1042 of configuration files to cover files specified with the -C option if
1043 they are going to be used with root privileges, not just the default
1046 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1047 option (effectively making it always true).
1049 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1050 files to be used while preserving root privileges.
1052 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1053 that rogue child processes cannot use them.
1055 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1056 run-time user, instead of root.
1058 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1059 Exim run-time user without dropping privileges.
1061 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1062 result string, instead of calling string_vformat() twice with the same
1065 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1066 for other users. Others should always drop root privileges if they use
1067 -C on the command line, even for a whitelisted configure file.
1069 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1071 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1077 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1078 $data_path, and $header_path variables; fixed documentation bugs and
1081 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1082 exipick to access non-standard spools, including the "frozen" queue
1085 NM/01 Bugzilla 965: Support mysql stored procedures.
1086 Patch from Alain Williams
1088 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1090 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1091 Patch from Andreas Metzler
1093 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1094 Patch from Kirill Miazine
1096 NM/05 Bugzilla 671: Added umask to procmail example.
1098 JJ/03 installed exipick 20100323.0, fixing doc bug
1100 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1101 directory. Notification and patch from Dan Rosenberg.
1103 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1105 TK/02 Improve log output when DKIM signing operation fails.
1107 MH/01 Treat the transport option dkim_domain as a colon separated
1108 list, not as a single string, and sign the message with each element,
1109 omitting multiple occurences of the same signer.
1111 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1112 Bugzilla 985, 986. Patch by Simon Arlott
1114 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1115 Patch by Simon Arlott
1117 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1118 MBX locking. Notification from Dan Rosenberg.
1124 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1126 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1128 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1130 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1132 NM/04 Bugzilla 915: Flag broken perl installation during build.
1138 TK/01 Added patch by Johannes Berg that expands the main option
1139 "spamd_address" if it starts with a dollar sign.
1141 TK/02 Write list of recipients to X-Envelope-Sender header when building
1142 the mbox-format spool file for content scanning (suggested by Jakob
1145 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1146 (http://www.dcc-servers.net/) support via dccifd. Activated by
1147 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1149 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1150 by Mark Daniel Reidel <mr@df.eu>.
1152 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1153 When building exim an external PCRE library is now needed -
1154 PCRE is a system library on the majority of modern systems.
1155 See entry on PCRE_LIBS in EDITME file.
1157 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1158 conversation. Added nologin parameter to request.
1159 Patch contributed by Kirill Miazine.
1161 TF/01 Do not log submission mode rewrites if they do not change the address.
1163 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1165 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1166 log files in place. Contributed by Roberto Lima.
1168 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1170 TF/03 Bugzilla 615: When checking the local_parts router precondition
1171 after a local_part_suffix or local_part_prefix option, Exim now
1172 does not use the address's named list lookup cache, since this
1173 contains cached lookups for the whole local part.
1175 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1176 Robert Millan. Documentation is in experimental-spec.txt.
1178 TF/04 Bugzilla 668: Fix parallel build (make -j).
1180 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1182 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1183 Patch provided by Jan Srzednicki.
1185 TF/05 Leading white space used to be stripped from $spam_report which
1186 wrecked the formatting. Now it is preserved.
1188 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1189 that they are available at delivery time.
1191 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1193 TF/08 TLS error reporting now respects the incoming_interface and
1194 incoming_port log selectors.
1196 TF/09 Produce a more useful error message if an SMTP transport's hosts
1197 setting expands to an empty string.
1199 NM/06 Bugzilla 744: EXPN did not work under TLS.
1200 Patch provided by Phil Pennock.
1202 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1203 Patch provided by Richard Godbee.
1205 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1206 acl_smtp_notquit, added index entry.
1208 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1209 Patch provided by Eugene Bujak.
1211 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1212 accept(). Patch provided by Maxim Dounin.
1214 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1215 Patch provided by Phil Pennock.
1217 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1219 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1220 Patch provided by Brad "anomie" Jorsch.
1222 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1223 Patch provided by Dean Brooks.
1225 TK/05 Add native DKIM support (does not depend on external libraries).
1227 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1228 Patch provided by Graeme Fowler.
1230 NM/16 Bugzilla 851: Documentation example syntax fix.
1232 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1234 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1237 NM/19 Bugzilla 745: TLS version reporting.
1238 Patch provided by Phil Pennock.
1240 NM/20 Bugzilla 167: bool: condition support.
1241 Patch provided by Phil Pennock.
1243 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1244 clients. Patch provided by Phil Pennock.
1246 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1247 Patch provided by Brad "anomie" Jorsch.
1249 NM/23 Bugzilla 687: Fix misparses in eximstats.
1250 Patch provided by Heiko Schlittermann.
1252 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1253 Patch provided by Heiko Schlittermann.
1255 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1256 plus update to original patch.
1258 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1260 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1261 Patch provided by David Brownlee.
1263 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1265 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1267 NM/30 Bugzilla 888: TLS documentation bugfixes.
1269 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1271 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1272 Unlike the original bugzilla I have changed all shell scripts in src tree.
1274 NM/33 Bugzilla 898: Transport filter timeout fix.
1275 Patch by Todd Rinaldo.
1277 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1278 Patch by Serge Demonchaux.
1280 NM/35 Bugzilla 39: Base64 decode bug fixes.
1281 Patch by Jakob Hirsch.
1283 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1285 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1287 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1289 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1295 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1296 ALT-N's libdkim that I have put here:
1297 http://duncanthrax.net/exim-experimental/
1299 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1300 were called 'true' and it seems that C99 defines that as a reserved
1301 keyword to be used with 'bool' variable types. That means you could
1302 not include C99-style headers which use bools without triggering
1303 build errors in sieve.c.
1305 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1306 as mailq or other aliases. Changed the --help handling significantly
1307 to do whats expected. exim_usage() emits usage/help information.
1309 SC/01 Added the -bylocaldomain option to eximstats.
1311 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1313 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1315 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1321 PH/01 Another patch from the Sieve maintainer.
1323 PH/02 When an IPv6 address is converted to a string for single-key lookup
1324 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1325 dots are used instead of colons so that keys in lsearch files need not
1326 contain colons. This was done some time before quoting was made available
1327 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1328 (notated using the quote facility) so as to distinguish them from IPv4
1329 keys. This meant that lookups for IP addresses in host lists did not work
1330 for iplsearch lookups.
1332 This has been fixed by arranging for IPv6 addresses to be expressed with
1333 colons if the lookup type is iplsearch. This is not incompatible, because
1334 previously such lookups could never work.
1336 The situation is now rather anomolous, since one *can* have colons in
1337 ordinary lsearch keys. However, making the change in all cases is
1338 incompatible and would probably break a number of configurations.
1340 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1343 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1344 conversion specification without a maximum field width, thereby enabling
1345 a rogue spamd server to cause a buffer overflow. While nobody in their
1346 right mind would setup Exim to query an untrusted spamd server, an
1347 attacker that gains access to a server running spamd could potentially
1348 exploit this vulnerability to run arbitrary code as the Exim user.
1350 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1351 $primary_hostname instead of what libspf2 thinks the hosts name is.
1353 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1354 a directory entry by the name of the lookup key. Previously, if a
1355 symlink pointed to a non-existing file or a file in a directory that
1356 Exim lacked permissions to read, a lookup for a key matching that
1357 symlink would fail. Now it is enough that a matching directory entry
1358 exists, symlink or not. (Bugzilla 503.)
1360 PH/03 The body_linecount and body_zerocount variables are now exported in the
1363 PH/04 Added the $dnslist_matched variable.
1365 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1366 This means they are set thereafter only if the connection becomes
1369 PH/06 Added the client_condition to authenticators so that some can be skipped
1370 by clients under certain conditions.
1372 PH/07 The error message for a badly-placed control=no_multiline_responses left
1373 "_responses" off the end of the name.
1375 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1377 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1378 (without spaces) instead of just copying the configuration text.
1380 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1382 PH/11 Added $max_received_linelength.
1384 PH/12 Added +ignore_defer and +include_defer to host lists.
1386 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1387 way in which PCRE > 7.0 is built.
1389 PH/14 Implemented queue_only_load_latch.
1391 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1392 MAIL command. The effect was to mangle the value on 64-bit systems.
1394 PH/16 Another patch from the Sieve maintainer.
1396 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1398 PH/18 If a system quota error occurred while trying to create the file for
1399 a maildir delivery, the message "Mailbox is full" was not appended to the
1400 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1401 only to a quota excession during the actual writing of the file.
1403 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1404 characters?) which causes problems in log lines. The DN values are now
1405 passed through string_printing() before being added to log lines.
1407 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1408 and InterBase are left for another time.)
1410 PH/21 Added message_body_newlines option.
1412 PH/22 Guard against possible overflow in moan_check_errorcopy().
1414 PH/23 POSIX allows open() to be a macro; guard against that.
1416 PH/24 If the recipient of an error message contained an @ in the local part
1417 (suitably quoted, of course), incorrect values were put in $domain and
1418 $local_part during the evaluation of errors_copy.
1424 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1425 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1428 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1429 issue a MAIL command.
1431 PH/02 In an ACL statement such as
1433 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1435 if a client was not listed at all, or was listed with a value other than
1436 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1437 the condition was not true (as it should be), so access was not denied.
1438 The bug was that the ! inversion was incorrectly passed on to the second
1439 item. This has been fixed.
1441 PH/03 Added additional dnslists conditions == and =& which are different from
1442 = and & when the dns lookup returns more than one IP address.
1444 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1445 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1447 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1448 FSYNC, which compiles an option called disable_fsync that allows for
1449 bypassing fsync(). The documentation is heavily laced with warnings.
1451 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1453 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1454 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1455 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1456 including adding "make clean"; (3) Added -fPIC when compiling the test
1457 dynamically loaded module, to get rid of a warning.
1459 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1460 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1461 = 0s. The bug is otherwise harmless.
1463 PH/07 There was a bug in the dovecot authenticator such that the value of
1464 $auth1 could be overwritten, and so not correctly preserved, after a
1465 successful authentication. This usually meant that the value preserved by
1466 the server_setid option was incorrect.
1468 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1470 PH/09 Installed PCRE release 7.0.
1472 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1473 run for batched SMTP input. It is now run at the start of every message
1474 in the batch. While fixing this I discovered that the process information
1475 (output by running exiwhat) was not always getting set for -bs and -bS
1476 input. This is fixed, and it now also says "batched" for BSMTP.
1478 PH/11 Added control=no_pipelining.
1480 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1481 patch, slightly modified), and move the expansion of helo_data till after
1482 the connection is made in the smtp transport (so it can use these
1485 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1487 PH/14 Added log_selector = +pid.
1489 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1491 PH/16 Add ${if forany and ${if forall.
1493 PH/17 Added dsn_from option to vary the From: line in DSNs.
1495 PH/18 Flush SMTP output before performing a callout, unless control =
1496 no_callout_flush is set.
1498 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1499 was true (the default) a successful delivery failed to delete the retry
1500 item, thus causing premature timeout of the address. The bug is now
1503 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1505 PH/21 Long custom messages for fakedefer and fakereject are now split up
1506 into multiline reponses in the same way that messages for "deny" and
1507 other ACL rejections are.
1509 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1510 with slight modification.
1512 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1513 draft, changing the syntax and factoring some duplicate code".
1515 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1516 for deliveries of the second and subsequent messages over the same SMTP
1519 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1520 ${reduce, with only minor "tidies".
1522 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1524 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1525 expansion side effects.
1527 PH/27 When a message times out after an over-quota error from an Exim-imposed
1528 quota, the bounce message says "mailbox is full". This message was not
1529 being given when it was a system quota that was exceeded. It now should
1532 MH/03 Made $recipients available in local_scan(). local_scan() already has
1533 better access to the recipient list through recipients_list[], but
1534 $recipients can be useful in postmaster-provided expansion strings.
1536 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1537 in the case of a MAIL command with additional options following the
1538 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1539 were accidentally chopped off.
1541 PH/29 SMTP synchronization checks are implemented when a command is read -
1542 there is a check that no more input is waiting when there shouldn't be
1543 any. However, for some commands, a delay in an ACL can mean that it is
1544 some time before the response is written. In this time, more input might
1545 arrive, invalidly. So now there are extra checks after an ACL has run for
1546 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1547 pipelining has not been advertised.
1549 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1551 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1552 returned in the SMTP response when smtp_return_error_details was false.
1553 This has been fixed.
1555 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1556 instead of the C I/O that was originally supplied, because problems were
1557 reported on Solaris.
1559 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1560 Exim which did not show up earlier: it was assuming that a call to
1561 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1562 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1563 was a macro that became an assignment, so it seemed to work. This has
1564 changed to a proper function call with a void return, hence the compile
1565 error. Exim's code has been fixed.
1567 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1570 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1572 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1574 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1575 because it turns out to be too big or there is a timeout) there is no
1576 "Completed" line in the log. When some messages of this type were
1577 selected by exigrep, they were listed as "not completed". Others were
1578 picked up by some special patterns. I have improved the selection
1579 criteria to be more general.
1581 PH/38 The host_find_failed option in the manualroute router can now be set
1582 to "ignore", to completely ignore a host whose IP address cannot be
1583 found. If all hosts are ignored, the behaviour is controlled by the new
1584 host_all_ignored option.
1586 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1587 homing or because of multiple MX records with /mx) generated more than
1588 one IP address, and the following item turned out to be the local host,
1589 all the secondary addresses of the first item were incorrectly removed
1590 from the list, along with the local host and any following hosts (which
1591 is what is supposed to happen).
1593 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1594 whoever called Exim into the -H file. In the case of the daemon it was
1595 behaving confusingly. When first started, it used values for whoever
1596 started the daemon, but after a SIGHUP it used the Exim user (because it
1597 calls itself on a restart). I have changed the code so that it now always
1600 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1601 message are rejected with the same error (e.g. no authentication or bad
1602 sender address), and a DATA command is nevertheless sent (as can happen
1603 with PIPELINING or a stupid MUA), the error message that was given to the
1604 RCPT commands is included in the rejection of the DATA command. This is
1605 intended to be helpful for MUAs that show only the final error to their
1608 PH/42 Another patch from the Sieve maintainer.
1610 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1611 Eximstats - Fixed some broken HTML links and added missing column headers
1613 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1614 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1616 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1618 PH/43 Yet another patch from the Sieve maintainer.
1620 PH/44 I found a way to check for a TCP/IP connection going away before sending
1621 the response to the final '.' that terminates a message, but only in the
1622 case where the client has not sent further data following the '.'
1623 (unfortunately, this is allowed). However, in many cases there won't be
1624 any further data because there won't be any more messages to send. A call
1625 to select() can be used: if it shows that the input is "ready", there is
1626 either input waiting, or the socket has been closed. An attempt to read
1627 the next input character can distinguish the two cases. Previously, Exim
1628 would have sent an OK response which the client would never have see.
1629 This could lead to message repetition. This fix should cure that, at
1630 least in a lot of common cases.
1632 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1633 advertised in response to EHLO.
1639 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1640 fixed by 4.65/MH/01 (is this a record?) are fixed:
1642 (i) An empty string was always treated as zero by the numeric comparison
1643 operators. This behaviour has been restored.
1645 (ii) It is documented that the numeric comparison operators always treat
1646 their arguments as decimal numbers. This was broken in that numbers
1647 starting with 0 were being interpreted as octal.
1649 While fixing these problems I realized that there was another issue that
1650 hadn't been noticed. Values of message_size_limit (both the global option
1651 and the transport option) were treated as octal if they started with 0.
1652 The documentation was vague. These values are now always treated as
1653 decimal, and I will make that clear in the documentation.
1659 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1660 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1663 MH/01 Don't check that the operands of numeric comparison operators are
1664 integers when their expansion is in "skipping" mode (fixes bug
1665 introduced by 4.64-PH/07).
1667 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1668 child addresses, Exim now panics and dies. Previously, because the count
1669 is held in a short int, deliveries were likely to be lost. As such a
1670 large number of recipients for a single message is ridiculous
1671 (performance will be very, very poor), I have chosen to impose a limit
1672 rather than extend the field.
1678 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1679 leftover -K file (the existence of which was triggered by #402).
1680 While we were at it, introduced process PID as part of the -K
1681 filename. This should rule out race conditions when creating
1684 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1685 processing considerably. Previous code took too long for large mails,
1686 triggering a timeout which in turn triggers #401.
1688 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1689 in the DK code in transports.c. sendfile() is not really portable,
1690 hence the _LINUX specificness.
1692 TF/01 In the add_headers option to the mail command in an Exim filter,
1693 there was a bug that Exim would claim a syntax error in any
1694 header after the first one which had an odd number of characters
1697 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1698 callout verification, Exim cached a "reject" for the entire domain. This
1699 is correct for most verifications, but it is not correct for a recipient
1700 verification with use_sender or use_postmaster set, because in that case
1701 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1702 case of MAIL FROM:<> rejection from other early rejections (e.g.
1703 rejection of HELO). When verifying a recipient using a non-null MAIL
1704 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1705 Whatever the result of the callout, the value of the domain cache is
1706 left unchanged (for any other kind of callout, getting as far as trying
1707 RCPT means that the domain itself is ok).
1709 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1712 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1713 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1714 session, the daemon ignores ECONNECT errors and logs others; it now
1715 ignores EPIPE as well.
1717 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1718 (quoted-printable decoding).
1720 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1721 later the small subsequent patch to fix an introduced bug.
1723 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1725 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1727 PH/08 An error is now given if message_size_limit is specified negative.
1729 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1730 to be given (somewhat) arbitrary names.
1732 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1735 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1736 miscellaneous code fixes
1738 PH/10 Added the log_reject_target ACL modifier to specify where to log
1741 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1742 hostname. This is wrong, because it relates to the incoming message (and
1743 probably the interface on which it is arriving) and not to the outgoing
1744 callout (which could be using a different interface). This has been
1745 changed to use the value of the helo_data option from the smtp transport
1746 instead - this is what is used when a message is actually being sent. If
1747 there is no remote transport (possible with a router that sets up host
1748 addresses), $smtp_active_hostname is used.
1750 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1751 tweaks were necessary in order to get it to work (see also 21 below):
1752 (a) The code assumed that strncpy() returns a negative number on buffer
1753 overflow, which isn't the case. Replaced with Exim's string_format()
1755 (b) There were several signed/unsigned issues. I just did the minimum
1756 hacking in of casts. There is scope for a larger refactoring.
1757 (c) The code used strcasecmp() which is not a standard C function.
1758 Replaced with Exim's strcmpic() function.
1759 (d) The code set only $1; it now sets $auth1 as well.
1760 (e) A simple test gave the error "authentication client didn't specify
1761 service in request". It would seem that Dovecot has changed its
1762 interface. Fortunately there's a specification; I followed it and
1763 changed what the client sends and it appears to be working now.
1765 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1768 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1769 address A is aliased to B and C, where B exists and C does not. Without
1770 -v the output is "A verified" because verification stops after a
1771 successful redirection if more than one address is generated. However,
1772 with -v the child addresses are also verified. Exim was outputting "A
1773 failed to verify" and then showing the successful verification for C,
1774 with its parentage. It now outputs "B failed to verify", showing B's
1775 parentage before showing the successful verification of C.
1777 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1778 look up a TXT record in a specific list after matching in a combined
1781 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1782 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1783 they consult the DNS. I had assumed they would set it the way they
1784 wanted; and indeed my experiments on Linux seem to show that in some
1785 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1786 To be on the safe side, however, I have now made the interface to
1787 host_find_byname() similar to host_find_bydns(), with an argument
1788 containing the DNS resolver options. The host_find_byname() function now
1789 sets these options at its start, just as host_find_bydns() does. The smtp
1790 transport options dns_qualify_single and dns_search_parents are passed to
1791 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1792 of host_find_byname() use the default settings of RES_DEFNAMES
1793 (qualify_single) but not RES_DNSRCH (search_parents).
1795 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1796 spool_read_header() do less string testing, by means of a preliminary
1797 switch on the second character of optional "-foo" lines. (This is
1798 overdue, caused by the large number of possibilities that now exist.
1799 Originally there were few.) While I was there, I also converted the
1800 str(n)cmp tests so they don't re-test the leading "-" and the first
1801 character, in the hope this might squeeze out yet more improvement.
1803 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1804 flag allowing group syntax was set by the header_syntax check but not
1805 turned off, possible causing trouble later; (2) The flag was not being
1806 set at all for the header_verify test, causing "group"-style headers to
1807 be rejected. I have now set it in this case, and also caused header_
1808 verify to ignore an empty address taken from a group. While doing this, I
1809 came across some other cases where the code for allowing group syntax
1810 while scanning a header line wasn't quite right (mostly, not resetting
1811 the flag correctly in the right place). These bugs could have caused
1812 trouble for malformed header lines. I hope it is now all correct.
1814 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1815 with the "reply" argument non-NULL. The code, however (which originally
1816 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1817 but it didn't always do it. This confused somebody who was copying the
1818 code for some other use. I have removed all the tests.
1820 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1821 feature that was used to support insecure browsers during the U.S. crypto
1822 embargo. It requires special client support, and Exim is probably the
1823 only MTA that supported it -- and would never use it because real RSA is
1824 always available. This code has been removed, because it had the bad
1825 effect of slowing Exim down by computing (never used) parameters for the
1826 RSA_EXPORT functionality.
1828 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1829 authenticator to fail if there's a tab character in the incoming data
1830 (there should never be unless someone is messing about, as it's supposed
1831 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1832 if the connection is using TLS or if the remote IP is the same as the
1833 local IP, and the "valid-client-cert option" if a client certificate has
1836 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1837 authenticators. This can be used for authorization after authentication
1838 succeeds. (In the case of plaintext, it servers for both authentication
1841 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1842 if any retry times were supplied.
1844 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1845 connection, where there is no client IP address to check. In this
1846 situation, the verify now always succeeds.
1848 PH/25 Applied John Jetmore's -Mset patch.
1850 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1852 PH/27 In a string expansion for a processed (not raw) header when multiple
1853 headers of the same name were present, leading whitespace was being
1854 removed from all of them, but trailing whitespace was being removed only
1855 from the last one. Now trailing whitespace is removed from each header
1856 before concatenation. Completely empty headers in a concatenation (as
1857 before) are ignored.
1859 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1860 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1862 PH/29 [Removed. This was a change that I later backed out, and forgot to
1863 correct the ChangeLog entry (that I had efficiently created) before
1864 committing the later change.]
1866 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1867 address errors (4xx response to RCPT) over the same connection as other
1868 messages routed to the same hosts. Such deliveries are always "forced",
1869 so retry times are not inspected. This resulted in far too many retries
1870 for the affected addresses. The effect occurred only when there were more
1871 hosts than the hosts_max_try setting in the smtp transport when it had
1872 the 4xx errors. Those hosts that it had tried were not added to the list
1873 of hosts for which the message was waiting, so if all were tried, there
1874 was no problem. Two fixes have been applied:
1876 (i) If there are any address or message errors in an SMTP delivery, none
1877 of the hosts (tried or untried) are now added to the list of hosts
1878 for which the message is waiting, so the message should not be a
1879 candidate for sending over the same connection that was used for a
1880 successful delivery of some other message. This seems entirely
1881 reasonable: after all the message is NOT "waiting for some host".
1882 This is so "obvious" that I'm not sure why it wasn't done
1883 previously. Hope I haven't missed anything, but it can't do any
1884 harm, as the worst effect is to miss an optimization.
1886 (ii) If, despite (i), such a delivery is accidentally attempted, the
1887 routing retry time is respected, so at least it doesn't keep
1888 hammering the server.
1890 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1891 in ${readsocket because some servers need this prod.
1893 PH/32 Added some extra debug output when updating a wait-xxx database.
1895 PH/33 The hint "could be header name not terminated by colon", which has been
1896 given for certain expansion errors for a long time, was not being given
1897 for the ${if def:h_colon_omitted{... case.
1899 PH/34 The spec says: "With one important exception, whenever a domain list is
1900 being scanned, $domain contains the subject domain." There was at least
1901 one case where this was not true.
1903 PH/35 The error "getsockname() failed: connection reset by peer" was being
1904 written to the panic log as well as the main log, but it isn't really
1905 panic-worthy as it just means the connection died rather early on. I have
1906 removed the panic log writing for the ECONNRESET error when getsockname()
1909 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1910 runs only) independently of the message's sender address. This meant
1911 that, if the 4xx error was in fact related to the sender, a different
1912 message to the same recipient with a different sender could confuse
1913 things. In particualar, this can happen when sending to a greylisting
1914 server, but other circumstances could also provoke similar problems.
1915 I have changed the default so that the retry time for these errors is now
1916 based a combination of the sender and recipient addresses. This change
1917 can be overridden by setting address_retry_include_sender=false in the
1920 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1921 remote server are returned as part of bounce messages. This was not
1922 happening for LMTP over a pipe (the lmtp transport), but now it is the
1923 same for both kinds of LMTP.
1925 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1926 in header lines that were in fact CNAMEs. This is no longer the case.
1928 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1929 and queue runs started by the daemon processed all messages. This has
1930 been fixed so that -R and -S can now usefully be given with -q<time>.
1932 PH/40 Import PCRE release 6.7 (fixes some bugs).
1934 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1936 PH/42 Give an error if -q is specified more than once.
1938 PH/43 Renamed the variables $interface_address and $interface_port as
1939 $received_ip_address and $received_port, to make it clear that these
1940 values apply to message reception, and not to the outgoing interface when
1941 a message is delivered. (The old names remain recognized, of course.)
1943 PH/44 There was no timeout on the connect() call when using a Unix domain
1944 socket in the ${readsocket expansion. There now is.
1946 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1947 be meaningful with "accept".
1949 SC/01 Eximstats V1.43
1950 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1952 SC/02 Eximstats V1.44
1953 Use a glob alias rather than an array ref in the generated
1954 parser. This improves both readability and performance.
1956 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1957 Collect SpamAssassin and rejection statistics.
1958 Don't display local sender or destination tables unless
1959 there is data to show.
1960 Added average volumes into the top table text output.
1962 SC/04 Eximstats V1.46
1963 Collect data on the number of addresses (recipients)
1964 as well as the number of messages.
1966 SC/05 Eximstats V1.47
1967 Added 'Message too big' to the list of mail rejection
1968 reasons (thanks to Marco Gaiarin).
1970 SC/06 Eximstats V1.48
1971 Mainlog lines which have GMT offsets and are too short to
1972 have a flag are now skipped.
1974 SC/07 Eximstats V1.49 (Alain Williams)
1975 Added the -emptyok flag.
1977 SC/08 Eximstats V1.50
1978 Fixes for obtaining the IP address from reject messages.
1980 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1981 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1982 whitesspace changes from 4.64-PH/27
1984 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1987 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1988 are found, allow negative numbers in numeric criteria)
1990 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1992 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1993 to match changes made in 4.64-PH/43
1995 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1997 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1998 "domain" is actually an IP address, and force a failure. This locks out
1999 those revolvers/nameservers that support "A-for-A" lookups, in
2000 contravention of the specifications.
2002 PH/48 When a host name was looked up from an IP address, and the subsequent
2003 forward lookup of the name timed out, the host name was left in
2004 $sender_host_name, contrary to the specification.
2006 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2007 restricted to single-key lookups, Exim was not diagnosing an error if
2008 * or *@ was used with a query-style lookup.
2010 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2012 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2013 long ago, but noone interested enough thought of it. Let's just say that
2014 the "1.1" means that there are some new functions that weren't there at
2015 some point in the past.
2017 PH/51 Error processing for expansion failure of helo_data from an smtp
2018 transport during callout processing was broken.
2020 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2021 tested/used via the -bh/-bhc/-bs options.
2023 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2024 bug, fixed in subsequent PCRE releases).
2026 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2027 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2029 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2035 SC/01 Use a glob alias rather than an array ref in eximstats generated
2036 parser. This improves both readability and performance.
2038 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2039 Don't display local sender or destination tables in eximstats unless
2040 there is data to show.
2041 Added average volumes into the eximstats top table text output.
2043 SC/03 Collect data on the number of addresses (recipients) as well
2044 as the number of messages in eximstats.
2046 TF/01 Correct an error in the documentation for the redirect router. Exim
2047 does (usually) call initgroups() when daemonizing.
2049 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2050 with consistent privilege compared to when running as a daemon.
2052 TF/03 Note in the spec that $authenticated_id is not set for local
2053 submissions from trusted users.
2055 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2056 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2058 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2059 by adding some example configuration directives to the default
2060 configuration file. A little bit of work is required to uncomment the
2061 directives and define how usernames and passwords are checked, but
2062 there is now a framework to start from.
2064 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2065 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2066 without this. I don't know how relevant this is to other LDAP libraries.
2068 PH/02 Add the verb name to the "unknown ACL verb" error.
2070 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2072 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2074 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2075 directory not expanded when it should be if an expanded home directory
2076 was set for the address (which is overridden by the transport).
2078 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2081 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2082 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2083 because it is too late at that time, and has no effect.
2085 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2086 security issue with \' (bugzilla #107). I could not use the
2087 PQescapeStringConn() function, because it needs a PGconn value as one of
2090 PH/08 When testing addresses using -bt, indicate those final addresses that
2091 are duplicates that would not cause an additional delivery. At least one
2092 person was confused, thinking that -bt output corresponded to deliveries.
2093 (Suppressing duplicates isn't a good idea as you lose the information
2094 about possibly different redirections that led to the duplicates.)
2096 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2097 systems where poll() doesn't work, in particular OS X.
2099 PH/10 Added more information to debugging output for retry time not reached.
2101 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2102 operations in malware.c.
2104 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2107 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2108 syslog_duplication set false, log lines that would normally be written
2109 both the the main log and to the reject log were not written to syslog at
2112 PH/14 In the default configuration, change the use of "message" in ACL warn
2113 statements to "add_header".
2115 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2116 not followed by a command (e.g. "seen endif").
2118 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2119 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2122 PH/17 Added extra conditions to the default value of delay_warning_condition
2126 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2127 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2128 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2131 The Auto-Submitted: and various List- headers are standardised, whereas I
2132 don't think Precedence: ever was.
2134 PH/18 Refactored debugging code in route_finduser() to show more information,
2135 in particular, the error code if getpwnam() issues one.
2137 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2138 This is apparently needed in addition to the PH/07 change above to avoid
2139 any possible encoding problems.
2141 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2142 but not after initializing Perl.
2144 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2145 output them only if debugging. By default they are written stderr,
2146 apparently, which is not desirable.
2148 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2151 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2154 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2156 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2157 authenticated or an ident call has been made. Suppress the default
2158 values for $authenticated_id and $authenticated_sender (but permit -oMai
2159 and -oMas) when testing with -bh.
2161 PH/24 Re-jigged the order of the tests in the default configuration so that the
2162 tests for valid domains and recipients precede the DNS black list and CSA
2163 tests, on the grounds that those ones are more expensive.
2165 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2166 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2167 This bug exists in every version of Exim that I still have, right back to
2170 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2171 However, an attempt to turn on case-sensitivity in a regex key by
2172 including (?-i) didn't work because the subject string was already
2173 lowercased, and the effects were non-intuitive. It turns out that a
2174 one-line patch can be used to allow (?-i) to work as expected.
2180 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2181 other effects) broke the use of negated acl sub-conditions.
2183 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2186 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2187 "Deny" causes Exim to reject the incoming connection with a 554 error.
2188 Unfortunately, if there is a major crisis, such as a disk failure,
2189 tcp-wrappers gives "deny", whereas what one would like would be some
2190 kind of temporary error. A kludge has been added to help with this.
2191 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2192 554 error is used if errno is still zero or contains ENOENT (which occurs
2193 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2196 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2198 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2199 errors. Otherwise a message that provokes a temporary error (when other
2200 messages do not) can cause a whole host to time out.
2202 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2203 addresses were routed directly to files or pipes from a redirect router.
2204 File deliveries just didn't batch; pipe deliveries might have suffered
2207 PH/06 A failure to get a lock for a hints database would erroneously always say
2208 "Failed to get write lock", even when it was really a read lock.
2210 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2211 of 0600. This has been changed to use the value of the lockfile_mode
2212 option (which defaults to 0600).
2214 PH/08 Applied small patch from the Sieve maintainer.
2216 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2217 folder from quota calculations, a direct delivery into this folder messed
2218 up the contents of the maildirsize file. This was because the regex was
2219 used only to exclude .Trash (or whatever) when the size of the mailbox
2220 was calculated. There was no check that a delivery was happening into an
2221 excluded directory. This bug has been fixed by ignoring all quota
2222 processing for deliveries into excluded directories.
2224 PH/10 Added the maildirfolder_create_regex option to appendfile.
2230 PH/01 The code for finding all the local interface addresses on a FreeBSD
2231 system running IPv6 was broken. This may well have applied to all BSD
2232 systems, as well as to others that have similar system calls. The broken
2233 code found IPv4 interfaces correctly, but gave incorrect values for the
2234 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2235 that it would not match correctly against @[] and not recognize the IPv6
2238 PH/02 The ipliteral router was not recognizing addresses of the form user@
2239 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2241 PH/03 Added disable_ipv6.
2243 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2244 decoded form, because it is most often used to construct To: headers
2245 lines in autoreplies, and the decoded form may well be syntactically
2246 invalid. However, $reply_address has leading white space removed, and all
2247 newlines turned into spaces so that the autoreply transport does not
2250 PH/05 If group was specified without a user on a router, and no group or user
2251 was specified on a transport, the group from the router was ignored.
2253 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2254 for visible compile-time settings that can be used to change these
2255 numbers, for those that want even more. Backwards compatibility with old
2256 spool files has been maintained. However, going back to a previous Exim
2257 release will lost any variables that are in spool files.
2259 PH/07 Two small changes when running in the test harness: increase delay when
2260 passing a TCP/IP connection to a new process, in case the original
2261 process has to generate a bounce, and remove special handling of
2262 127.0.0.2 (sic), which is no longer necessary.
2264 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2265 be the same on different OS.
2267 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2270 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2271 whether --show-vars was specified or not
2273 JJ/02 exipick: Added support for new ACL variable spool format introduced
2276 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2277 syntactically invalid From: or Reply-to: line, and a filter used this to
2278 generate an autoreply, and therefore failed to obtain an address for the
2279 autoreply, Exim could try to deliver to a non-existent relative file
2280 name, causing unrelated and misleading errors. What now happens is that
2281 it logs this as a hard delivery error, but does not attempt to create a
2284 PH/11 The exinext utility has a -C option for testing purposes, but although
2285 the given file was scanned by exinext itself; it wasn't being passed on
2286 when Exim was called.
2288 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2289 an end-of-file indication when reading a command response.
2291 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2292 compiled. In many other places in Exim, IPv6 addresses are always
2293 recognized, so I have changed this. It also means that IPv4 domain
2294 literals of the form [IPV4:n.n.n.n] are now always recognized.
2296 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2297 used if the router is not running as root, for example, when verifying at
2298 ACL time, or when using -bh. The debugging output from this situation was
2299 non-existent - all you got was a failure to exec. I have made two
2302 (a) Failures to set uid/gid, the current directory, or a process leader
2303 in a subprocess such as that created by queryprogram now generate
2304 suitable debugging ouput when -d is set.
2306 (b) The queryprogram router detects when it is not running as root,
2307 outputs suitable debugging information if -d is set, and then runs
2308 the subprocess without attempting to change uid/gid.
2310 PH/15 Minor change to Makefile for building test_host (undocumented testing
2313 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2314 additional section of a DNS packet that returns MX or SRV records.
2315 Instead, it always explicitly searches for A/AAAA records. This avoids
2316 major problems that occur when a DNS server includes only records of one
2317 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2318 fixed another bug: if SRV records were looked up and the corresponding
2319 address records were *not* found in the additional section, the port
2320 values from the SRV records were lost.
2322 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2323 using the correct key (the original address) when searching the retry
2324 rules in order to find which one to use for generating the retry hint.
2326 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2327 adding the default one. Similarly, if it contains a Reply-To: header, the
2328 errors_reply_to option, if set, is not used.
2330 PH/19 When calculating a retry time, Exim used to measure the "time since
2331 failure" by looking at the "first failed" field in the retry record. Now
2332 it does not use this if it is later than than the arrival time of the
2333 message. Instead it uses the arrival time. This makes for better
2334 behaviour in cases where some deliveries succeed, thus re-setting the
2335 "first failed" field. An example is a quota failure for a huge message
2336 when small messages continue to be delivered. Without this change, the
2337 "time since failure" will always be short, possible causing more frequent
2338 delivery attempts for the huge message than are intended.
2339 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2341 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2342 $1, $2, $3) because the numerical variables can be reset during some
2343 expansion items (e.g. "match"), thereby losing the authentication data.
2345 PH/21 Make -bV show the size of off_t variables so that the test suite can
2346 decide whether to run tests for quotas > 2G.
2348 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2349 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2350 filecount value is greater than 2G or if a quota value is greater than 2G
2351 on a system where the size of off_t is not greater than 4, a panic error
2354 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2355 never match. The debug and -bh output now contains an explicit error
2356 message indicating a malformed IPv4 address or mask.
2358 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2359 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2360 PH/23 above applies.
2362 PH/25 Do not write to syslog when running in the test harness. The only
2363 occasion when this arises is a failure to open the main or panic logs
2364 (for which there is an explicit test).
2366 PH/26 Added the /no_tell option to "control=freeze".
2368 PH/27 If a host name lookup failed very early in a connection, for example, if
2369 the IP address matched host_lookup and the reverse lookup yielded a name
2370 that did not have a forward lookup, an error message of the form "no IP
2371 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2372 could be logged. Now it outputs the IP address instead of "NULL".
2374 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2375 allows the sender and the authenticated sender to be set when
2376 submitting a message from within Exim. Since child_open_exim() is
2377 documented for local_scan(), the new function should be too.
2379 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2380 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2381 results in an empty string is now treated as unset.
2383 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2385 PH/31 Added sender_verify_fail logging option.
2387 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2388 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2389 tidied the source and removed it altogether.
2391 PH/33 When a queue run was abandoned because the load average was too high, a
2392 log line was always written; now it is written only if the queue_run log
2393 selector is set. In addition, the log line for abandonment now contains
2394 information about the queue run such as the pid. This is always present
2395 in "start" and "stop" lines but was omitted from the "abandon" line.
2397 PH/34 Omit spaces between a header name and the colon in the error message that
2398 is given when verify = headers_syntax fails (if there are lots of them,
2399 the message gets confusing).
2401 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2402 names, as there are now some PTR records that contain slashes. This check
2403 is only to protect against broken name servers that fall over on strange
2404 characters, so the fact that it applies to all lookups doesn't matter.
2406 PH/36 Now that the new test suite is complete, we can remove some of the
2407 special code in Exim that was needed for the old test suite. For example,
2408 sorting DNS records because real resolvers return them in an arbitrary
2409 order. The new test suite's fake resolver always returns records in the
2412 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2413 bounces) except when queue_only is set, to avoid logging races between
2414 the different processes.
2416 PH/38 Panic-die if .include specifies a non-absolute path.
2418 PH/39 A tweak to the "H" retry rule from its user.
2420 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2421 a label. They prevented compilation on older perls.
2423 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2424 a warning to be raised on newish perls.
2426 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2427 on queue. Changes to match documented behaviour of showing count of
2428 messages matching specified criteria.
2430 PH/40 Changed the default ident timeout from 30s to 5s.
2432 PH/41 Added support for the use of login_cap features, on those BSD systems
2433 that have them, for controlling the resources used by pipe deliveries.
2435 PH/42 The content-scanning code uses fopen() to create files in which to put
2436 message data. Previously it was not paying any attention to the mode of
2437 the files. Exim runs with umask(0) because the rest of the code creates
2438 files with open(), and sets the required mode explicitly. Thus, these
2439 files were ending up world-writeable. This was not a big issue, because,
2440 being within the spool directory, they were not world-accessible. I have
2441 created a function called modefopen, which takes an additional mode
2442 argument. It sets umask(777), creates the file, chmods it to the required
2443 mode, then resets the umask. All the relevant calls to fopen() in the
2444 content scanning code have been changed to use this function.
2446 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2447 to 24 hours. This avoids potential overflow problems when processing G
2448 and H retry rules. I suspect nobody ever tinkers with this value.
2450 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2452 PH/45 When the plaintext authenticator is running as a client, the server's
2453 challenges are checked to ensure they are valid base64 strings. By
2454 default, the authentication attempt is cancelled if an invalid string is
2455 received. Setting client_ignore_invalid_base64 true ignores these errors.
2456 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2457 they are received. Thus, the responses can be made to depend on the
2458 challenges. If an invalid string is ignored, an empty string is placed in
2461 PH/46 Messages that are created by the autoreply transport now contains a
2462 References: header, in accordance with RFCs 2822 and 3834.
2464 PH/47 Added authenticated_sender_force to the smtp transport.
2466 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2468 PH/49 Installed latest patch from the Sieve maintainer.
2470 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2471 was also set, the appendfile transport was unnecessarily scanning a
2472 directory of message files (e.g. for maildir delivery) to find the count
2473 of files (along with the size), even though it did not need this
2474 information. It now does the scan only if it needs to find either the
2475 size of the count of files.
2477 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2479 PH/52 Two bugs concerned with error handling when the smtp transport is
2482 (i) Exim was not creating retry information for temporary errors given
2483 for individual recipients after the DATA command when the smtp transport
2484 was used in LMTP mode. This meant that they could be retried too
2485 frequently, and not timed out correctly.
2487 (ii) Exim was setting the flag that allows error details to be returned
2488 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2489 recipients that were returned after the DATA command.
2491 PH/53 This is related to PH/52, but is more general: for any failing address,
2492 when detailed error information was permitted to be returned to the
2493 sender, but the error was temporary, then after the final timeout, only
2494 "retry timeout exceeded" was returned. Now it returns the full error as
2495 well as "retry timeout exceeded".
2497 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2498 do this, and (what is worse) MTAs that accept it.
2500 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2501 will now be deprecated.
2503 PH/56 New os.c-cygwin from the Cygwin maintainer.
2505 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2506 formats (previously only available in exim formats via -bpr, -bpru,
2507 and -bpra. Now also available in native and exiqgrep formats)
2509 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2510 with very large, slow to parse queues
2512 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2514 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2516 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2517 responses to authentication challenges, though it was showing the
2518 challenges; (ii) I've removed the CR characters from the debug output for
2521 PH/58 Allow for the insertion of a newline as well as a space when a string
2522 is turned into more than one encoded-word during RFC 2047 encoding. The
2523 Sieve code now uses this.
2525 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2526 data_4xx, lost_connection, tls_required.
2528 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2529 message was being sent as an SMTP response.
2531 PH/61 Add -l and -k options to exicyclog.
2533 PH/62 When verifying, if an address was redirected to one new address, so that
2534 verification continued, and the new address failed or deferred after
2535 having set something in $address_data, the value of $address_data was not
2536 passed back to the ACL. This was different to the case when no
2537 redirection occurred. The value is now passed back in both cases.
2539 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2540 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2541 use login_cap.h, so on its own it isn't the distinguishing feature. The
2542 new name refers directly to the setclassresources() function.
2544 PH/65 Added configuration files for NetBSD3.
2546 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2548 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2549 is preferred over IPv4.
2551 PH/68 The bounce_return_message and bounce_return_body options were not being
2552 honoured for bounces generated during the reception of non-SMTP messages.
2553 In particular, this applied to messages rejected by the ACL. This bug has
2554 been fixed. However, if bounce_return_message is true and bounce_return_
2555 body is false, the headers that are returned for a non-SMTP message
2556 include only those that have been read before the error was detected.
2557 (In the case of an ACL rejection, they have all been read.)
2559 PH/69 The HTML version of the specification is now built in a directory called
2560 spec_html instead of spec.html, because the latter looks like a path with
2561 a MIME-type, and this confuses some software.
2563 PH/70 Catch two compiler warnings in sieve.c.
2565 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2566 function verify_get_ident() calls ip_connect() to connect a socket, but
2567 if the "connect()" function timed out, ip_connect() used to close the
2568 socket. However, verify_get_ident() also closes the socket later, and in
2569 between Exim writes to the log, which may get opened at this point. When
2570 the socket was closed in ip_connect(), the log could get the same file
2571 descriptor number as the socket. This naturally causes chaos. The fix is
2572 not to close the socket in ip_connect(); the socket should be closed by
2573 the function that creates it. There was only one place in the code where
2574 this was missing, in the iplookup router, which I don't think anybody now
2575 uses, but I've fixed it anyway.
2577 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2578 well as to direct DNS lookups. Otherwise the handling of names in host
2579 lists is inconsistent and therefore confusing.
2585 PH/01 Two changes to the default runtime configuration:
2587 (1) Move the checks for relay_from_hosts and authenticated clients from
2588 after to before the (commented out) DNS black list checks.
2590 (2) Add control=submission to the relay_from_hosts and authenticated
2591 clients checks, on the grounds that messages accepted by these
2592 statements are most likely to be submissions.
2594 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2596 (1) Generate an error if the third argument for the ${prvs expansion is
2599 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2602 (3) Reset the variables that are obtained from the first argument of
2603 ${prvscheck and used in the second argument before leaving the code,
2604 because their memory is reclaimed, so using them afterwards may do
2607 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2608 one (it's much easier than Tom thought :-).
2610 (5) Because of (4), we can now allow for the use of $prvscheck_result
2611 inside the third argument.
2613 PH/03 For some reason, the default setting of PATH when running a command from
2614 a pipe transport was just "/usr/bin". I have changed it to
2617 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2618 anything to be listed in the output from -bV.
2620 PH/05 When a filter generated an autoreply, the entire To: header line was
2621 quoted in the delivery log line, like this:
2623 => >A.N.Other <ano@some.domain> <original@ddress> ...
2625 This has been changed so that it extracts the operative address. There
2626 may be more than one such address. If so, they are comma-separated, like
2629 => >ano@some.domain,ona@other.domain <original@ddress> ...
2631 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2632 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2633 not being looked up in the rDNS to get a host name, Exim was showing the
2634 IP address twice in Received: lines, even though the IP addresses were
2635 identical. For example:
2637 Received: from [1.2.3.4] (helo=[1.2.3.4])
2639 However, if the real host name was known, it was omitting the HELO data
2640 if it matched the actual IP address. This has been tidied up so that it
2641 doesn't show the same IP address twice.
2643 PH/07 When both +timestamp and +memory debugging was on, the value given by
2644 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2645 called by the debug printing, thereby overwriting the timestamp buffer.
2646 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2648 PH/08 When the original message was included in an autoreply transport, it
2649 always said "this is a copy of the message, including all the headers",
2650 even if body_only or headers_only was set. It now gives an appropriate
2653 PH/09 Applied a patch from the Sieve maintainer which:
2655 o fixes some comments
2656 o adds the (disabled) notify extension core
2657 o adds some debug output for the result of if/elsif tests
2658 o points to the current vacation draft in the documentation
2659 and documents the missing references header update
2663 o fixes a bug in processing the envelope test (when testing
2664 multiple envelope elements, the last element determinted the
2667 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2668 Electronic Mail") by including:
2670 Auto-submitted: auto-generated
2672 in the messages that it generates (bounce messages and others, such as
2673 warnings). In the case of bounce messages for non-SMTP mesages, there was
2674 also a typo: it was using "Auto_submitted" (underscore instead of
2675 hyphen). Since every message generated by Exim is necessarily in response
2676 to another message, thes have all been changed to:
2678 Auto-Submitted: auto-replied
2680 in accordance with these statements in the RFC:
2682 The auto-replied keyword:
2684 - SHOULD be used on messages sent in direct response to another
2685 message by an automatic process,
2687 - MUST NOT be used on manually-generated messages,
2689 - MAY be used on Delivery Status Notifications (DSNs) and Message
2690 Disposition Notifications (MDNs),
2692 - MUST NOT be used on messages generated by automatic or periodic
2693 processes, except for messages which are automatic responses to
2696 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2697 to the default Received: header definition.
2699 PH/12 Added log selector acl_warn_skipped (default on).
2701 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2702 variables because (a) they are in the wrong storage pool and (b) even if
2703 they were copied, it wouldn't work properly because of the caching.
2705 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2706 checking when decoding. Apparently there are clients that generate
2707 overlong encoded strings. Why am I not surprised?
2709 PH/15 If the first argument of "${if match_address" was not empty, but did not
2710 contain an "@" character, Exim crashed. Now it writes a panic log message
2711 and treats the condition as false.
2713 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2715 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2716 extension "envelope-auth". The code is finished and in agreement with
2717 other implementations, but there is no documentation so far and in fact,
2718 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2719 not changing the active code.
2721 Print executed "if" and "elsif" statements when debugging is used. This
2722 helps a great deal to understand what a filter does.
2724 Document more things not specified clearly in RFC3028. I had all this
2725 sorted out, when out of a sudden new issues came to my mind. Oops."
2727 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2730 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2731 canonical form (as documented). However, after a host name lookup from
2732 the IP address, check_host() was doing a simple string comparison with
2733 addresses acquired from the DNS when checking that the found name did
2734 have the original IP as one of its addresses. Since any found IPv6
2735 addresses are likely to be in abbreviated form, the comparison could
2736 fail. Luckily, there already exists a function for doing the comparison
2737 by converting both addresses to binary, so now that is used instead of
2738 the text comparison.
2740 PH/20 There was another similar case to PH/19, when a complete host name was
2741 given in a host list; looking up its IP address could give an abbreviated
2742 form, whereas the current host's name might or might not be abbreviated.
2743 The same fix has been applied.
2749 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2750 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2753 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2754 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2756 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2758 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2759 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2760 Also, for "Auto-Submitted", treat anything other than "no" as
2761 non-personal, in accordance with RFC 3834. (Previously it treated
2762 anything starting "auto-" as non-personal.)
2764 TF/01 The control=submission/name=... option had a problem with syntax
2765 errors if the name included a slash character. The /name= option
2766 now slurps the rest of the string, so it can include any characters
2767 but it must come last in the list of options (after /sender_retain
2770 PH/05 Some modifications to the interface to the fake nameserver for the new
2778 TK/01 Added the "success_on_redirect" address verification option. See
2779 NewStuff for rationale and an example.
2781 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2783 PH/02 Patch to exigrep to allow it to work on syslog lines.
2785 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2786 fread() to skip over the body file's header line, because in Cygwin the
2787 header line is locked and is inaccessible.
2789 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2790 co-exist for some time) to make it clear that it is the Exim ID that is
2791 referenced, not the Message-ID: header line.
2793 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2794 string_format() function, because snprintf() does not exist on all
2797 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2798 ${stat: expansion item.
2800 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2801 protocol synchronization error", to keep the pedants happy.
2803 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2804 well as for IRIX systems, when gcc is being used. See the host.c source
2807 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2809 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2812 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2813 transport and to the smtp transport in LMTP mode.
2815 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2817 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2818 run a filter in a subprocess. This could lead to confusion in subsequent
2819 lookups in the parent process. There should also be a search_tidyup() at
2820 the end of the subprocess.
2822 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2823 only if the host matched helo_try_verify_hosts, which caused the
2824 verification to occur when the EHLO/HELO command was issued. The ACL just
2825 tested the remembered result. Now, if a previous verification attempt has
2826 not happened, "verify = helo" does it there and then.
2828 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2830 TK/03 Fix log output including CR from clamd.
2832 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2833 exist provoked a memory error which could cause a segfault.
2835 PH/15 Installed PCRE 6.2
2837 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2839 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2840 of the problem. Specifically, suggested +O2 rather than +O1 for the
2843 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2845 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2846 record, the port was not correctly logged when the outgoing_port log
2847 selector was set (it logged the transort's default port).
2849 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2850 fallback_hosts, and "hosts" in the smtp transport.
2852 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2853 host errors such as "Connection refused".
2855 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2856 authentication with radiusclient 0.4.9:
2858 - Error returned from rc_read_config was caught wrongly
2859 - Username/password not passed on to radius server due to wrong length.
2861 The presumption is that some radiusclient API changes for 4.51/PH/17
2862 were not taken care of correctly. The code is still untested by me (my
2863 Linux distribution still has 0.3.2 of radiusclient), but it was
2864 contributed by a Radius user.
2866 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2867 expanding the "port" option of the smtp transport.
2869 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2870 while reading a MIME header. Thanks to Tom Hughes for a patch.
2872 PH/24 Include config.h inside local_scan.h so that configuration settings are
2875 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2876 that in an ACL for RCPT (for example), you can examine exactly what was
2879 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2880 commands, but it was not correctly comparing the address with the actual
2881 client host address. Thus, it would show the EHLO address in Received:
2882 header lines when this was not necessary.
2884 PH/27 Added the % operator to ${eval:}.
2886 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2887 it should be ignoring failures (because with -C, for example, it has lost
2888 privilege). It wasn't ignoring creation failures other than "already
2891 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2894 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2895 because an input line was too long, either on its own, or by virtue of
2896 too many continuations, the temporary file was not being removed, and the
2897 return code was incorrect.
2899 PH/31 Missing "BOOL" in function definition in filtertest.c.
2901 PH/32 Applied Sieve patches from the maintainer.
2903 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2905 PH/33 Added "verify = not_blind".
2907 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2908 Local/Makefile (with some defaults set). These are used in built scripts
2909 such as exicyclog, but they have never been used in the exim_install
2910 script (though there are many overriding facilities there). I have
2911 arranged that the exim_install script now takes note of these two
2914 PH/35 Installed configuration files for Dragonfly.
2916 PH/36 When a locally submitted message by a trusted user did not contain a
2917 From: header, and the sender address was obtained from -f or from an SMTP
2918 MAIL command, and the trusted user did not use -F to supply a sender
2919 name, $originator_name was incorrectly used when constructing a From:
2920 header. Furthermore, $originator_name was used for submission mode
2921 messages from external hosts without From: headers in a similar way,
2922 which is clearly wrong.
2924 PH/37 Added control=suppress_local_fixups.
2926 PH/38 When log_selector = +received_sender was set, and the addition of the
2927 sender made the log line's construction buffer exactly full, or one byte
2928 less than full, an overflow happened when the terminating "\n" was
2931 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2932 when the result of a list match is failure because a DNS lookup failed.
2934 PH/40 RM_COMMAND is now used in the building process.
2936 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2937 the "build-* directories that it finds.
2939 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2940 address in a domain literal was a prefix of an interface address.
2942 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2943 when verifying a sender address, unless rewrite_headers is false.
2945 PH/44 Wrote a long comment about why errors_to addresses are verified as
2946 recipients, not senders.
2948 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2949 the ratelimit ACL was added.
2951 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2953 PH/46 Added extra information about PostgreSQL errors to the error string.
2955 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2956 avoiding the need to install special zones in a real server. This is
2957 backwards compatible; if it can't find the fake resolver, it drops back.
2958 Thus, both old and new test suites can be run.
2960 TF/02 Added util/ratelimit.pl
2962 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2963 clock is set back in time.
2965 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2966 Candler <B.Candler@pobox.com>.
2968 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2969 OK for addresses that are the result of redirections.
2971 PH/48 A number of further additions for the benefit of the new test suite,
2972 including a fake gethostbyname() that interfaces to the fake DNS resolver
2975 TF/06 The fix for widen_domains has also been applied to qualify_single and
2976 search_parents which are the other dnslookup options that can cause
2979 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2982 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2984 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2985 service" if it encountered an MX record with an empty target hostname.
2986 The message is now "an MX or SRV record indicated no SMTP service".
2988 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2989 if the DNS of the sending site is misconfigured. This is quite a
2990 common situation. This change restores the behaviour of treating a
2991 helo verification defer as a failure.
2993 PH/51 If self=fail was set on a router, the bounce message did not include the
2994 actual error message.
3000 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3002 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3003 command itself ended in error, the underlying message about the transport
3004 filter timeout was being overwritten with the pipe command error. Now the
3005 underlying error message should be appended to the second error message.
3007 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3009 PH/02 Reduce the amount of output that "make" produces by default. Full output
3010 can still be requested.
3012 PH/03 The warning log line about a condition test deferring for a "warn" verb
3013 was being output only once per connection, rather than after each
3014 occurrence (because it was using the same function as for successful
3015 "warn" verbs). This seems wrong, so I have changed it.
3017 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3018 it should not have, which might have caused a crash in the right
3019 circumstances, but probably never did.
3021 PH/04 Installed a modified version of Tony Finch's patch to make submission
3022 mode fix the return path as well as the Sender: header line, and to
3023 add a /name= option so that you can make the user's friendly name appear
3026 TF/03 Added the control = fakedefer ACL modifier.
3028 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3029 Mark Lowes for thorough testing.
3031 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3033 TK/03 Merged latest SRS patch from Miles Wilton.
3035 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3036 with the definition in sysexits.h (which is #included earlier).
3037 Fortunately, Exim does not actually use EX_OK. The code used to try to
3038 preserve the sysexits.h value, by assumimg that macro definitions were
3039 scanned for macro replacements. I have been disabused of this notion,
3040 so now the code just undefines EX_OK before #including unistd.h.
3042 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3043 in the smtp transport. When a block could not be written in a single
3044 write() function, the timeout was being re-applied to each part-write.
3045 This seems wrong - if the receiver was accepting one byte at a time it
3046 would take for ever. The timeout is now adjusted when this happens. It
3047 doesn't have to be particularly precise.
3049 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3050 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3052 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3053 without a domain if the check to <postmaster@domain> fails.
3055 SC/01 Eximstats: added -xls and the ability to specify output files
3056 (patch written by Frank Heydlauf).
3058 SC/02 Eximstats: use FileHandles for outputing results.
3060 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3062 SC/04 Eximstats: fixed display of large numbers with -nvr option
3064 SC/05 Eximstats: fixed merging of reports with empty tables.
3066 SC/06 Eximstats: added the -include_original_destination flag
3068 SC/07 Eximstats: removed tabs and trailing whitespace.
3070 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3072 TK/06 MBOX spool code: Add real "From " MBOX separator line
3073 so the .eml file is really in mbox format (even though
3074 most programs do not really care). Patch from Alex Miller.
3076 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3077 The latter is generated from $received_to and is only set if the
3078 message has one envelope recipient. SA can use these headers,
3079 obviously out-of-the-box. Patch from Alex Miller.
3081 PH/08 The ${def test on a variable was returning false if the variable's
3082 value was "0", contrary to what the specification has always said!
3083 The result should be true unless the variable is empty.
3085 PH/09 The syntax error of a character other than { following "${if
3086 def:variable_name" (after optional whitespace) was not being diagnosed.
3087 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3088 accidental colon was present, for example, could give incorrect results.
3090 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3091 result is used (not including appendfile, where other changes are about
3094 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3095 This involved changing a lot of size variables from int to off_t. It
3096 should work with maildirs and everything.
3098 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3099 spamd dying while we are connected to it.
3101 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3104 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3105 function for BDB 4.3.
3107 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3109 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3110 the sources for PCRE have been re-arranged and more files are now
3113 PH/15 The code I had for printing potentially long long variables in PH/11
3114 above was not the best (it lost precision). The length of off_t variables
3115 is now inspected at build time, and an appropriate printing format (%ld
3116 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3117 to be "long long int" or "long int". This is needed for the internal
3118 formatting function string_vformat().
3120 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3121 the configuration file to be ":syslog", then the script "guesses" where
3122 the logs files are, rather than using the compiled in default. In our
3123 case the guess is not the same as the compiled default, so the script
3124 suddenly stopped working when I started to use syslog. The patch checks
3125 to see if log_file_path is "". If so, it attempts to read it from exim
3126 with no configuration file to get the compiled in version, before it
3127 falls back to the previous guessing code."
3129 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3130 implementing BATV in an Exim configuration. See NewStuff for the gory
3133 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3134 Makefile that are specific to HP-UX.
3136 PH/18 If the "use_postmaster" option was set for a recipient callout together
3137 with the "random" option, the postmaster address was used as the MAIL
3138 FROM address for the random test, but not for the subsequent recipient
3139 test. It is now used for both.
3141 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3142 patch removes a few documentation additions to RFC 3028, because the
3143 latest draft now contains them. It adds the new en;ascii-case comparator
3144 and a new error check for 8bit text in MIME parts. Comparator and
3145 require names are now matched exactly. I enabled the subaddress
3146 extension, but it is not well tested yet (read: it works for me)."
3148 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3149 rework some of the code of TK/09 above to avoid the hardwired use of
3150 "%lld" and "long long". Replaced the call to snprintf() with a call to
3153 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3154 records point to non-existent hosts", "retry timeout exceeded", and
3155 "retry time not reached for any host after a long failure period".
3157 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3158 experimental DomainKeys support:
3160 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3161 (2) On an error such as an illegally used "control", the wrong name for
3162 the control was given.
3164 These problems did NOT occur unless DomainKeys support was compiled.
3166 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3168 PH/24 Added ${if match_ip condition.
3170 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3171 databases so that it will be absolutely obvious if a crash occurs in the
3172 DB library. This is a regular occurrence (often caused by mis-matched
3175 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3176 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3177 picked up on a user's system that detects such things. There doesn't seem
3178 to be a gcc warning option for this - only an attribute that has to be
3179 put on the function's prototype. It seems that in Fedora Core 4 they have
3180 set this on a number of new functions. No doubt there will be more in due
3183 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3184 specify a transport. However, if an address that was verified by such a
3185 router was the subject of a callout, Exim crashed because it tried to
3186 read the rcpt_include_affixes from the non-existent transport. Now it
3187 just assumes that the setting of that option is false. This bug was
3188 introduced by 4.51/PH/31.
3190 PH/28 Changed -d+all to exclude +memory, because that information is very
3191 rarely of interest, but it makes the output a lot bigger. People tend to
3192 do -d+all out of habit.
3194 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3195 code in os-type was giving problems when libc.so lives in lib64, like on
3198 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3199 aren't the modern standard, and it seems that some systems' include files
3200 don't always have them. Exim was already checking for some of the newer
3201 ones like T_AAAA, and defining it itself. I've added checks for all the
3202 record types that Exim uses.
3204 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3205 not automatically generating a new one, as it is supposed to. This
3206 prevented TLS from working. If the file did exist, but contained invalid
3207 data, a new version was generated, as expected. It was only the case of a
3208 non-existent file that was broken.
3210 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3211 with a change in libdomainkeys > 0.64.
3213 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3214 from DNS. If the selector record carries the flag, it now has
3215 precedence over the domain-wide flag.
3217 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3219 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3220 the use of a port name that isn't defined in /etc/services) occurred, the
3221 message was deferred as in a normal delivery, and thus remained on the
3222 spool, instead of being failed because of the mua_wrapper setting. This
3223 is now fixed, and I tidied up some of the mua_wrapper messages at the
3226 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3227 the messages in a hash of arrays rather than using individual hashes.
3228 This is a bit cleaner and results in dramatic memory savings, albeit
3229 at a slight CPU cost.
3231 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3232 as requested by Marc Sherman.
3234 SC/10 Eximstats: added histograms for user specified patterns as requested
3237 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3239 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3240 fopen() in the content-scanning modules that did not already have it.
3246 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3247 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3249 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3251 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3253 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3254 to test Sieve filters that use "vacation".
3256 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3257 that changes the way the GnuTLS parameters are stored in the cache file.
3258 The new format can be generated externally. For backward compatibility,
3259 if the data in the cache doesn't make sense, Exim assumes it has read an
3260 old-format file, and it generates new data and writes a new file. This
3261 means that you can't go back to an older release without removing the
3264 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3265 work if there are any delivery delays because "one_time" forces the
3266 parent to be marked "delivered", so its unseen clone is never tried
3267 again. For this reason, Exim now forbids the simultaneous setting of
3270 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3271 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3272 entry if you want to know the details. The fix, however, neglected to
3273 consider the case when local delivery batching is involved. The test for
3274 "previously delivered" was not happening when checking to see if an
3275 address could be batched with a previous (undelivered) one; under
3276 certain circumstances this could lead to multiple deliveries to the same
3279 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3280 in its include files, and this causes problems building Exim.
3282 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3283 header_syntax) but Exim was just ignoring anything given after a slash.
3284 In particular, this caused confusion with an attempt to use "verify =
3285 reverse_host_lookup/defer_ok". An error is now given when options are
3286 supplied for verify items that do not have them. (Maybe reverse_host_
3287 lookup should have a defer_ok option, but that's a different point.)
3289 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3290 defined by RFC 821) to 2048, because there were problems with some AUTH
3291 commands, and RFC 1869 says the size should be increased for extended
3292 SMTP commands that take arguments.
3294 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3297 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3298 "unknown" error; now it says that the functionality isn't in the binary.
3300 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3301 an address' error message when a string expansion fails (syntax or
3302 whatever). Otherwise the password may appear in the log. Following change
3303 PH/42 below, there is no longer a chance of it appearing in a bounce
3306 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3308 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3309 of its addresses was ever tried. (Bugzilla bug #2.)
3311 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3312 the result incorrectly in the debug output. (It correctly added a newline
3313 to what was transported.)
3315 TF/01 Added $received_time.
3317 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3318 commented out examples of how to interface to a virus scanner and to
3319 SpamAssassin. Also added commented examples of av_scanner and
3320 spamd_address settings.
3322 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3323 and controls are allowed in which ACLs. There were a couple of minor
3324 errors. Some of the entries in the conditions table (which is a table of
3325 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3326 negation of where the condition IS allowed.
3328 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3330 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3331 header file does not have a version number, so I've had to invent a new
3332 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3333 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3334 radiusclient), but it was contributed by a Radius user.
3336 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3337 files or directories, for OpenSSL.
3339 PH/19 When an Exim process that is running as root has to create an Exim log
3340 file, it does so in a subprocess that runs as exim:exim so as to get the
3341 ownership right at creation (otherwise, other Exim processes might see
3342 the file with the wrong ownership). There was no test for failure of this
3343 fork() call, which would lead to the process getting stuck as it waited
3344 for a non-existent subprocess. Forks do occasionally fail when resources
3345 run out. I reviewed all the other calls to fork(); they all seem to check
3348 PH/20 When checking for unexpected SMTP input at connect time (before writing
3349 the banner), Exim was not dealing correctly with a non-positive return
3350 from the read() function. If the client had disconnected by this time,
3351 the result was a log entry for a synchronization error with an empty
3352 string after "input=" when read() returned zero. If read() returned -1
3353 (an event I could not check), uninitialized data bytes were printed.
3354 There were reports of junk text (parts of files, etc) appearing after
3357 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3359 PH/22 Added support for macro redefinition, and (re)definition in between
3360 driver and ACL definitions.
3362 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3363 forgetting to use the resulting value; it was using the unexpanded value.
3365 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3366 hadn't been configured. The fix is from Juergen Kreileder, who
3367 understands it better than I do:
3369 "Here's what I see happening with three configured cyrus_sasl
3370 authenticators configured (plain, login, cram-md5):
3372 On startup auth_cyrus_sasl_init() gets called for each of these.
3373 This means three calls to sasl_listmech() without a specified mech_list.
3374 => SASL tests which mechs of all available mechs actually work
3375 => three warnings about OTP not working
3376 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3378 With the patch, sasl_listmech() also gets called three times. But now
3379 SASL's mech_list option is set to the server_mech specified in the the
3380 authenticator. Or in other words, the answer from sasl_listmech()
3381 gets limited to just the mech you're testing for (which is different
3383 => the return list contains just 'plain' or 'login', 'cram-md5' or
3384 nothing depending on the value of ob->server_mech.
3386 I've just tested the patch: Authentication still works fine,
3387 unavailable mechs specified in the exim configuration are still
3388 caught, and the auth.log warnings about OTP are gone."
3390 PH/25 When debugging is enabled, the contents of the command line are added
3391 to the debugging output, even when log_selector=+arguments is not
3394 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3395 answer is "GNU", and only if the return is "GNU/something" is the answer
3398 PH/27 $acl_verify_message is now set immediately after the failure of a
3399 verification in an ACL, and so is available in subsequent modifiers. In
3400 particular, the message can be preserved by coding like this:
3402 warn !verify = sender
3403 set acl_m0 = $acl_verify_message
3405 Previously, $acl_verify_message was set only while expanding "message"
3406 and "log_message" when a very denied access.
3408 PH/28 Modified OS/os.c-Linux with
3410 -#ifndef OS_LOAD_AVERAGE
3411 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3413 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3414 nomenclature these days.)
3416 PH/29 Installed patch from the Sieve maintainer that adds the options
3417 sieve_useraddress and sieve_subaddress to the redirect router.
3419 PH/30 In these circumstances:
3420 . Two addresses routed to the same list of hosts;
3421 . First host does not offer TLS;
3422 . First host accepts first address;
3423 . First host gives temporary error to second address;
3424 . Second host offers TLS and a TLS session is established;
3425 . Second host accepts second address.
3426 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3427 and peerdn, if requested) that were in fact used only for the second
3430 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3431 attention to any local part prefix or suffix that was matched by the
3432 router that accepted the address. It now behaves in the same way as it
3433 does for delivery: the affixes are removed from the local part unless
3434 rcpt_include_affixes is set on the transport.
3436 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3437 timeout during the DATA phase of an incoming message.
3439 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3440 applied a patch sanctioned by the Sieve maintainer.
3442 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3443 the uid or gid is negative. A case of a negative gid caused this to be
3444 noticed. The fix allows for either to be negative.
3446 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3447 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3450 PH/36 The change PH/12 above was broken. Fixed it.
3452 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3453 the grounds that routing the same address twice would always produce the
3454 same answer. This might have been true once, but it is certainly no
3455 longer true now. Routing a child address may depend on the previous
3456 routing that produced that child. Some complicated redirection strategies
3457 went wrong when messages had multiple recipients, and made Exim's
3458 behaviour dependent on the order in which the addresses were given.
3460 I have moved the duplicate checking until after the routing is complete.
3461 Exim scans the addresses that are assigned to local and remote
3462 transports, and removes any duplicates. This means that more work will be
3463 done, as duplicates will always all be routed, but duplicates are
3464 presumably rare, so I don't expect this is of any significance.
3466 For deliveries to pipes, files, and autoreplies, the duplicate checking
3467 still happens during the routing process, since they are not going to be
3470 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3471 It corrects a timeout issue with spamd. This is Ian's comment: "The
3472 background is that sometimes spamd either never reads data from a
3473 connection it has accepted, or it never writes response data. The exiscan
3474 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3475 blindly assumes that writes won't block so it may never time out."
3477 PH/39 Allow G after quota size as well as K and M.
3479 PH/40 The value set for $authenticated_id in an authenticator may not contain
3480 binary zeroes or newlines because the value is written to log lines and
3481 to spool files. There was no check on this. Now the value is run through
3482 the string_printing() function so that such characters are converted to
3483 printable escape sequences.
3485 PH/41 $message_linecount is a new variable that contains the total number of
3486 lines in the message. Compare $body_linecount, which is the count for the
3489 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3490 bounce and delay warning messages, except in certain special cases, which
3493 (a) An SMTP error message from a remote host;
3494 (b) A message specified in a :fail: redirection;
3495 (c) A message specified in a "fail" command in a system filter;
3496 (d) A message specified in a FAIL return from the queryprogram router;
3497 (e) A message specified by the cannot_route_message router option.
3499 In these cases only, Exim does include the error details in bounce and
3500 warning messages. There are also a few cases where bland messages such
3501 as "unrouteable address" or "local delivery error" are given.
3503 PH/43 $value is now also set for the "else" part of a ${run expansion.
3505 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3506 being worked on, but at least Exim now implements the latest version to
3509 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3510 process to complete was treated as a delivery failure, a timeout while
3511 writing the message to the pipe was logged, but erroneously treated as a
3512 successful delivery. Such timeouts include transport filter timeouts. For
3513 consistency with the overall process timeout, these timeouts are now
3514 treated as errors, giving rise to delivery failures by default. However,
3515 there is now a new Boolean option for the pipe transport called
3516 timeout_defer, which, if set TRUE, converts the failures into defers for
3517 both kinds of timeout. A transport filter timeout is now identified in
3520 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3521 systems where "make" and "gmake" are different, calling "gmake" at top
3522 level broke things. I've arranged for the value of $(MAKE) to be passed
3523 from the Makefile to this script so that it can call the same version of
3527 A note about Exim versions 4.44 and 4.50
3528 ----------------------------------------
3530 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3531 changes of various kinds. As a consequence, a big documentation update was
3532 needed. This delayed the release for rather longer than seemed good, especially
3533 in the light of a couple of (minor) security issues. Therefore, the changes
3534 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3535 release. So 4.44 and 4.50 are in effect two different branches that both start
3538 I have left the 4.50 change log unchanged; it contains all the changes since
3539 4.43. The change log for 4.44 is below; many of its items are identical to
3540 those for 4.50. This seems to be the most sensible way to preserve the
3541 historical information.
3547 1. Minor wording change to the doc/README.SIEVE file.
3549 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3550 computation of the current number of files was incorrect.
3552 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3553 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3556 4. Give more explanation in the error message when the command for a transport
3557 filter fails to execute.
3559 5. There are several places where Exim runs a non-Exim command in a
3560 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3561 was being done only for the command run by the queryprogram router. It is
3562 now done for all such subprocesses. The other cases are: ${run, transport
3563 filters, and the commands run by the lmtp and pipe transports.
3565 6. Added CONFIGURE_GROUP build-time option.
3567 7. Some older OS have a limit of 256 on the maximum number of file
3568 descriptors. Exim was using setrlimit() to set 1000 as a large value
3569 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3570 systems. I've change it so that if it can't get 1000, it tries for 256.
3572 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3573 was an oversight, and furthermore, ever since the addition of extra
3574 controls (e.g. 4.43/32), the checks on when to allow different forms of
3575 "control" were broken. There should now be diagnostics for all cases when a
3576 control that does not make sense is encountered.
3578 9. Added the /retain_sender option to "control=submission".
3580 10. $recipients is now available in the predata ACL (oversight).
3582 11. Tidy the search cache before the fork to do a delivery from a message
3583 received from the command line. Otherwise the child will trigger a lookup
3584 failure and thereby defer the delivery if it tries to use (for example) a
3585 cached ldap connection that the parent has called unbind on.
3587 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3588 of $address_data from the recipient verification was clobbered by the
3589 sender verification.
3591 13. The value of address_data from a sender verification is now available in
3592 $sender_address_data in subsequent conditions in the ACL statement.
3594 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3596 15. Added a new option "connect=<time>" to callout options, to set a different
3599 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3600 was its contents. (It was OK if the option was not defined at all.)
3602 17. A "Completed" log line is now written for messages that are removed from
3603 the spool by the -Mrm option.
3605 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3606 information about exactly what failed.
3608 19. Added -dd to debug only the daemon process.
3610 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3611 handles timeouts, both on the server side and network timeouts. Renamed the
3612 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3614 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3615 It is now set to "smtps".
3617 22. $host_address is now set to the target address during the checking of
3618 ignore_target_hosts.
3620 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3621 being passed; this would have caused $sender_host_name to have been used if
3622 matching the list had actually called for a host name (not very likely,
3623 since this list is usually IP addresses). A host name is now passed as
3626 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3627 code that specifies a non-restarting handler (typically sigaction() in
3628 modern systems) in an attempt to fix a rare and obscure crash bug.
3630 25. Narrowed the window for a race in the daemon that could cause it to ignore
3631 SIGCHLD signals. This is not a major problem, because they are used only to
3632 wake it up if nothing else does.
3634 26. A malformed maildirsize file could cause Exim to calculate negative values
3635 for the mailbox size or file count. Odd effects could occur as a result.
3636 The maildirsize information is now recalculated if the size or filecount
3639 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3640 support for a long time. Removed HAVE_SYS_VFS_H.
3642 28. Installed the latest version of exipick from John Jetmore.
3644 29. In an address list, if the pattern was not a regular expression, an empty
3645 subject address (from a bounce message) matched only if the pattern was an
3646 empty string. Non-empty patterns were not even tested. This was the wrong
3647 because it is perfectly reasonable to use an empty address as part of a
3648 database query. An empty address is now tested by patterns that are
3649 lookups. However, all the other forms of pattern expect the subject to
3650 contain a local part and a domain, and therefore, for them, an empty
3651 address still always fails if the pattern is not itself empty.
3653 30. Exim went into a mad DNS loop when attempting to do a callout where the
3654 host was specified on an smtp transport, and looking it up yielded more
3655 than one IP address.
3657 31. Re-factored the code for checking spool and log partition space into a
3658 function that finds that data and another that does the check. The former
3659 is then used to implement four new variables: $spool_space, $log_space,
3660 $spool_inodes, and $log_inodes.
3662 32. The RFC2047 encoding function was originally intended for short strings
3663 such as real names; it was not keeping to the 75-character limit for
3664 encoded words that the RFC imposes. It now respects the limit, and
3665 generates multiple encoded words if necessary. To be on the safe side, I
3666 have increased the buffer size for the ${rfc2047: expansion operator from
3669 33. It is now permitted to omit both strings after an "if" condition; if the
3670 condition is true, the result is "true". As before, when the second string
3671 is omitted, a false condition yields an empty string. This makes it less
3672 cumbersome to write custom ACL and router conditions.
3674 34. Failure to deliver a bounce message always caused it to be frozen, even if
3675 there was an errors_to setting on the router. The errors_to setting is now
3678 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3679 canonical form (fully expanded) before being placed in
3680 $sender_host_address.
3682 36. The table in the code that translates DNS record types into text (T_A to
3683 "A" for instance) was missing entries for NS and CNAME. It is just possible
3684 that this could have caused confusion if both these types were looked up
3685 for the same domain, because the text type is used as part of Exim's
3686 per-process caching. But the chance of anyone hitting this buglet seems
3689 37. The dnsdb lookup has been extended in a number of ways.
3691 (1) There is a new type, "zns", which walks up the domain tree until it
3692 finds some nameserver records. It should be used with care.
3694 (2) There is a new type, "mxh", which is like "mx" except that it returns
3695 just the host names, not the priorities.
3697 (3) It is now possible to give a list of domains (or IP addresses) to be
3698 looked up. The behaviour when one of the lookups defers can be
3699 controlled by a keyword.
3701 (4) It is now possible to specify the separator character for use when
3702 multiple records are returned.
3704 38. The dnslists ACL condition has been extended: it is now possible to supply
3705 a list of IP addresses and/or domains to be looked up in a particular DNS
3708 39. Added log_selector=+queue_time_overall.
3710 40. When running the queue in the test harness, wait just a tad after forking a
3711 delivery process, to get repeatability of debugging output.
3713 41. Include certificate and key file names in error message when GnuTLS fails
3714 to set them up, because the GnuTLS error message doesn't include the name
3715 of the failing file when there is a problem reading it.
3717 42. Allow both -bf and -bF in the same test run.
3719 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3721 44. The "Exiscan patch" is now merged into the mainline Exim source.
3723 45. Sometimes the final signoff response after QUIT could fail to get
3724 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3725 < 0 before doing a fflush(). This bug looks as though it goes back to the
3726 introduction of TLS in release 3.20, but "sometimes" must have been rare
3727 because the tests only now provoked it.
3729 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3730 (this can affect the format of dates).
3732 47. exim_tidydb, when checking for the continued existence of a message for
3733 which it has found a message-specific retry record, was not finding
3734 messages that were in split spool directories. Consequently, it was
3735 deleting retry records that should have stayed in existence.
3737 48. Steve fixed some bugs in eximstats.
3739 49. The SPA authentication driver was not abandoning authentication and moving
3740 on to the next authenticator when an expansion was forced to fail,
3741 contradicting the general specification for all authenticators. Instead it
3742 was generating a temporary error. It now behaves as specified.
3744 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3745 (the order specifies the preference for clients). The order is now AES256,
3746 AES128, 3DES, ARCFOUR128.
3748 51. Small patch to Sieve code - explicitly set From: when generating an
3751 52. Exim crashed if a remote delivery caused a very long error message to be
3752 recorded - for instance if somebody sent an entire SpamAssassin report back
3753 as a large number of 550 error lines. This bug was coincidentally fixed by
3754 increasing the size of one of Exim's internal buffers (big_buffer) that
3755 happened as part of the Exiscan merge. However, to be on the safe side, I
3756 have made the code more robust (and fixed the comments that describe what
3759 53. Now that there can be additional text after "Completed" in log lines (if
3760 the queue_time_overall log selector is set), a one-byte patch to exigrep
3761 was needed to allow it to recognize "Completed" as not the last thing in
3764 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3765 patch that reportedly fixes this has been added. I am not expert enough to
3766 create a test for it. This is what the patch creator wrote:
3768 "I found a little strange behaviour of ldap code when working with
3769 Windows 2003 AD Domain, where users was placed in more than one
3770 Organization Units. When I tried to give exim partial DN, the exit code
3771 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3772 But simultaneously result of request was absolutely normal ldap result,
3773 so I produce this patch..."
3775 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3776 REFERENCE, so I have modified the code to exclude the patch when that macro
3779 55. Some experimental protocols are using DNS PTR records for new purposes. The
3780 keys for these records are domain names, not reversed IP addresses. The
3781 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3782 leaves it alone. Component reversal etc. now happens only for IP addresses.
3785 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3787 57. Double the size of the debug message buffer (to 2048) so that more of very
3788 long debug lines gets shown.
3790 58. The exicyclog utility now does better if the number of log files to keep
3791 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3793 59. Two changes related to the smtp_active_hostname option:
3795 (1) $smtp_active_hostname is now available as a variable.
3796 (2) The default for smtp_banner uses $smtp_active_hostname instead
3797 of $primary_hostname.
3799 60. The host_aton() function is supposed to be passed a string that is known
3800 to be a valid IP address. However, in the case of IPv6 addresses, it was
3801 not checking this. This is a hostage to fortune. Exim now panics and dies
3802 if the condition is not met. A case was found where this could be provoked
3803 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3804 components; fortuitously, this particular loophole had already been fixed
3805 by change 4.50/55 above.
3807 If there are any other similar loopholes, the new check in host_aton()
3808 itself should stop them being exploited. The report I received stated that
3809 data on the command line could provoke the exploit when Exim was running as
3810 exim, but did not say which command line option was involved. All I could
3811 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3812 running as the user.
3815 61. There was a buffer overflow vulnerability in the SPA authentication code
3816 (which came originally from the Samba project). I have added a test to the
3817 spa_base64_to_bits() function which I hope fixes it.
3820 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3821 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3823 63. The daemon start-up calls getloadavg() while still root for those OS that
3824 need the first call to be done as root, but it missed one case: when
3825 deliver_queue_load_max is set with deliver_drop_privilege. This is
3826 necessary for the benefit of the queue runner, because there is no re-exec
3827 when deliver_drop_privilege is set.
3829 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3830 This has been fixed.
3832 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3833 was in use, was not putting the data itself into the right store pool;
3834 consequently, it could be overwritten for a subsequent message in the same
3835 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3838 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3840 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3841 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3842 the function were treating the return as a boolean value, which happened to
3843 work because 0=false and not-0=true, but is not correct code.
3845 68. The host_aton() function was not handling scoped IPv6 addresses (those
3846 with, for example, "%eth0" on the end) correctly.
3848 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3849 negated items (that is, ~something) in unsigned ints. Some compilers
3850 apparently mutter when there is no cast.
3852 70. If an address verification called from an ACL failed, and did not produce a
3853 user-specific message (i.e. there was only a "system" message), nothing was
3854 put in $acl_verify_message. In this situation, it now puts the system
3857 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3858 session; change 4.31/43 added the unwanted input to the log line - except
3859 that it did not do this in the start of session case. It now does.
3861 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3862 This is wrong and can cause the other end to generate a synchronization
3863 error if it is another Exim or anything else that does the synchronization
3864 check. A QUIT command is no longer sent after a timeout.
3866 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3867 during host lookups.
3869 74. The defer_ok option of callout verification was not working if it was used
3870 when verifying addresses in header lines, that is, for this case:
3872 verify = header_sender/callout=defer_ok
3874 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3875 those file descriptors could be used for SMTP connections. If anything
3876 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3877 could be sent to the SMTP client, causing chaos. The daemon now opens
3878 stdin, stdout, and stderr to /dev/null when it puts itself into the
3881 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3882 log by default. The user can override this with suitable Perl magic.
3884 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3885 the log message when discard triggers, was not working for the DATA ACL or
3886 for the non-SMTP ACL.
3888 78. Error message wording change in sieve.c.
3890 79. If smtp_accept_max_per_host was set, the number of connections could be
3891 restricted to fewer than expected, because the daemon was trying to set up
3892 a new connection before checking whether the processes handling previous
3893 connections had finished. The check for completed processes is now done
3894 earlier. On busy systems, this bug wouldn't be noticed because something
3895 else would have woken the daemon, and it would have reaped the completed
3898 80. If a message was submitted locally by a user whose login name contained one
3899 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3900 It caused a spool format error. I have fixed the spool reading code. A
3901 related problem was that the "from" clause in the Received: line became
3902 illegal because of the space(s). It is now covered by ${quote_local_part.
3904 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3907 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3909 83. Patch from Sieve maintainer to fix unterminated string problem in
3910 "vacation" handling.
3912 84. Some minor changes to the Linux configuration files to help with other
3913 OS variants using glibc.
3915 85. One more patch for Sieve to update vacation handling to latest spec.
3918 ----------------------------------------------------
3919 See the note above about the 4.44 and 4.50 releases.
3920 ----------------------------------------------------
3926 1. Change 4.43/35 introduced a bug that caused file counts to be
3927 incorrectly computed when quota_filecount was set in an appendfile
3930 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3931 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3934 3. Give more explanation in the error message when the command for a transport
3935 filter fails to execute.
3937 4. There are several places where Exim runs a non-Exim command in a
3938 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3939 was being done only for the command run by the queryprogram router. It is
3940 now done for all such subprocesses. The other cases are: ${run, transport
3941 filters, and the commands run by the lmtp and pipe transports.
3943 5. Some older OS have a limit of 256 on the maximum number of file
3944 descriptors. Exim was using setrlimit() to set 1000 as a large value
3945 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3946 systems. I've change it so that if it can't get 1000, it tries for 256.
3948 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3949 was an oversight, and furthermore, ever since the addition of extra
3950 controls (e.g. 4.43/32), the checks on when to allow different forms of
3951 "control" were broken. There should now be diagnostics for all cases when a
3952 control that does not make sense is encountered.
3954 7. $recipients is now available in the predata ACL (oversight).
3956 8. Tidy the search cache before the fork to do a delivery from a message
3957 received from the command line. Otherwise the child will trigger a lookup
3958 failure and thereby defer the delivery if it tries to use (for example) a
3959 cached ldap connection that the parent has called unbind on.
3961 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3962 of $address_data from the recipient verification was clobbered by the
3963 sender verification.
3965 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3966 was its contents. (It was OK if the option was not defined at all.)
3968 11. A "Completed" log line is now written for messages that are removed from
3969 the spool by the -Mrm option.
3971 12. $host_address is now set to the target address during the checking of
3972 ignore_target_hosts.
3974 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3975 being passed; this would have caused $sender_host_name to have been used if
3976 matching the list had actually called for a host name (not very likely,
3977 since this list is usually IP addresses). A host name is now passed as
3980 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3981 code that specifies a non-restarting handler (typically sigaction() in
3982 modern systems) in an attempt to fix a rare and obscure crash bug.
3984 15. Narrowed the window for a race in the daemon that could cause it to ignore
3985 SIGCHLD signals. This is not a major problem, because they are used only to
3986 wake it up if nothing else does.
3988 16. A malformed maildirsize file could cause Exim to calculate negative values
3989 for the mailbox size or file count. Odd effects could occur as a result.
3990 The maildirsize information is now recalculated if the size or filecount
3993 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3994 support for a long time. Removed HAVE_SYS_VFS_H.
3996 18. Updated exipick to current release from John Jetmore.
3998 19. Allow an empty sender to be matched against a lookup in an address list.
3999 Previously the only cases considered were a regular expression, or an
4002 20. Exim went into a mad DNS lookup loop when doing a callout where the
4003 host was specified on the transport, if the DNS lookup yielded more than
4006 21. The RFC2047 encoding function was originally intended for short strings
4007 such as real names; it was not keeping to the 75-character limit for
4008 encoded words that the RFC imposes. It now respects the limit, and
4009 generates multiple encoded words if necessary. To be on the safe side, I
4010 have increased the buffer size for the ${rfc2047: expansion operator from
4013 22. Failure to deliver a bounce message always caused it to be frozen, even if
4014 there was an errors_to setting on the router. The errors_to setting is now
4017 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4018 canonical form (fully expanded) before being placed in
4019 $sender_host_address.
4021 24. Updated eximstats to version 1.33
4023 25. Include certificate and key file names in error message when GnuTLS fails
4024 to set them up, because the GnuTLS error message doesn't include the name
4025 of the failing file when there is a problem reading it.
4027 26. Expand error message when OpenSSL has problems setting up cert/key files.
4030 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4031 (this can affect the format of dates).
4033 28. exim_tidydb, when checking for the continued existence of a message for
4034 which it has found a message-specific retry record, was not finding
4035 messages that were in split spool directories. Consequently, it was
4036 deleting retry records that should have stayed in existence.
4038 29. eximstats updated to version 1.35
4039 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4040 1.35 - bugfix such that pie charts by volume are generated correctly
4042 30. The SPA authentication driver was not abandoning authentication and moving
4043 on to the next authenticator when an expansion was forced to fail,
4044 contradicting the general specification for all authenticators. Instead it
4045 was generating a temporary error. It now behaves as specified.
4047 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4048 (the order specifies the preference for clients). The order is now AES256,
4049 AES128, 3DES, ARCFOUR128.
4051 31. Small patch to Sieve code - explicitly set From: when generating an
4054 32. Exim crashed if a remote delivery caused a very long error message to be
4055 recorded - for instance if somebody sent an entire SpamAssassin report back
4056 as a large number of 550 error lines. This bug was coincidentally fixed by
4057 increasing the size of one of Exim's internal buffers (big_buffer) that
4058 happened as part of the Exiscan merge. However, to be on the safe side, I
4059 have made the code more robust (and fixed the comments that describe what
4062 33. Some experimental protocols are using DNS PTR records for new purposes. The
4063 keys for these records are domain names, not reversed IP addresses. The
4064 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4065 leaves it alone. Component reversal etc. now happens only for IP addresses.
4068 34. The host_aton() function is supposed to be passed a string that is known
4069 to be a valid IP address. However, in the case of IPv6 addresses, it was
4070 not checking this. This is a hostage to fortune. Exim now panics and dies
4071 if the condition is not met. A case was found where this could be provoked
4072 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4073 components; fortuitously, this particular loophole had already been fixed
4074 by change 4.50/55 or 4.44/33 above.
4076 If there are any other similar loopholes, the new check in host_aton()
4077 itself should stop them being exploited. The report I received stated that
4078 data on the command line could provoke the exploit when Exim was running as
4079 exim, but did not say which command line option was involved. All I could
4080 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4081 running as the user.
4084 35. There was a buffer overflow vulnerability in the SPA authentication code
4085 (which came originally from the Samba project). I have added a test to the
4086 spa_base64_to_bits() function which I hope fixes it.
4089 36. The daemon start-up calls getloadavg() while still root for those OS that
4090 need the first call to be done as root, but it missed one case: when
4091 deliver_queue_load_max is set with deliver_drop_privilege. This is
4092 necessary for the benefit of the queue runner, because there is no re-exec
4093 when deliver_drop_privilege is set.
4095 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4096 was in use, was not putting the data itself into the right store pool;
4097 consequently, it could be overwritten for a subsequent message in the same
4098 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4101 38. Sometimes the final signoff response after QUIT could fail to get
4102 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4103 < 0 before doing a fflush(). This bug looks as though it goes back to the
4104 introduction of TLS in release 3.20, but "sometimes" must have been rare
4105 because the tests only now provoked it.
4111 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4112 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4113 Now it returns an integer. A number of calls were still expecting a T/F
4114 return. Fortuitously, in all cases, the tests worked in OK situations,
4115 which is the norm. However, things would have gone wrong on any write
4116 failures on the smtp file descriptor. This function is used when sending
4117 messages over SMTP and also when doing verify callouts.
4119 2. When Exim is called to do synchronous delivery of a locally submitted
4120 message (the -odf or -odi options), it no longer closes stderr before doing
4123 3. Implemented the mua_wrapper option.
4125 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4127 5. Implemented the functions header_remove(), header_testname(),
4128 header_add_at_position(), and receive_remove_recipient(), and exported them
4131 6. If an ACL "warn" statement specified the addition of headers, Exim already
4132 inserted X-ACL-Warn: at the start if there was no header name. However, it
4133 was not making this test for the second and subsequent header lines if
4134 there were newlines in the string. This meant that an invalid header could
4135 be inserted if Exim was badly configured.
4137 7. Allow an ACL "warn" statement to add header lines at the start or after all
4138 the Received: headers, as well as at the end.
4140 8. Added the rcpt_4xx retry error code.
4142 9. Added postmaster_mailfrom=xxx to callout verification option.
4144 10. Added mailfrom=xxxx to the callout verification option, for verify=
4147 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4148 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4150 12. Inserted some casts to stop certain compilers warning when using pointer
4151 differences as field lengths or precisions in printf-type calls (mostly
4152 affecting debugging statements).
4154 13. Added optional readline() support for -be (dynamically loaded).
4156 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4157 same clock tick as a message's arrival, so that its received time was the
4158 same as the "first fail" time on the retry record, and that message
4159 remained on the queue past the ultimate address timeout, every queue runner
4160 would try a delivery (because it was past the ultimate address timeout) but
4161 after another failure, the ultimate address timeout, which should have then
4162 bounced the address, did not kick in. This was a "< instead of <=" error;
4163 in most cases the first failure would have been in the next clock tick
4164 after the received time, and all would be well.
4166 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4167 being recognized when the domain list was tested by the match_domain
4168 condition in an expansion string.
4170 16. Added the ${str2b64: operator.
4172 17. Exim was always calling setrlimit() to set a large limit for the number of
4173 processes, without checking whether the existing limit was already
4174 adequate. (It did check for the limit on file descriptors.) Furthermore,
4175 errors from getrlimit() and setrlimit() were being ignored. Now they are
4176 logged to the main and panic logs, but Exim does carry on, to try to do its
4177 job under whatever limits there are.
4179 18. Imported PCRE 5.0.
4181 19. Trivial typo in log message " temporarily refused connection" (the leading
4184 20. If the log selector return_path_on_delivery was set and an address was
4185 redirected to /dev/null, the delivery process crashed because it assumed
4186 that a return path would always be set for a "successful" delivery. In this
4187 case, the whole delivery is bypassed as an optimization, and therefore no
4190 21. Internal re-arrangement: the function for sending a challenge and reading
4191 a response while authentication was assuming a zero-terminated challenge
4192 string. It's now changed to take a pointer and a length, to allow for
4193 binary data in such strings.
4195 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4197 23. Exim was not respecting finduser_retries when seeking the login of the
4198 uid under which it was called; it was always trying 10 times. (The default
4199 setting of finduser_retries is zero.) Also, it was sleeping after the final
4200 failure, which is pointless.
4202 24. Implemented tls_on_connect_ports.
4204 25. Implemented acl_smtp_predata.
4206 26. If the domain in control=submission is set empty, Exim assumes that the
4207 authenticated id is a complete email address when it generates From: or
4208 Sender: header lines.
4210 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4211 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4212 chown and chgrp in /bin and hostname in /usr/bin.
4214 28. Exim was keeping the "process log" file open after each use, just as it
4215 does for the main log. This opens the possibility of it remaining open for
4216 long periods when the USR1 signal hits a daemon. Occasional processlog
4217 errors were reported, that could have been caused by this. Anyway, it seems
4218 much more sensible not to leave this file open at all, so that is what now
4221 29. The long-running daemon process does not normally write to the log once it
4222 has entered its main loop, and it closes the log before doing so. This is
4223 so that log files can straightforwardly be renamed and moved. However,
4224 there are a couple of unusual error situations where the daemon does write
4225 log entries, and I had neglected to close the log afterwards.
4227 30. The text of an SMTP error response that was received during a remote
4228 delivery was being truncated at 512 bytes. This is too short for some of
4229 the long messages that one sometimes sees. I've increased the limit to
4232 31. It is now possible to make retry rules that apply only when a message has a
4233 specific sender, in particular, an empty sender.
4235 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4236 it possible to be selective about when SMTP synchronization is enforced.
4238 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4240 32. Implemented hosts_connection_nolog.
4242 33. Added an ACL for QUIT.
4244 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4247 35. Added mailbox_size and mailbox_filecount to appendfile.
4249 36. Added control = no_multiline_responses to ACLs.
4251 37. There was a bug in the logic of the code that waits for the clock to tick
4252 in the case where the clock went backwards by a substantial amount such
4253 that the microsecond fraction of "now" was more than the microsecond
4254 fraction of "then" (but the whole seconds number was less).
4256 38. Added support for the libradius Radius client library this is found on
4257 FreeBSD (previously only the radiusclient library was supported).
4263 1. When certain lookups returned multiple values in the form name=value, the
4264 quoting of the values was not always being done properly. Specifically:
4265 (a) If the value started with a double quote, but contained no whitespace,
4267 (b) If the value contained whitespace other than a space character (i.e.
4268 tabs or newlines or carriage returns) it was not quoted.
4269 This fix has been applied to the mysql and pgsql lookups by writing a
4270 separate quoting function and calling it from the lookup code. The fix
4271 should probably also be applied to nisplus, ibase and oracle lookups, but
4272 since I cannot test any of those, I have not disturbed their existing code.
4274 2. A hit in the callout cache for a specific address caused a log line with no
4275 reason for rejecting RCPT. Now it says "Previous (cached) callout
4276 verification failure".
4278 3. There was an off-by-one bug in the queryprogram router. An over-long
4279 return line was truncated at 256 instead of 255 characters, thereby
4280 overflowing its buffer with the terminating zero. As well as fixing this, I
4281 have increased the buffer size to 1024 (and made a note to document this).
4283 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4284 when Exim is waiting for an SMTP response from a remote server, Exim
4285 restarts its select() call on the socket, thereby resetting its timeout.
4286 This is not a problem when such interrupts are rare. Somebody set up a cron
4287 job to run exiwhat every 2 minutes, which is less than the normal select()
4288 timeout (5 or 10 minutes). This meant that the select() timeout never
4289 kicked in because it was always reset. I have fixed this by comparing the
4290 time when an interrupt arrives with the time at the start of the first call
4291 to select(). If more time than the timeout has elapsed, the interrupt is
4292 treated as a timeout.
4294 5. Some internal re-factoring in preparation for the addition of Sieve
4295 extensions (by MH). In particular, the "personal" test is moved to a
4296 separate function, and given an option for scanning Cc: and Bcc: (which is
4297 not set for Exim filters).
4299 6. When Exim created an email address using the login of the caller as the
4300 local part (e.g. when creating a From: or Sender: header line), it was not
4301 quoting the local part when it contained special characters such as @.
4303 7. Installed new OpenBSD configuration files.
4305 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4306 try to make them clearer.
4308 9. Callout options, other than the timeout value, were being ignored when
4309 verifying sender addresses in header lines. For example, when using
4311 verify = header_sender/callout=no_cache
4313 the cache was (incorrectly) being used.
4315 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4316 only the Cygwin environment.
4318 11. When return_path_on_delivery was set as a log selector, if different remote
4319 addresses in the same message used different return paths and parallel
4320 remote delivery occurred, the wrong values would sometimes be logged.
4321 (Whenever a remote delivery process finished, the return path value from
4322 the most recently started remote delivery process was logged.)
4324 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4325 lines when AUTH and/or TLS are in use. This is the "received protocol"
4326 field. Exim used to use "asmtp" for authenticated SMTP, without any
4327 indication (in the protocol name) for TLS use. Now it follows the RFC and
4328 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4329 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4330 names appear in log lines as well as in Received: header lines.
4332 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4333 extensions, and comparison tests, and to fix some bugs.
4335 14. Changes to the "personal" filter test:
4337 (1) The test was buggy in that it was just doing the equivalent of
4338 "contains" tests on header lines. For example, if a user's address was
4339 anne@some.where, the "personal" test would incorrectly be true for
4341 To: susanne@some.where
4343 This test is now done by extracting each address from the header in turn,
4344 and checking the entire address. Other tests that are part of "personal"
4345 are now done using regular expressions (for example, to check local parts
4346 of addresses in From: header lines).
4348 (2) The list of non-personal local parts in From: addresses has been
4349 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4350 taken from the Sieve specification recommendations.
4352 (3) If the message contains any header line starting with "List-" it is
4353 treated as non-personal.
4355 (4) The test for "circular" in the Subject: header line has been removed
4356 because it now seems ill-conceived.
4358 15. Minor typos in src/EDITME comments corrected.
4360 16. Installed latest exipick from John Jetmore.
4362 17. If headers_add on a router specified a text string that was too long for
4363 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4364 of string_sprintf() is now avoided.
4366 18. $message_body_size was not set (it was always zero) when running the DATA
4367 ACL and the local_scan() function.
4369 19. For the "mail" command in an Exim filter, no default was being set for
4370 the once_repeat time, causing a random time value to be used if "once" was
4371 specified. (If the value happened to be <= 0, no repeat happened.) The
4372 default is now 0s, meaning "never repeat". The "vacation" command was OK
4373 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4374 (I found it when inspecting the code).
4376 20. There is now an overall timeout for performing a callout verification. It
4377 defaults to 4 times the callout timeout, which applies to individual SMTP
4378 commands during the callout. The overall timeout applies when there is more
4379 than one host that can be tried. The timeout is checked before trying the
4380 next host. This prevents very long delays if there are a large number of
4381 hosts and all are timing out (e.g. when the network connections are timing
4382 out). The value of the overall timeout can be changed by specifying an
4383 additional sub-option for "callout", called "maxwait". For example:
4385 verify = sender/callout=5s,maxwait=20s
4387 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4388 to the end before writing, but this should make it even safer).
4390 22. Exim was forgetting that it had advertised PIPELINING for the second and
4391 subsequent messages on an SMTP connection. It was also not resetting its
4392 memory on STARTTLS and an internal HELO.
4394 23. When Exim logs an SMTP synchronization error within a session, it now
4395 records whether PIPELINING has been advertised or not.
4397 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4398 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4399 rather than long int.
4401 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4403 26. Added the never_mail option to autoreply.
4409 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4410 crash if the getsockname() call failed; this can happen if a connection is
4411 closed very soon after it is established. The problem was simply in the
4412 order in which certain operations were done, causing Exim to try to write
4413 to the SMTP stream before it had set up the file descriptor. The bug has
4414 been fixed by making things happen in the correct order.
4420 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4421 before the connection was closed, thus losing the rejection response.
4423 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4424 some early Solaris releases, but causes trouble in current releases where
4425 socklen_t is defined.
4427 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4430 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4433 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4434 that releasing the top of it at the end releases what was used for sub-
4435 expansions (unless the block got too big). However, discard this block if
4436 the first thing is a variable or header, so that we can use its block when
4437 it is dynamic (useful for very large $message_headers, for example).
4439 6. Lookups now cache *every* query, not just the most recent. A new, separate
4440 store pool is used for this. It can be recovered when all lookup caches are
4441 flushed. Lookups now release memory at the end of their result strings.
4442 This has involved some general refactoring of the lookup sources.
4444 7. Some code has been added to the store_xxx() functions to reduce the amount
4445 of flapping under certain conditions.
4447 8. log_incoming_interface used to affect only the <= reception log lines. Now
4448 it causes the local interface and port to be added to several more SMTP log
4449 lines, for example "SMTP connection from", and rejection lines.
4451 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4453 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4455 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4456 could be overwritten at the end of the current message (or the start of a
4457 new message if it was set in a HELO ACL). The value is now preserved for
4458 the duration of the SMTP connection.
4460 12. If a transport had a headers_rewrite setting, and a matching header line
4461 contained an unqualified address, that address was qualified, even if it
4462 did not match any rewriting rules. The underlying bug was that the values
4463 of the flags that permit the existence of unqualified sender and recipient
4464 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4465 non-local messages, and by -bnq for local messages) were not being
4466 preserved with the message after it was received.
4468 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4469 "next input=" as part of the text comprising the host identity instead of
4470 the correct text. The code was using the same buffer for two different
4471 strings. However, depending on which order the printing function evaluated
4472 its arguments, the bug did not always show up. Under Linux, for example, my
4473 test suite worked just fine.
4475 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4476 doesn't work with some older versions of Perl. It has been changed to "my",
4477 which in any case is probably the better facility to use.
4479 15. A really picky compiler found some instances of statements for creating
4480 error messages that either had too many or two few arguments for the format
4483 16. The size of the buffer for calls to the DNS resolver has been increased
4484 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4485 for addresses that have a lot of PTR records. This alleviates a problem; it
4486 does not fully solve it.
4488 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4489 buffer now truncates the list and logs the incident, which is the same
4490 action as happens when Exim is looking up a host name and its aliases.
4491 Previously in this situation something unpredictable would happen;
4492 sometimes it was "internal error: store_reset failed".
4494 18. If a server dropped the connection unexpectedly when an Exim client was
4495 using GnuTLS and trying to read a response, the client delivery process
4496 crashed while trying to generate an error log message.
4498 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4499 string, for example:
4501 warn message = H1: something\nH2: something
4503 the text was added as a single header line from Exim's point of view
4504 though it ended up OK in the delivered message. However, searching for the
4505 second and subsequent header lines using $h_h2: did not work. This has been
4506 fixed. Similarly, if a system filter added multiple headers in this way,
4507 the routers could not see them.
4509 20. Expanded the error message when iplsearch is called with an invalid key to
4510 suggest using net-iplsearch in a host list.
4512 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4513 ACLs are no longer actually imposed (and a message to that effect is
4516 22. If a "gecos" field in a passwd entry contained escaped characters, in
4517 particular, if it contained a \" sequence, Exim got it wrong when building
4518 a From: or a Sender: header from that name. A second bug also caused
4519 incorrect handling when an unquoted " was present following a character
4520 that needed quoting.
4522 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4523 was not being matched caselessly.
4525 24. Arranged for all hyphens in the exim.8 source to be escaped with
4528 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4529 properly. Recipient callout cache records were still being keyed to include
4530 the sender, even when use_sender was set false. This led to far more
4531 callouts that were necessary. The sender is no longer included in the key
4532 when use_sender is false.
4534 26. Added "control = submission" modifier to ACLs.
4536 27. Added the ${base62d: operator to decode base 62 numbers.
4538 28. dnsdb lookups can now access SRV records.
4540 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4541 the configuration file.
4543 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4544 (-v) mode. This makes the output for a verbose queue run more intelligible.
4546 31. Added a use_postmaster feature to recipient callouts.
4548 32. Added the $body_zerocount variable, containing the number of binary zero
4549 bytes in the message body.
4551 33. The time of last modification of the "new" subdirectory is now used as the
4552 "mailbox time last read" when there is a quota error for a maildir
4555 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4557 35. Added +ignore_unknown as a special item in host lists.
4559 36. Code for decoding IPv6 addresses in host lists is now included, even if
4560 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4561 address was recognized as an IP address, but was then not correctly decoded
4562 into binary, causing unexpected and incorrect effects when compared with
4569 1. Very minor rewording of debugging text in manualroute to say "list of
4570 hosts" instead of "hostlist".
4572 2. If verify=header_syntax was set, and a header line with an unqualified
4573 address (no domain) and a large number of spaces between the end of the
4574 name and the colon was received, the reception process suffered a buffer
4575 overflow, and (when I tested it) crashed. This was caused by some obsolete
4576 code that should have been removed. The fix is to remove it!
4578 3. When running in the test harness, delay a bit after writing a bounce
4579 message to get a bit more predictability in the log output.
4581 4. Added a call to search_tidyup() just before forking a reception process. In
4582 theory, someone could use a lookup in the expansion of smtp_accept_max_
4583 per_host which, without the tidyup, could leave open a database connection.
4585 5. Added the variables $recipient_data and $sender_data which get set from a
4586 lookup success in an ACL "recipients" or "senders" condition, or a router
4587 "senders" option, similar to $domain_data and $local_part_data.
4589 6. Moved the writing of debug_print from before to after the "senders" test
4592 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4593 problems for message scanning, either using a data ACL, or using
4594 local_scan() because the Received: header was not generated till after they
4595 were called (in order to set the time as the time of reception completion).
4596 I have revised the way this works. The header is now generated after the
4597 body is received, but before the ACL or local_scan() are called. After they
4598 are run, the timestamp in the header is updated.
4604 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4605 before starting a queue runner without re-exec. This happened only when
4606 deliver_drop_privilege was set or when the Exim user was set to root. The
4607 effect of the bug was that timeouts during subsequent deliveries caused
4608 crashes instead of being properly handled. The handler is now left at its
4609 default (and expected) setting.
4611 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4612 message, again when deliver_drop_privilege is set or Exim is run as root.
4613 The bug described in (1) was not present in this case, but the tidying up
4614 of the other signals was missing. I have made the two cases consistent.
4616 3. The ignore_target_hosts setting on a manualroute router was being ignored
4617 for hosts that were looked up using the /MX notation.
4619 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4622 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4623 operated on the sender address. After changing the $sender_address to <>
4624 for the sender address verify, Exim was re-instated it as the original
4625 (before rewriting) address, but remembering that it had rewritten it, so it
4626 wasn't rewriting it again. This bug also had the effect of breaking the
4627 sender address verification caching when the sender address was rewritten.
4629 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4630 This has been changed so that if the ip literal address matches
4631 ignore_target_hosts, the router declines.
4633 7. Added expansion conditions match_domain, match_address, and match_local_
4634 part (NOT match_host).
4636 8. The placeholder for the Received: header didn't have a length field set.
4638 9. Added code to Exim itself and to exim_lock to test for a specific race
4639 condition that could lead to file corruption when using MBX delivery. The
4640 issue is with the lockfile that is created in /tmp. If this file is removed
4641 after a process has opened it but before that process has acquired a lock,
4642 there is the potential for a second process to recreate the file and also
4643 acquire a lock. This could lead to two Exim processes writing to the file
4644 at the same time. The added code performs the same test as UW imapd; it
4645 checks after acquiring the lock that its file descriptor still refers to
4646 the same named file.
4648 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4649 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4652 11. Added the smtp_active_hostname option. If used, this will typically be made
4653 to depend on the incoming interface address. Because $interface_address is
4654 not set up until the daemon has forked a reception process, error responses
4655 that can happen earlier (such as "too many connections") no longer contain
4658 12. If an expansion in a condition on a "warn" statement fails because a lookup
4659 defers, the "warn" statement is abandoned, and the next ACL statement is
4660 processed. Previously this caused the whole ACL to be aborted.
4662 13. Added the iplsearch lookup type.
4664 14. Added ident_timeout as a log selector.
4666 15. Added tls_certificate_verified as a log selector.
4668 16. Added a global option tls_require_ciphers (compare the smtp transport
4669 option of the same name). This controls incoming TLS connections.
4671 17. I finally figured out how to make tls_require_ciphers do a similar thing
4672 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4673 before starting the TLS session.
4675 18. Tabs are now shown as \t in -bP output.
4677 19. If the log selector return_path_on_delivery was set, Exim crashed when
4678 bouncing a message because it had too many Received: header lines.
4680 20. If two routers both had headers_remove settings, and the first one included
4681 a superfluous trailing colon, the final name in the first list and the
4682 first name in the second list were incorrectly joined into one item (with a
4683 colon in the middle).
4689 1. Added -C and -D options to the exinext utility, mainly to make it easier
4690 to include in the automated testing, but these could be helpful when
4691 multiple configurations are in use.
4693 2. The exinext utility was not formatting the output nicely when there was
4694 an alternate port involved in the retry record key, nor when there was a
4695 message id as well (for retries that were specific to a specific message
4696 and a specific host). It was also confused by IPv6 addresses, because of
4697 the additional colons they contain. I have fixed the IPv4 problem, and
4698 patched it up to do a reasonable job for IPv6.
4700 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4701 delivery, the log line now contains "pipelined" if PIPELINING was used.
4703 4. An SMTP transport process used to panic and die if the bind() call to set
4704 an explicit outgoing interface failed. This has been changed; it is now
4705 treated in the same way as a connect() failure.
4707 5. A reference to $sender_host_name in the part of a conditional expansion
4708 that was being skipped was still causing a DNS lookup. This no longer
4711 6. The def: expansion condition was not recognizing references to header lines
4712 that used bh_ and bheader_.
4714 7. Added the _cache feature to named lists.
4716 8. The code for checking quota_filecount in the appendfile transport was
4717 allowing one more file than it should have been.
4719 9. For compatibility with Sendmail, the command line option
4727 and sets the incoming protocol and host name (for trusted callers). The
4728 host name and its colon can be omitted when only the protocol is to be set.
4729 Note the Exim already has two private options, -pd and -ps, that refer to
4730 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4731 "s", but I don't think that's a major issue.
4733 10. A number of refactoring changes to the code, none of which should affect
4736 (a) The number of logging options was getting close to filling up the
4737 32-bit word that was used as a bit map. I have split them into two classes:
4738 those that are passed in the argument to log_write(), and those that are
4739 only ever tested independently outside of that function. These are now in
4740 separate 32-bit words, so there is plenty of room for expansion again.
4741 There is no change in the user interface or the logging behaviour.
4743 (b) When building, for example, log lines, the code previously used a
4744 macro that called string_cat() twice, in order to add two strings. This is
4745 not really sufficiently general. Furthermore, there was one instance where
4746 it was actually wrong because one of the argument was used twice, and in
4747 one call a function was used. (As it happened, calling the function twice
4748 did not affect the overall behaviour.) The macro has been replaced by a
4749 function that can join an arbitrary number of extra strings onto a growing
4752 (c) The code for expansion conditions now uses a table and a binary chop
4753 instead of a serial search (which was left over from when there were very
4754 few conditions). Also, it now recognizes conditions like "pam" even when
4755 the relevant support is not compiled in: a suitably worded error message is
4756 given if an attempt is made to use such a condition.
4758 11. Added ${time_interval:xxxxx}.
4760 12. A bug was causing one of the ddress fields not to be passed back correctly
4761 from remote delivery subprocesses. The field in question was not being
4762 subsequently used, so this caused to problems in practice.
4764 13. Added new log selectors queue_time and deliver_time.
4766 14. Might have fixed a bug in maildirsizefile handling that threw up
4767 "unexpected character" debug warnings, and recalculated the data
4768 unnecessarily. In any case, I expanded the warning message to give more
4771 15. Added the message "Restricted characters in address" to the statements in
4772 the default ACL that block characters like @ and % in local parts.
4774 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4775 Three changes have been made:
4777 (a) There was a serious bug; a negative response to MAIL caused the whole
4778 recipient domain to be cached as invalid, thereby blocking all messages
4779 to all local parts at the same domain, from all senders. This bug has
4780 been fixed. The domain is no longer cached after a negative response to
4781 MAIL if the sender used is not empty.
4783 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4786 (c) A new callout option, "use_sender" has been added for people who want
4787 the modified behaviour.
4793 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4796 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4797 indeed breaks things for older releases.
4799 3. Added additional logging to the case where there is a problem reading data
4800 from a filter that is running in a subprocess using a pipe, in order to
4801 try to track down a specific problem.
4803 4. Testing facility fudge: when running in the test harness and attempting
4804 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4805 getting "No route to host". Convert this to a timeout.
4807 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4810 6. Some OS don't have socklen_t but use size_t instead. This affects the
4811 fifth argument of getsockopt() amongst other things. This is now
4812 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4813 can be set for individual OS. I have set it for SunOS5, OSF1, and
4814 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4815 some earlier ones do not.
4817 7. Change 4.30/15 was not doing the test caselessly.
4819 8. The standard form for an IPv6 address literal was being rejected by address
4820 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4821 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4822 this, as well as the form without the "IPv6" on the front (but only when
4823 address literals are enabled, of course).
4825 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4827 10. Exim crashed if a message with an empty sender address specified by -f
4828 encountered a router with an errors_to setting. This could be provoked only
4829 by a command such as
4833 where an empty string was supplied; "<>" did not hit this bug.
4835 11. Installed PCRE release 4.5.
4837 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4838 remained set. It is now erased.
4840 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4841 times from message ids (which are base 36 rather than the normal 62).
4843 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4844 were being counted as actual protocol errors, and logged if the log
4845 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4846 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4847 and DATA following a set of rejected RCPTs do not count as protocol errors.
4848 In other words, Exim assumes they were pipelined, though this may not
4849 actually be the case. Of course, in all cases the client gets an
4850 appropriate error code.
4852 15. If a lookup fails in an ACL condition, a message about the failure may
4853 be available; it is used if testing the ACL cannot continue, because most
4854 such messages specify what the cause of the deferral is. However, some
4855 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4856 that caused an old message to be retained and used if a later statement
4857 caused a defer, replacing the real cause of the deferral.
4859 16. If an IP address had so many PTR records that the DNS lookup buffer
4860 was not large enough to hold them, Exim could crash while trying to process
4861 the truncated data. It now detects and logs this case.
4863 17. Further to 4.21/58, another change has been made: if (and only if) the
4864 first line of a message (the first header line) ends with CRLF, a bare LF
4865 in a subsequent header line has a space inserted after it, so as not to
4866 terminate the header.
4868 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4869 unnecessarily, used atoi() instead of strtol(), and didn't check the
4870 termination when getting file sizes from file names by regex.
4872 19. Completely re-implemented the support for maildirsize files, in the light
4873 of a number of problems with the previous contributed implementation
4874 (4.30/29). In particular:
4876 . If the quota is zero, the maildirsize file is maintained, but no quota is
4879 . If the maildir directory does not exist, it is created before any attempt
4880 to write a maildirsize file.
4882 . The quota value in the file is just a cache; if the quota is changed in
4883 the transport, the new value overrides.
4885 . A regular expression is available for excluding directories from the
4888 20. The autoreply transport checks the characters in options that define the
4889 message's headers; it allows continued headers, but it was checking with
4890 isspace() after an embedded newline instead of explicitly looking for a
4893 21. If all the "regular" hosts to which an address was routed had passed their
4894 expiry times, and had not reached their retry times, the address was
4895 bounced, even if fallback hosts were defined. Now Exim should go on to try
4898 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4899 equivalent code in the SMTP transport. Some hosts send humungous responses
4900 to HELO/EHLO, more than 1024 it seems.
4902 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4903 gives compiler warnings in some environments. I've now done it "properly",
4906 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4907 (because of problems with the built-in one) was declared to return uschar *
4908 instead of char *, causing compiler failure.
4910 25. Fixed a file descriptor leak when processing alias/forward files.
4912 26. Fixed a minor format string issue in dbfn.c.
4914 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4916 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4917 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4920 29. When a sender address is verified, it is cached, to save repeating the test
4921 when there is more than one recipient in a message. However, when the
4922 verification involves a callout, it is possible for different callout
4923 options to be set for different recipients. It is too complicated to keep
4924 track of this in the cache, so now Exim always runs a verification when a
4925 callout is required, relying on the callout cache for the optimization.
4926 The overhead is duplication of the address routing, but this should not be
4929 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4930 to be verified with callout=postmaster, and the main callout worked but the
4931 postmaster check failed, the verification correctly failed. However, if a
4932 subsequent RCPT command asked for sender verification *without* the
4933 postmaster check, incorrect caching caused this verification also to fail,
4936 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4937 it was not caching the DNS options (qualify_single, search_parents) that
4938 were used when the lookup failed. A subsequent lookup with different
4939 options therefore always gave the same answer, though there were cases
4940 where it should not have. (Example: a "domains = !$mx_any" option on a
4941 dnslookup router: the "domains" option is always processed without any
4942 widening, but the router might have qualify_single set.) Now Exim uses the
4943 cached value only when the same options are set.
4945 32. Added John Jetmore's "exipick" utility to the distribution.
4947 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4948 than a timeout (e.g. a certificate is required, and is not provided), an
4949 Exim server now closes the connection immediately. Previously it waited for
4950 the client to close - but if the client is SSL, it seems that they each
4951 wait for each other, leading to a delay before one of them times out.
4953 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4954 maintained 0.8.x compatibility because I don't think many are using it, and
4955 it is clearly obsolete.
4957 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4960 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4961 client certificate was expired. A simple patch fixes this, though I don't
4962 understand the full logic of why the verify callback is called multiple
4965 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4966 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4967 which causes problems with some clients (such as the Certicom SSL Plus
4968 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4969 disables the coutermeasure allowing Eudora to connect."
4971 38. Exim was not checking that a write() to a log file succeeded. This could
4972 lead to Bad Things if a log got too big, in particular if it hit a file
4973 size limit. Exim now panics and dies if it cannot write to a log file, just
4974 as it does if it cannot open a log file.
4976 39. Modified OS/Makefile-Linux so that it now contains
4978 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4980 The two -D definitions ensure that Exim is compiled with large file
4981 support, which makes it possible to handle log files that are bigger than
4984 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4985 instance) a domain was checked against a named list that involved a lookup,
4986 causing $domain_data to be set, then another domain was checked against the
4987 same list, then the first domain was re-checked, the value of $domain_data
4988 after the final check could be wrong. In particular, if the second check
4989 failed, it could be set empty. This bug probably also applied to
4992 41. The strip_trailing_dot option was not being applied to the address given
4993 with the -f command-line option.
4995 42. The code for reading a message's header from the spool was incrementing
4996 $received_count, but never initializing it. This meant that the value was
4997 incorrect (doubled) while delivering a message in the same process in which
4998 it was received. In the most common configuration of Exim, this never
4999 happens - a fresh exec is done - but it can happen when
5000 deliver_drop_privilege is set.
5002 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5003 it now includes up to 150 characters of the unexpected data in the log
5006 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5007 and building data strings. The size of both of these buffers was 10 000
5008 bytes - far larger than anybody would *ever* want, thought I. Needless to
5009 say, somebody hit the limit. I have increased the maximum line length to
5010 20 000 and the maximum data length of concatenated lines to 100 000. I have
5011 also fixed two bugs, because there was no checking on these buffers. Tsk,
5012 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5013 buffer is too small.
5015 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5016 lsearch lookups. Now it does.
5018 46. When parsing a route_list item in a manualroute router, a fixed-length
5019 buffer was used for the list of hosts. I made this 1024 bytes long,
5020 thinking that nobody would ever have a list of hosts that long. Wrong.
5021 Somebody had a whole pile of complicated expansion conditions, and the
5022 string was silently truncated, leading to an expansion error. It turns out
5023 that it is easier to change to an unlimited length (owing to other changes
5024 that have happened since this code was originally written) than to build
5025 structure for giving a limitation error. The length of the item that
5026 expands into the list of hosts is now unlimited.
5028 47. The lsearch lookup could not handle data where the length of text line was
5029 more than 4095 characters. Such lines were truncated, leading to shortened
5030 data being returned. It should now handle lines of any length.
5032 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5033 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5036 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5037 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5038 no longer happens in comment lines. A list of replacements is now placed
5039 at the head of all of the source files, except those whose only change is
5040 to replace PERL_COMMAND in the very first #! line.
5042 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5043 messages on the queue, with a bottom-up merge sort, using code contributed
5044 by Michael Haardt. This should make operations like -bp somewhat faster on
5045 large queues. It won't affect queue runners, except when queue_run_in_order
5048 51. Installed eximstats 1.31 in the distribution.
5050 52. Added support for SRV lookups to the dnslookup router.
5052 53. If an ACL referred to $message_body or $message_body_end, the value was not
5053 reset for any messages that followed in the same SMTP session.
5055 54. The store-handling optimization for building very long strings was not
5056 differentiating between the different store pools. I don't think this
5057 actually made any difference in practice, but I've tidied it.
5059 55. While running the routers to verify a sender address, $sender_address
5060 was still set to the sender address. This is wrong, because when routing to
5061 send a bounce to the sender, it would be empty. Therefore, I have changed
5062 it so that, while verifying a sender address, $sender_address is set to <>.
5063 (There is no change to what happens when verifying a recipient address.)
5065 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5066 target A or AAAA records (if not already returned) without resetting the
5067 qualify_single or search_parents options of the DNS resolver. These are
5068 inappropriate in this case because the targets of MX and SRV records must
5069 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5070 target that, when qualified, matched something in the local domain. These
5071 two options are now turned off when doing these lookups.
5073 57. It seems that at least some releases of Reiserfs (which does not have the
5074 concept of a fixed number of inodes) returns zero and not -1 for the
5075 number of available inodes. This interacted badly with check_spool_inodes,
5076 which assumed that -1 was the "no such thing" setting. What I have done is
5077 to check that the total number of inodes is greater than zero before doing
5078 the test of how many are available.
5080 58. When a "warn" ACL statement has a log_message modifier, the message is
5081 remembered, and not repeated. This is to avoid a lot of repetition when a
5082 message has many recipients that cause the same warning to be written.
5083 Howewer, Exim was preserving the list of already written lines for an
5084 entire SMTP session, which doesn't seem right. The memory is now reset if a
5085 new message is started.
5087 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5088 debugging output unless log_rewrite was also set.
5090 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5091 of (int)(handle) when we know that handle contains (void *)(-1).
5093 61. The Exim daemon panic-logs an error return when it closes the incoming
5094 connection. However "connection reset by peer" seems to be common, and
5095 isn't really an error worthy of noting specially, so that particular error
5098 62. When Exim is trying to find all the local interfaces, it used to panic and
5099 die if the ioctl to get the interface flags failed. However, it seems that
5100 on at least one OS (Solaris 9) it is possible to have an interface that is
5101 included in the list of interfaces, but for which you get a failure error
5102 for this call. This happens when the interface is not "plumbed" into a
5103 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5104 failure of the "get flags" call assumes that the interface is down.
5106 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5107 makes life easier for people who are doing arithmetic on fields extracted
5108 from dates, where you often get leading zeros that should not be
5109 interpreted as octal.
5111 64. Added qualify_domain to the redirect router, to override the global
5114 65. If a pathologically long header line contained very many addresses (the
5115 report of this problem mentioned 10 000) and each of them was rewritten,
5116 Exim could use up a very large amount of memory. (It kept on making new
5117 copies of the header line as it rewrote, and never released the old ones.)
5118 At the expense of a bit more processing, the header rewriting function has
5119 been changed so that it no longer eats memory in this way.
5121 66. The generation of the Received: header has been moved from the time that a
5122 message starts to be received, to the time that it finishes. The timestamp
5123 in the Received: header should now be very close to that of the <= log
5124 line. There are two side-effects of this change:
5126 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5127 logged header lines no longer include the local Received: line, because
5128 it has not yet been created. The same applies to a copy of the message
5129 that is returned to a non-SMTP sender when a message is rejected.
5131 (b) When a filter file is tested using -bf, no additional Received: header
5132 is added to the test message. After some thought, I decided that this
5135 This change does not affect the value of $received_for. It is still set
5136 after address rewriting, but before local_scan() is called.
5138 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5140 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5141 gave an unhelpful panic error message, and a defer error. I have managed to
5142 change this behaviour so that it now rejects any supplied certificate,
5143 which seems right, as the list of acceptable certificates is empty.
5145 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5146 gave an unhelpful defer error. I have not managed to make this reject any
5147 supplied certificates, but the error message it gives is "no certificate
5148 supplied", which is not helpful.
5150 70. exigrep's output now also includes lines that are not associated with any
5151 message, but which match the given pattern. Implemented by a patch from
5152 Martin Sluka, which also tidied up the Perl a bit.
5154 71. Recipient callout verification, like sender verification, was using <> in
5155 the MAIL FROM command. This isn't really the right thing, since the actual
5156 sender may affect whether the remote host accepts the recipient or not. I
5157 have changed it to use the actual sender in the callout; this means that
5158 the cache record is now keyed on a recipient/sender pair, not just the
5159 recipient address. There doesn't seem to be a real danger of callout loops,
5160 since a callout by the remote host to check the sender would use <>.
5161 [SEE ABOVE: changed after hitting problems.]
5163 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5164 temporary errors. However, in the case of such a code being given after
5165 the end of a data transmission (i.e. after ".") Exim was failing to write
5166 a retry record for the message. (Yes, there was some broken host that was
5167 actually sending 8xx at this point.)
5169 73. An unknown lookup type in a host list could cause Exim to panic-die when
5170 the list was checked. (An example that provoked this was putting <; in the
5171 middle of a list instead of at the start.) If this happened during a DATA
5172 ACL check, a -D file could be left lying around. This kind of configuration
5173 error no longer causes Exim to die; instead it causes a defer errror. The
5174 incident is still logged to the main and panic logs.
5176 74. Buglet left over from Exim 3 conversion. The message "too many messages
5177 in one connection" was written to the rejectlog but not the mainlog, except
5178 when address rewriting (yes!) was being logged.
5180 75. Added write_rejectlog option.
5182 76. When a system filter was run not as root (that is, when system_filter_user
5183 was set), the values of the $n variables were not being returned to the
5184 main process; thus, they were not subsequently available in the $sn
5187 77. Added +return_path_on_delivery log selector.
5189 78. A connection timeout was being treated differently from recipients deferred
5190 when testing hosts_max_try with a message that was older than the host's
5191 retry timeout. (The host should not be counted, thus allowing all hosts to
5192 be tried at least once before bouncing.) This may have been the cause of an
5193 occasionally reported bug whereby a message would remain on the queue
5194 longer than the retry timeout, but would be bounced if a delivery was
5195 forced. I say "may" because I never totally pinned down the problem;
5196 setting up timeout/retry tests is difficult. See also the next item.
5198 79. The ultimate address timeout was not being applied to errors that involved
5199 a combination of host plus message (for example, a timeout on a MAIL
5200 command). When an address resolved to a number of possible hosts, and they
5201 were not all tried for each delivery (e.g. because of hosts_max_try), a
5202 message could remain on the queue longer than the retry timeout.
5204 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5207 81. Fixed an obscure SMTP outgoing bug which required at least the following
5208 conditions: (a) there was another message waiting for the same server;
5209 (b) the server returned 5xx to all RCPT commands in the first message so
5210 that the message was not completed; (c) the server dropped the connection
5211 or gave a negative response to the RSET that Exim sends to abort the
5212 transaction. The observed case was a dropped connection after DATA that had
5213 been sent in pipelining mode. That is, the server had advertised PIPELINING
5214 but was not implementing it correctly. The effect of the bug was incorrect
5215 behaviour, such as trying another host, and this could lead to a crash.
5221 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5222 and daemon.c were passed as pointers to ints; they should have been
5223 pointers to socklen_t variables (which are typically unsigned ints).
5225 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5228 3. Fixed a really odd bug that affected only the testing scheme; patching a
5229 certain fixed string in the binary changed the value of another string that
5230 happened to be identical to the end of the original first string.
5232 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5233 name", it returns that address as the IP address. On some operating
5234 systems (e.g. Solaris), it also passes back the IP address string as the
5235 "host name". However, on others (e.g. Linux), it passes back an empty
5236 string. Exim wasn't checking for this, and was changing the host name to an
5237 empty string, assuming it had been canonicized.
5239 5. Although rare, it is permitted to have more than one PTR record for a given
5240 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5241 all the names associated with an address, because they do in Solaris.
5242 However, it seems that they do not in Linux for data that comes from the
5243 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5244 I found this out when I moved to a new Linux workstation and tried to run
5245 the Exim test suite.
5247 To get round this problem I have changed the code so that it now does its
5248 own call to the DNS to look up PTR records when searching for a host name.
5249 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5250 addresses that are only in /etc/hosts are still found.
5252 This behaviour is, however, controlled by an option called host_lookup_
5253 order, which defaults to "bydns:byaddr". If people want to use the other
5254 order, or indeed, just use one or the other means of lookup, they can
5255 specify it in this variable.
5257 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5258 some operating systems, this comes back from gethostbyaddr() as an empty
5259 string, and this is what Exim used to test for. However, it seems that in
5260 other systems, "." is yielded. Exim now tests for this case too.
5262 7. The values of check_spool_space and check_log_space are now held internally
5263 as a number of kilobytes instead of an absolute number of bytes. If a
5264 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5265 kilobyte. This means that much larger values can be stored.
5267 8. Exim monitor: an attempt to get the action menu when not actually pointing
5268 at a message produces an empty menu entitled "No message selected". This
5269 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5270 no entries in it ("Shell widget menu has zero width and/or height"). So I
5271 have added a single, blank menu entry in this case.
5273 9. Added ${quote_local_part.
5275 10. MIME decoding is now applied to the contents of Subject: header lines when
5278 11. Now that a reference to $sender_host_address automatically causes a reverse
5279 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5280 host lookup before query-style lookups in lists that might use this
5281 variable. This has therefore been abolished, and the "net-" prefix is no
5282 longer necessary for query-style lookups.
5284 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5285 have been a typo for LFLAGS, so it has been changed.
5287 13. The install script calls Exim with "-C /dev/null" in order to find the
5288 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5289 to be output. Howeve, since Exim outputs its version number before the
5290 error, it didn't break the script. It just looked ugly. I fixed this by
5291 always allowing "-C /dev/null" if the caller is root.
5293 14. Ignore overlarge ACL variable number when reading spool file - insurance
5294 against a later release with more variables having written the file.
5296 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5297 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5298 this, as well as the form without the "IPv6" on the front.
5300 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5301 OS/Makefile-Darwin file.
5303 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5304 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5306 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5307 variable. However, some OS use ints for inodes. I've added cast to long int
5308 to get rid of the compiler warning.
5310 19. I had forgotten to lock out "/../" in configuration file names when
5311 ALT_CONFIG_PREFIX was set.
5313 20. Routers used for verification do not need to specify transports. However,
5314 if such a router generated a host list, and callout was configured, Exim
5315 crashed, because it could not find a port number from the (non-existent)
5316 transport. It now assumes port 25 in this circumstance.
5318 21. Added the -t option to exigrep.
5320 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5321 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5322 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5323 them is set, LOOKUP_LSEARCH is forced.
5325 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5327 24. Added sender and host information to the "rejected by local_scan()" log
5328 line; previously there was no indication of these.
5330 25. Added .include_if_exists.
5332 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5333 Linux. It turns out that not all file systems support this. Apparently some
5334 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5335 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5336 not supported on the file descriptor, is now ignored when Exim is trying to
5337 sync a directory. This applies only to Linux.
5339 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5341 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5342 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5345 29. Added support for maildirsize files from supplied patch (modified a bit).
5347 30. The use of :fail: followed by an empty string could lead Exim to respond to
5348 sender verification failures with (e.g.):
5350 550 Verification failed for <xxx>
5351 550 Sender verify failed
5353 where the first response line was missing the '-' that indicates it is not
5354 the final line of the response.
5356 31. The loop for finding the name of the user that called Exim had a hardwired
5357 limit of 10; it now uses the value of finduser_retries, which is used for
5358 all other user lookups.
5360 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5363 33. Exim was neglecting to zero errno before one call of strtol() when
5364 expanding a string and expecting an integer value. On some systems this
5365 resulted in spurious "integer overflow" errors. Also, it was casting the
5366 result into an int without checking.
5368 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5369 did not work. The code looks as if it has *never* worked, though it appears
5370 to have been documented since at least releast 1.62. I have made it work.
5372 35. The "timeout_DNS" error in retry rules, also documented since at least
5373 1.62, also never worked. As it isn't clear exactly what this means, and
5374 clearly it isn't a major issue, I have abolished the feature by treating it
5375 as "timeout", and writing a warning to the main and panic logs.
5377 36. The display of retry rules for -brt wasn't always showing the error code
5380 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5381 timeout_connect_A, timeout_connect_MX.
5383 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5384 to the empty sender.
5386 39. The daemon was not analysing the content of -oX till after it had closed
5387 stderr and disconnected from the controlling terminal. This meant that any
5388 syntax errors were only noted on the panic log, and the return code from
5389 the command was 0. By re-arranging the code a little, I've made the
5390 decoding happen first, so such errors now appear on stderr, and the return
5391 code is 1. However, the actual setting up of the sockets still happens in
5392 the disconnected process, so errors there are still only recorded on the
5395 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5396 connections (as happens on some IP stacks) was logged at start up time as
5397 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5398 it from "IPv6 and IPv4", which means that two separate sockets are being
5401 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5402 says whether AF_INET or AF_INET6 was passed as an argument.
5404 42. Exiwhat output was messed up when time zones were included in log
5407 43. Exiwhat now gives more information about the daemon's listening ports,
5408 and whether -tls-on-connect was used.
5410 44. The "port" option of the smtp transport is now expanded.
5412 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5413 silently ignored. Now an error message is written to the main and panic
5416 46. There's a new ACL modifier called "logwrite" which writes to a log file
5417 as soon as it is encountered.
5419 47. Added $local_user_uid and $local_user_gid at routing time.
5421 48. Exim crashed when trying to verify a sender address that was being
5424 49. Exim was recognizing only a space character after ".include". It now also
5425 recognizes a tab character.
5427 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5428 extracting the relevant information from the specification. The man page no
5429 longer contains scrambled data for the -d option, and I've added a section
5430 at the front about calling Exim under different names.
5432 51. Added "extra_headers" argument to the "mail" command in filter files.
5434 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5437 53. Installed eximstats 1.29.
5439 54. Added transport_filter_timeout as a generic transport option.
5441 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5442 Cc: header lines. This was required by RFC 822, but it not required by RFC
5445 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5446 incoming messages that did not have them. Now it does so only if the
5447 message originates locally, that is, if there is no associated remote host
5448 address. When Resent- header lines are present, this applies to the Resent-
5449 lines rather than the non-Resent- lines.
5451 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5452 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5454 58. Messages for configuration errors now include the name of the main
5455 configuration file - useful now that there may be more than one file in a
5456 list (.included file names were always shown).
5458 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5459 for those rare installations that do not start the daemon as root or run it
5460 setuid root. I've cut out the call to initgroups() if the daemon is not
5463 60. The Exim user and group can now be bound into the binary as text strings
5464 that are looked up at the start of Exim's processing.
5466 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5468 62. Added $mailstore_basename variable.
5470 63. Installed patch to sieve.c from Michael Haardt.
5472 64. When Exim failed to open the panic log after failing to open the main log,
5473 the original message it was trying to log was written to stderr and debug
5474 output, but if they were not available (the usual case in production), it
5475 was lost. Now it is written to syslog before the two lines that record the
5476 failures to open the logs.
5478 65. Users' Exim filters run in subprocesses under the user's uid. It is
5479 possible for a "deliver" command or an alias in a "personal" command to
5480 provoke an address rewrite. If logging of address rewriting is configured,
5481 this fails because the process is not running as root or exim. There may be
5482 a better way of dealing with this, but for the moment (because 4.30 needs
5483 to be released), I have disabled address rewrite logging when running a
5484 filter in a non-root, non-exim process.
5490 1. The buildconfig auxiliary program wasn't quoting the value set for
5491 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5492 not defined. This bug was masked in 4.22 by the effect that was fixed in
5495 2. Some messages that were rejected after a message id was allocated were
5496 shown as "incomplete" by exigrep. It no longer does this for messages that
5497 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5499 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5500 have allow_domain_literals set, the ID did not get logged in the <= line.
5501 Domain literals are now always recognized in Message-ID: header lines.
5503 4. The first argument for a ${extract expansion item is the key name or field
5504 number. Leading and trailing spaces in this item were not being ignored,
5505 causing some misleading effects.
5507 5. When deliver_drop_privilege was set, single queue runner processes started
5508 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5509 same command in the process it spins off) were not dropping privilege.
5511 6. When the daemon running as "exim" started a queue runner, it always
5512 re-executed Exim in the spun-off process. This is a waste of effort when
5513 deliver_drop_privilege is set. The new process now just calls the
5514 queue-runner function directly.
5520 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5523 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5524 ignored. Though the use of -oP was forcing the writing of a pid file, it
5525 was always written to the default place.
5527 3. If the message "no IP address found for host xxxx" is generated during
5528 incoming verification, it is now followed by identification of the incoming
5529 connection (so you can more easily find what provoked it).
5531 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5533 5. Added some features to "harden" Exim a bit more against certain attacks:
5535 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5536 be put in Local/Makefile. This is like the never_users runtime option,
5537 but it cannot be overridden. The default setting is "root".
5539 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5540 prefix string with which any file named in a -C command line option
5543 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5544 is retained for -C and -D only if the caller of Exim is root. Without
5545 it, the exim user may also use -C and -D and retain privilege.
5547 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5548 command line option is disabled.
5550 6. Macro names set by the -D option must start with an upper case letter, just
5551 like macro names defined in the configuration file.
5553 7. Added "dereference=" facility to LDAP.
5555 8. Two instances of the typo "uknown" in the source files are fixed.
5557 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5558 the Configure-Makefile script screwed up while processing it.
5560 10. Incorporated PCRE 4.4.
5562 11. The SMTP synchronization check was not operating right at the start of an
5563 SMTP session. For example, it could not catch a HELO sent before the client
5564 waited for the greeting. There is now a check for outstanding input at the
5565 point when the greeting is written. Because of the duplex, asynchronous
5566 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5567 way, but not yet received, when the check is performed.
5569 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5570 on TCP/IP sockets, because this apparently causes some broken clients to
5573 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5574 unchanged) from the Cygwin maintainer.
5576 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5577 was supported instead of testing for mbx. Effectively a typo.
5579 15. The spa authenticator server code was not checking that the input it
5580 received was valid base64.
5582 16. The debug output line for the "set" modifier in ACLs was not showing the
5583 name of the variable that was being set.
5585 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5587 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5588 lookup on its own. Something else was needed to trigger the lookup. For
5589 example, a match in host_lookup or the need for a host name in a host list.
5590 Now, if $sender_host_name is referenced and the host name has not yet been
5591 looked up, a lookup is performed. If the lookup fails, the variable remains
5592 empty, and $host_lookup_failed is set to "1".
5594 19. Added "eqi" as a case-independent comparison operator.
5596 20. The saslauthd authentication condition could segfault if neither service
5597 nor realm was specified.
5599 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5600 error message that was logged was misleading, and incoming SMTP
5601 connections were dropped. The message is now more accurate, and temporary
5602 errors are given to SMTP connections.
5604 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5605 (except RSET) with a 421 error, until QUIT is received. However, it was
5606 failing to send a response to QUIT.
5608 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5609 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5610 ACL is now run after the helo_try_verify_hosts code.
5612 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5613 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5614 case-independent, so other case variants are also recognized). Apparently
5615 some systems use these upper case variants.
5617 25. If more than two messages were waiting for the same host, and a transport
5618 filter was specified for the transport, Exim sent two messages over the
5619 same TCP/IP connection, and then failed with "socket operation on non-
5620 socket" when it tried to send the third.
5622 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5624 27. The extern definition of crypt16() in expand.c was not being excluded when
5625 the OS had its own crypt16() function.
5627 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5628 as a preferred synonym for return_size_limit, both as an option and as an
5631 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5633 30. Changed the default configuration ACL to relax the local part checking rule
5634 for addresses that are not in any local domains. For these addresses,
5635 slashes and pipe symbols are allowed within local parts, but the sequence
5636 /../ is explicitly forbidden.
5638 31. SPA server authentication was not clearing the challenge buffer before
5641 32. log_message in a "warn" ACL statement was writing to the reject log as
5642 well as to the main log, which contradicts the documentation and doesn't
5643 seem right (because no rejection is happening). So I have stopped it.
5645 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5646 However, I am unable to do any testing of this.
5648 34. Fixed an infelicity in the appendfile transport. When checking directories
5649 for a mailbox, to see if any needed to be created, it was accidentally
5650 using path names with one or more superfluous leading slashes; tracing
5651 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5653 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5654 contents are added to the log line that is written for every discarded
5655 recipient. (Previously a log_message setting was ignored.)
5657 36. The ${quote: operator now quotes the string if it is empty.
5659 37. The install script runs exim in order to find its version number. If for
5660 some reason other than non-existence or emptiness, which it checks, it
5661 could not run './exim', it was installing it with an empty version number,
5662 i.e. as "exim-". This error state is now caught, and the installation is
5665 38. An argument was missing from the function that creates an error message
5666 when Exim fails to connect to the socket for saslauthd authentication.
5667 This could cause Exim to crash, or give a corrupted message.
5669 39. Added isip, isip4, and isip6 to ${if conditions.
5671 40. The ACL variables $acl_xx are now saved with the message, and can be
5672 accessed later in routers, transports, and filters.
5674 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5675 strings in the file are not string-expanded.
5677 42. If a MAIL command specified a SIZE value that was too large to fit into an
5678 int variable, the check against message_size_limit failed. Such values are
5679 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5680 day this will have to be increased, but I don't think I want to be around
5681 when emails are that large.
5688 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5689 iconv() is not standard in FreeBSD.
5691 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5692 IPv6 enabled. The observed symptom was a segmentation fault on return from
5693 the function os_common_find_running_interfaces() in src/os.c.
5695 3. In the check_special_case() function in daemon.c I had used "errno" as an
5696 argument name, which causes warnings on some systems. This was basically a
5697 typo, since it was named "eno" in the comments!
5699 4. The code that waits for the clock to tick (at a resolution of some fraction
5700 of a second) so as to ensure message-id uniqueness was always waiting for
5701 at least one whole tick, when it could have waited for less. [This is
5702 almost certainly not relevant at current processor speeds, where it is
5703 unlikely to ever wait at all. But we try to future-proof.]
5705 5. The function that sleeps for a time interval that includes fractions of a
5706 second contained a race. It did not block SIGALRM between setting the
5707 timer, and suspending (a couple of lines later). If the interval was short
5708 and the sigsuspend() was delayed until after it had expired, the suspension
5709 never ended. On busy systems this could lead to processes getting stuck for
5712 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5713 process, before it forks any delivery processes. The open lookup caching
5714 mechanism meant that the open file or database connection was passed into
5715 the delivery process. The problem was that delivery processes always tidy
5716 up cached lookup data. This could cause a problem for the next delivery
5717 process started by the queue runner, because the external queue runner
5718 process does not know about the closure. So the next delivery process
5719 still has data in the lookup cache. In the case of a file lookup, there was
5720 no problem because closing a file descriptor in a subprocess doesn't affect
5721 the parent. However, if the lookup was caching a connection to a database,
5722 the connection was closed, and the second delivery process was likely to
5723 see errors such as "PGSQL: query failed: server closed the connection
5724 unexpectedly". The problem has been fixed by closing all cached lookups
5725 in a queue runner before running a delivery process.
5727 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5728 seem to have the "const" qualifier which it has on other OS. I've
5731 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5732 *of the same type* (client or server) with the same public name that an
5733 error should be diagnosed.
5735 9. When Exim looked up a host name for an IP address, but failed to find the
5736 original IP address when looking up the host name (a safety check), it
5737 output the message "<ip address> does not match any IP for NULL", which was
5738 confusing, to say the least. The bug was that the host name should have
5739 appeared instead of "NULL".
5741 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5742 user that is built into the binary, and the -C or -D options is used, root
5743 privilege is dropped before the configuration file is read. In addition,
5744 logging is switched to stderr instead of the normal log files. If the
5745 configuration then re-defines the Exim user, the unprivileged environment
5746 is probably not what is expected, so Exim logs a panic warning message (but
5749 However, if deliver_drop_privilege is set, the unprivileged state may well
5750 be exactly what is intended, so the warning has been cut out in that case,
5751 and Exim is allowed to try to write to its normal log files.
5757 1. smtp_return_error_details was not giving details for temporary sender
5758 or receiver verification errors.
5760 2. Diagnose a configuration error if two authenticators have the same public
5763 3. Exim used not to create the message log file for a message until the first
5764 delivery attempt. This could be confusing when incoming messages were held
5765 for policy or load reasons. The message log file is now created at the time
5766 the message is received, and an initial "Received" line is written to it.
5768 4. The automatically generated man page for command line options had a minor
5769 bug that caused no ill effects; however, a more serious problem was that
5770 the procedure for building the man page automatically didn't always
5771 operate. Consequently, release 4.20 contains an out-of-date version. This
5772 shouldn't happen again.
5774 5. When building Exim with embedded Perl support, the script that builds the
5775 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5776 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5778 6. The freeze_tell option was not being used for messages that were frozen on
5779 arrival, either by an ACL or by local_scan().
5781 7. Added the smtp_incomplete_transaction log selector.
5783 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5784 was accepting AUTH without a new EHLO.
5786 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5787 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5790 10. Logging of TCP/IP connections (when configured) now happens in the main
5791 daemon process instead of the child process, so that the TCP/IP connection
5792 count is more accurate (but it can never be perfect).
5794 11. The use of "drop" in a nested ACL was not being handled correctly in the
5795 outer ACL. Now, if condition failure induced by the nested "drop" causes
5796 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5797 or "require"), the connection is dropped.
5799 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5800 that yield "discard" can now be used with an "accept" or a "discard" verb,
5801 but an error is generated for any others (because I can't see a useful way
5802 to define what should happen).
5804 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5805 are now processed in the same way as lines in the Exim configuration file.
5806 In particular, continuation lines are supported.
5808 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5810 15. Added -ti meaning -t -i.
5812 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5813 domains, and warn by logging if others are found.
5815 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5816 returned from GIFCONF when Exim is trying to find the list of interfaces on
5817 a host. The code in os.c has been modified to copy each ifreq to an aligned
5818 structure in all cases.
5820 Also, in some cases, the returned ifreq's were being copied to a 'struct
5821 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5822 means the last couple of bytes of an IPv6 address could be chopped if the
5823 ifreq contained only a normal sockaddr (14 bytes storage).
5825 18. Named domain lists were not supported in the hosts_treat_as_local option.
5826 An entry such as +xxxx was not recognized, and was treated as a literal
5829 19. Ensure that header lines added by a DATA ACL are included in the reject log
5830 if the ACL subsequently rejects the message.
5832 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5833 MD5 (which is deprecated).
5835 21. When testing a filter file using -bf, Exim was writing a message when it
5836 took the sender from a "From " line in the message, but it was not doing so
5837 when it took $return_path from a Return-Path: header line. It now does.
5839 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5840 with a valid header line field name (a series of printing characters
5841 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5843 23. Changed "disc" in the source to "disk" to conform to the documentation and
5844 the book and for uniformity.
5846 24. Ignore Sendmail's -Ooption=value command line item.
5848 25. When execve() failed while trying to run a command in a pipe transport,
5849 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5850 could be confused with a return value of 69 from the command itself. This
5851 has been changed to 127, the value the shell returns if it is asked to run
5852 a non-existent command. The wording for the related log line suggests a
5853 non-existent command as the problem.
5855 26. If received_header_text expands to an empty string, do not add a Received:
5856 header line to the message. (Well, it adds a token one on the spool, but
5857 marks it "old" so that it doesn't get used or transmitted.)
5859 27. Installed eximstats 1.28 (addition of -nt option).
5861 28. There was no check for failure on the call to getsockname() in the daemon
5862 code. This can fail if there is a shortage of resources on the system, with
5863 ENOMEM, for example. A temporary error is now given on failure.
5865 29. Contrary to the C standard, it seems that in some environments, the
5866 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5867 program. Exim now does this explicitly; it affects the formatting of
5868 timestamps using strftime().
5870 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5871 complaints. I've now initialized all the variables, to avoid this.
5873 32. Header lines added by a system filter were not being "seen" during
5874 transport-time rewrites.
5876 33. The info_callback() function passed to OpenSSL is set up with type void
5877 (*)(SSL *, int, int), as described somewhere. However, when calling the
5878 function (actually a macro) that sets it up, the type void(*)() is
5879 expected. I've put in a cast to prevent warnings from picky compilers.
5881 34. If a DNS black list lookup found a CNAME record, but there were no A
5882 records associated with the domain it pointed at, Exim crashed.
5884 35. If a DNS black list lookup returned more than one A record, Exim ignored
5885 all but the first. It now scans all returned addresses if a particular IP
5886 value is being sought. In this situation, the contents of the
5887 $dnslist_value variable are a list of all the addresses, separated by a
5890 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5891 to accept a host name and all its aliases if the forward lookup for any of
5892 them yielded the IP address of the incoming connection. Now it accepts only
5893 those names whose forward lookup yields the correct IP address. Any other
5894 names are discarded. This closes a loophole whereby a rogue DNS
5895 administrator could create reverse DNS records to break through a
5896 wildcarded host restriction in an ACL.
5898 37. If a user filter or a system filter that ran in a subprocess used any of
5899 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5900 the wrong values were passed to the pipe command ($thisaddress had the
5901 value of $0, $0 had the value of $1, etc). This bug was introduced by
5902 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5904 38. Improved the line breaking for long SMTP error messages from ACLs.
5905 Previously, if there was no break point between 40 and 75 characters, Exim
5906 left the rest of the message alone. Two changes have been made: (a) I've
5907 reduced the minimum length to 35 characters; (b) if it can't find a break
5908 point between 35 and 75 characters, it looks ahead and uses the first one
5909 that it finds. This may give the occasional overlong line, but at least the
5910 remaining text gets split now.
5912 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5913 file descriptors might be low, and that setting 1000 would always raise it.
5914 It turns out that in some environments, the limit is already over 1000 and
5915 that lowering it causes trouble. So now Exim takes care not to decrease it.
5917 40. When delivering a message, the value of $return_path is set to $sender_
5918 address at the start of routing (routers may change the value). By an
5919 oversight, this default was not being set up when an address was tested by
5920 -bt or -bv, which affected the outcome if any router or filter referred to
5923 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5924 log only when "message" or "log_message" are set. However, if one of the
5925 conditions was an address verification, or a call to a nested ACL, the
5926 messages generated by the underlying test were being passed through. This
5927 no longer happens. The underlying message is available in $acl_verify_
5928 message for both "message" and "log_message" expansions, so it can be
5929 passed through if needed.
5931 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5932 new expansion $bh_ to give the encoded byte string without charset
5933 translation. Translation happens only if iconv() is available; HAVE_ICONV
5934 indicates this at build time. HEADERS_CHARSET gives the charset to
5935 translate to; headers_charset can change it in the configuration, and
5936 "headers charset" can change it in an individual filter file.
5938 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5939 that creates RFC 2047 encoded "words" labels them as that charset instead
5940 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5941 expansion operator; (ii) when Exim creates a From: line for a local
5942 message; (iii) when a header line is rewritten to include a "phrase" part.
5944 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5945 buggy, causing it to skip the first lines of messages whose message ID
5946 ended in 'D'. This would not have bitten before Exim release 4.14, because
5947 message IDs were unlikely to end in 'D' before then. The effect was to have
5948 incorrect size information for certain domains.
5950 45. #include "config.h" was missing at the start of the crypt16.c module. This
5951 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5954 46. If there was a timeout during a "random" callout check, Exim treated it as
5955 a failure of the random address, and carried on sending RSET and the real
5956 address. If the delay was just some slowness somewhere, the response to the
5957 original RCPT would be taken as a response to RSET and so on, causing
5958 mayhem of various kinds.
5960 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5961 when I implemented it. It didn't allow for the fact that some option values
5962 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5963 the right test for positive values.
5965 48. Domain names in DNS records are case-independent. Exim always looks them up
5966 in lower case. Some resolvers return domain names in exactly the case they
5967 appear in the zone file, that is, they may contain uppercase letters. Not
5968 all resolvers do this - some return always lower case. Exim was treating a
5969 change of case by a resolver as a change of domain, similar to a widening
5970 of a domain abbreviation. This triggered its re-routing code and so it was
5971 trying to route what was effectively the same domain again. This normally
5972 caused routing to fail (because the router wouldn't handle the domain
5973 twice). Now Exim checks for this case specially, and just changes the
5974 casing of the domain that it ultimately uses when it transmits the message
5977 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5980 50. If a filter generated a file delivery with a non-absolute name (possible if
5981 no home directory exists for the router), the forbid_file option was not
5984 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5985 the existing equality matching.
5987 52. Exim was using ints instead of ino_t variables in some places where it was
5988 dealing with inode numbers.
5990 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5991 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5992 called TMPDIR, and if it finds it is different, it changes its value.
5994 54. The smtp_printf() function is now made available to local_scan() so
5995 additional output lines can be written before returning. There is also an
5996 smtp_fflush() function to enable the detection of a dropped connection.
5997 The variables smtp_input and smtp_batched_input are exported to
6000 55. Changed the default runtime configuration: the message "Unknown user"
6001 has been removed from the ACL, and instead placed on the localuser router,
6002 using the cannot_route_message feature. This means that any verification
6003 failures that generate their own messages won't get overridden. Similarly,
6004 the "Unrouteable address" message that was in the ACL for unverifiable
6005 relay addresses has also been removed.
6007 56. Added hosts_avoid_esmtp to the smtp transport.
6009 57. The exicyclog script was not checking for the esoteric option
6010 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6011 will work only if exicyclog is run under the appropriate euid.
6013 58. Following a discussion on the list, the rules by which Exim recognises line
6014 endings on incoming messages have been changed. The -dropcr and drop_cr
6015 options are now no-ops, retained only for backwards compatibility. The
6016 following line terminators are recognized: LF CRLF CR. However, special
6017 processing applies to CR:
6019 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6020 nor a local message in the state where . is a terminator.
6022 (ii) If a bare CR is encountered in a header line, an extra space is added
6023 after the line terminator so as not to end the header. The reasoning
6024 behind this is that bare CRs in header lines are most likely either
6025 to be mistakes, or people trying to play silly games.
6027 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6028 was being incorrectly given as 18 bytes larger than it should have been.
6029 This is a VOB (very old bug).
6031 60. This may never have affected anything current, but just in case it has:
6032 When the local host is found other than at the start of a list of hosts,
6033 the local host, those with the same MX, and any that follow, are discarded.
6034 When the list in question was part of a longer list of hosts, the following
6035 hosts (not currently being processed) were also being discarded. This no
6036 longer happens. I'm not sure if this situation could ever has previously
6039 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6042 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6043 header. This is something that is recommended in a new Internet Draft, and
6044 is something that is documented as being done by Sendmail. There are two
6045 possible values. For messages generated by the autoreply transport, Exim
6048 Auto-Submitted: auto-replied
6050 whereas for all other generated messages (e.g. bounces) it adds
6052 Auto-Submitted: auto-generated
6054 63. The "personal" condition in filters now includes a test for the
6055 Auto-Submitted: header. If it contains the string "auto-" the message it
6056 not considered personal.
6058 64. Added rcpt_include_affixes as a generic transport option.
6060 65. Added queue_only_override (default true).
6062 66. Added the syslog_duplication option.
6064 67. If what should have been the first header line of a message consisted of
6065 a space followed by a colon, Exim was mis-interpreting it as a header line.
6066 It isn't of course - it is syntactically invalid and should therefore be
6067 treated as the start of the message body. The misbehaviour could have
6068 caused a number of strange effects, including loss of data in subsequent
6069 header lines, and spool format errors.
6071 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6072 client host had authenticated. This control can now be exercised by an ACL
6073 for more flexibility.
6075 69. By default, callouts do not happen when testing with -bh. There is now a
6076 variant, -bhc, which does actually run the callout code, including
6077 consulting and updating the callout cache.
6079 70. Added support for saslauthd authentication, courtesy of Alexander
6082 71. If statvfs() failed on the spool or log directories while checking their
6083 size for availability, Exim confusingly gave the error "space shortage".
6084 Furthermore, in debugging mode it crashed with a floating point exception.
6085 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6086 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6087 serious problem, Exim now writes to the main and panic logs when this
6088 happens, with details of the failure. It then refuses to accept the
6089 incoming message, giving the message "spool directory problem" or "log
6090 directory problem" with a 421 code for SMTP messages.
6092 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6093 0, 1, and 2 exist, because some OS complain for execs without them (see
6094 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6095 descriptors. However, the code omitted to check that the open succeeded,
6096 causing mysterious errors if for some reason the permissions on /dev/null
6097 got screwed. Now Exim writes a message to the main and panic logs, and
6098 bombs out if it can't open /dev/null.
6100 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6101 interact so that it is all more flexible. It is supposed to remain
6102 backwards compatible. Also added extra_local_interfaces.
6104 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6105 to bomb out with an assertion failure - to the client this appears as a
6106 connection drop. This problem occurs in the part of the code that was taken
6107 from the Samba project. Fortunately, the assertion is in a very simple
6108 function, so I have fixed this by reproducing the function inline in the
6109 one place where it is called, and arranging for authentication to fail
6110 instead of killing the process with assert().
6112 75. The SPA client code was not working when the server requested OEM rather
6113 than Unicode encoding.
6115 76. Added code to make require_files with a specific uid setting more usable in
6116 the case where statting the file as root fails - usually a non-root-mounted
6117 NFS file system. When this happens and the failure is EACCES, Exim now
6118 forks a subprocess and does the per-uid checking as the relevant uid.
6120 77. Added process_log_path.
6122 78. If log_file_path was not explicitly set, a setting of check_log_space or
6123 check_log_inodes was ignored.
6125 79. If a space check for the spool or log partitions fails, the incident is now
6126 logged. Of course, in the latter case the data may get lost...
6128 80. Added the %p formatting code to string_format() so that it can be used to
6129 print addresses in debug_print(). Adjusted all the address printing in the
6130 debugging in store.c to use %p rather than %d.
6132 81. There was a concern that a line of code in smtp_in.c could overflow a
6133 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6134 initially expressed, the concern was not well-founded, because trailing
6135 spaces are removed early. However, if the trailing spaces were followed by
6136 a NULL, they did not get removed, so the overflow was possible. Two fixes
6139 (a) I re-wrote the offending code in a cleaner fashion.
6140 (b) If an incoming SMTP command contains a NULL character, it is rejected
6143 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6144 runs initgroups(), so that if the Exim user is in any additional groups,
6145 they will be used during message reception.
6151 The change log for 4.20 and earlier releases has been archived.