1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
15 TL/02 The BSD's have an arc4random API. One of the functions to induce
16 adding randomness was arc4random_stir(), but it has been removed in
17 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
18 function when detected.
20 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
21 cause callback expansion.
23 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
24 syntax errors in an expansion can be treated as a string instead of
25 logging or causing an error, due to the internal use of bool_lax
26 instead of bool when processing it.
28 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
29 server certificates when making smtp deliveries.
31 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
33 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
35 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
37 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
38 Merged patch from Sebastian Wiedenroth.
40 JH/05 Fix results-pipe from transport process. Several recipients, combined
41 with certificate use, exposed issues where response data items split
42 over buffer boundaries were not parsed properly. This eventually
43 resulted in duplicates being sent. This issue only became common enough
44 to notice due to the introduction of conection certificate information,
45 the item size being so much larger. Found and fixed by Wolfgang Breyha.
47 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
48 size buffer was used, resulting in syntax errors when an expansion
51 JH/07 Add support for directories of certificates when compiled with a GnuTLS
52 version 3.3.6 or later.
54 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
55 is EXPERIMENTAL_EVENT, the main-configuration and transport options
56 both become "event_action", the variables become $event_name, $event_data
57 and $event_defer_errno. There is a new variable $verify_mode, usable in
58 routers, transports and related events. The tls:cert event is now also
59 raised for inbound connections, if the main configuration event_action
62 TL/06 In test suite, disable OCSP for old versions of openssl which contained
63 early OCSP support, but no stapling (appears to be less than 1.0.0).
65 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
66 server certificate names available under the smtp transport option
67 "tls_verify_cert_hostname" now do not permit multi-component wildcard
73 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
74 checkers that were complaining about end of non-void function with no
77 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
78 This was a regression intruduced in 4.83 by another bugfix.
80 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
82 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
83 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
89 TF/01 Correctly close the server side of TLS when forking for delivery.
91 When a message was received over SMTP with TLS, Exim failed to clear up
92 the incoming connection properly after forking off the child process to
93 deliver the message. In some situations the subsequent outgoing
94 delivery connection happened to have the same fd number as the incoming
95 connection previously had. Exim would try to use TLS and fail, logging
96 a "Bad file descriptor" error.
98 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
99 utilities have not been installed.
101 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
102 temporary space as the ACL may create new global variables.
104 TL/01 LDAP support uses per connection or global context settings, depending
105 upon the detected version of the libraries at build time.
107 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
108 to extract and use the src ip:port in logging and expansions as if it
109 were a direct connection from the outside internet. PPv2 support was
110 updated based on HAProxy spec change in May 2014.
112 JH/02 Add ${listextract {number}{list}{success}{fail}}.
114 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
115 Properly escape header and check for NULL return.
117 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
120 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
122 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
123 characters in header names, implemented as a verify condition.
124 Contributed by Michael Fischer v. Mollard.
126 TL/05 Rename SPF condition results err_perm and err_temp to standardized
127 results permerror and temperror. Previous values are deprecated but
128 still accepted. In a future release, err_perm and err_temp will be
129 completely removed, which will be a backward incompatibility if the
130 ACL tests for either of these two old results. Patch contributed by
131 user bes-internal on the mailing list.
133 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
135 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
136 selectors, in both main and reject logs.
138 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
141 JH/07 Add malware type "sock" for talking to simple daemon.
143 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
145 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
146 routers/transports under cutthrough routing.
148 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
149 numbers. Touch up "bool" conditional to keep the same definition.
151 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
153 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
155 JH/12 Expand items in router/transport headers_add or headers_remove lists
156 individually rather than the list as a whole. Bug 1452.
158 Required for reasonable handling of multiple headers_ options when
159 they may be empty; requires that headers_remove items with embedded
160 colons must have them doubled (or the list-separator changed).
162 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
163 view the policy declared in the DMARC record. Currently, $dmarc_status
164 is a combined value of both the record presence and the result of the
167 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
169 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
170 dnslookup router and the smtp transport (applying to the forward
173 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
174 of ldap servers used for a specific lookup. Patch provided by Heiko
177 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
178 New variable $lookup_dnssec_authenticated for observability.
180 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
181 Patch submitted by Lars Timman.
183 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
185 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
186 Requires trusted mode and valid format message id, aborts otherwise.
187 Patch contributed by Heiko Schlichting.
189 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
190 certextract with support for various fields. Bug 1358.
192 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
193 is requested by default, modifiable by smtp transport option
196 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
197 operate on certificate variables to give certificate fingerprints
198 Also new ${sha256:cert_variable}.
200 JH/23 The PRDR feature is moved from being Experimental into the mainline.
202 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
203 Christian Aistleitner.
205 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
207 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
208 file. Patch from Wolfgang Breyha.
210 JH/25 Expand the coverage of the delivery $host and $host_address to
211 client authenticators run in verify callout. Bug 1476.
213 JH/26 Port service names are now accepted for tls_on_connect_ports, to
214 align with daemon_smtp_ports. Bug 72.
216 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
217 support and error reporting did not work properly.
219 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
220 and is readable. Patch from Andrew Colin Kissa.
222 TL/14 Enhance documentation of ${run expansion and how it parses the
223 commandline after expansion, particularly in the case when an
224 unquoted variable expansion results in an empty value.
226 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
228 PP/02 Fix internal collision of T_APL on systems which support RFC3123
229 by renaming away from it. Addresses GH issue 15, reported by
232 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
234 TL/15 SECURITY: prevent double expansion in math comparison functions
235 (can expand unsanitized data). Not remotely exploitable.
242 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
244 PP/02 Make -n do something, by making it not do something.
245 When combined with -bP, the name of an option is not output.
247 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
250 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
251 $sender_host_name and config options to manage this, and basic check
254 PP/05 DSCP support for outbound connections and control modifier for inbound.
256 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
257 (Only plugin which currently uses this is kerberos4, which nobody should
258 be using, but we should make it available and other future plugins might
259 conceivably use it, even though it would break NAT; stuff *should* be
260 using channel bindings instead).
262 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
263 name; added for Sendmail compatibility; requires admin caller.
264 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
265 just ignore it); requires trusted caller.
266 Also parse but ignore: -Ac -Am -X<logfile>
269 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
271 TL/02 Add +smtp_confirmation as a default logging option.
273 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
274 Patch by Magnus Holmgren from 2007-02-20.
276 TL/04 Bugzilla 1281 - Spec typo.
277 Bugzilla 1283 - Spec typo.
278 Bugzilla 1290 - Spec grammar fixes.
280 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
282 TL/06 Add Experimental DMARC support using libopendmarc libraries.
284 TL/07 Fix an out of order global option causing a segfault. Reported to dev
285 mailing list by by Dmitry Isaikin.
287 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
289 JH/02 Support "G" suffix to numbers in ${if comparisons.
291 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
293 NM/01 Bugzilla 1197 - Spec typo
294 Bugzilla 1196 - Spec examples corrections
296 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
298 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
299 gnutls_enable_pkcs11, but renamed to more accurately indicate its
302 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
303 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
305 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
306 "acl {{name}{arg}...}", and optional args on acl condition
309 JH/05 Permit multiple router/transport headers_add/remove lines.
311 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
313 JH/07 Avoid using a waiting database for a single-message-only transport.
314 Performance patch from Paul Fisher. Bugzilla 1262.
316 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
319 JH/09 Add $headers_added variable, with content from use of ACL modifier
320 add_header (but not yet added to the message). Bugzilla 199.
322 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
323 Pulled from Bugzilla 817 by Wolfgang Breyha.
325 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
327 (nb: this is the same fix as in Exim 4.80.1)
329 JH/11 Add A= logging on delivery lines, and a client_set_id option on
332 JH/12 Add optional authenticated_sender logging to A= and a log_selector
335 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
337 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
338 advertise SMTP AUTH mechanism to us, instead of a generic
339 protocol violation error. Also, make Exim more robust to bad
340 data from the Dovecot auth socket.
342 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
344 When a queue runner is handling a message, Exim first routes the
345 recipient addresses, during which it prunes them based on the retry
346 hints database. After that it attempts to deliver the message to
347 any remaining recipients. It then updates the hints database using
350 So if a recipient address works intermittently, it can get repeatedly
351 deferred at routing time. The retry hints record remains fresh so the
352 address never reaches the final cutoff time.
354 This is a fairly common occurrence when a user is bumping up against
355 their storage quota. Exim had some logic in its local delivery code
356 to deal with this. However it did not apply to per-recipient defers
357 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
359 This change adds a proper retry rule check during routing so that the
360 final cutoff time is checked against the message's age. We only do
361 this check if there is an address retry record and there is not a
362 domain retry record; this implies that previous attempts to handle
363 the address had the retry_use_local_parts option turned on. We use
364 this as an approximation for the destination being like a local
365 delivery, as in LMTP.
367 I suspect this new check makes the old local delivery cutoff check
368 redundant, but I have not verified this so I left the code in place.
370 TF/02 Correct gecos expansion when From: is a prefix of the username.
372 Test 0254 submits a message to Exim with the header
376 When I ran the test suite under the user fanf2, Exim expanded
377 the header to contain my full name, whereas it should have added
378 a Resent-Sender: header. It erroneously treats any prefix of the
379 username as equal to the username.
381 This change corrects that bug.
383 GF/01 DCC debug and logging tidyup
384 Error conditions log to paniclog rather than rejectlog.
385 Debug lines prefixed by "DCC: " to remove any ambiguity.
387 TF/03 Avoid unnecessary rebuilds of lookup-related code.
389 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
390 Bug spotted by Jeremy Harris; was flawed since initial commit.
391 Would have resulted in OCSP responses post-SNI triggering an Exim
392 NULL dereference and crash.
394 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
396 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
397 Bug detection, analysis and fix by Samuel Thibault.
398 Bugzilla 1331, Debian bug #698092.
400 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
402 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
403 Server implementation by Todd Lyons, client by JH.
404 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
405 config variable "prdr_enable" controls whether the server
406 advertises the facility. If the client requests PRDR a new
407 acl_data_smtp_prdr ACL is called once for each recipient, after
408 the body content is received and before the acl_smtp_data ACL.
409 The client is controlled by bolth of: a hosts_try_prdr option
410 on the smtp transport, and the server advertisement.
411 Default client logging of deliveries and rejections involving
412 PRDR are flagged with the string "PRDR".
414 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
415 fclose(). Diagnosis by Todd Lyons.
417 PP/17 Update configure.default to handle IPv6 localhost better.
418 Patch by Alain Williams (plus minor tweaks).
421 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
422 This is now consistent with GnuTLS, and is now documented: the
423 previous undocumented portable approach to treating the option as
424 unset was to force an expansion failure. That still works, and
425 an empty string is now equivalent.
427 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
428 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
429 not performing validation itself.
431 PP/20 Added force_command boolean option to pipe transport.
432 Patch from Nick Koston, of cPanel Inc.
434 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
437 TF/04 Added udpsend ACL modifer and hexquote expansion operator
439 PP/21 Fix eximon continuous updating with timestamped log-files.
440 Broken in a format-string cleanup in 4.80, missed when I repaired the
441 other false fix of the same issue.
442 Report and fix from Heiko Schlichting.
445 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
446 Report from Prashanth Katuri.
448 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
449 It's SecureTransport, so affects any MacOS clients which use the
450 system-integrated TLS libraries, including email clients.
452 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
453 using a MIME ACL for non-SMTP local injection.
454 Report and assistance in diagnosis by Warren Baker.
456 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
458 JH/16 Fix comparisons for 64b. Bugzilla 1385.
460 TL/09 Add expansion variable $authenticated_fail_id to keep track of
461 last id that failed so it may be referenced in subsequent ACL's.
463 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
466 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
467 ldap library initialization, allowing self-signed CA's to be
468 used. Also properly sets require_cert option later in code by
469 using NULL (global ldap config) instead of ldap handle (per
470 session). Bug diagnosis and testing by alxgomz.
472 TL/12 Enhanced documentation in the ratelimit.pl script provided in
473 the src/util/ subdirectory.
475 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
476 renamed to Transport Post Delivery Action by Jeremy Harris, as
479 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
480 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
481 redis_servers = needs to be configured which will be used by the redis
482 lookup. Patch from Warren Baker, of The Packet Hub.
484 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
486 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
487 hostname or reverse DNS when processing a host list. Used suggestions
488 from multiple comments on this bug.
490 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
492 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
493 Missed a few lines, added it to make the runtest require no keyboard
496 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
497 contains upper case chars. Make router use caseful_local_part.
499 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
500 support when GnuTLS has been built with p11-kit.
506 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
508 This, or similar/improved, will also be change PP/11 of 4.82.
514 PP/01 Handle short writes when writing local log-files.
515 In practice, only affects FreeBSD (8 onwards).
516 Bugzilla 1053, with thanks to Dmitry Isaikin.
518 NM/01 Bugzilla 949 - Documentation tweak
520 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
523 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
525 PP/02 Implemented gsasl authenticator.
527 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
529 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
530 `pkg-config foo` for cflags/libs.
532 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
533 with rest of GSASL and with heimdal_gssapi.
535 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
536 `pkg-config foo` for cflags/libs for the TLS implementation.
538 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
539 properties get this fed in as external SSF. A number of robustness
540 and debugging improvements to the cyrus_sasl authenticator.
542 PP/08 cyrus_sasl server now expands the server_realm option.
544 PP/09 Bugzilla 1214 - Log authentication information in reject log.
545 Patch by Jeremy Harris.
547 PP/10 Added dbmjz lookup type.
549 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
551 PP/12 MAIL args handles TAB as well as SP, for better interop with
552 non-compliant senders.
553 Analysis and variant patch by Todd Lyons.
555 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
556 Bug report from Lars Müller <lars@samba.org> (via SUSE),
557 Patch from Dirk Mueller <dmueller@suse.com>
559 PP/13 tls_peerdn now print-escaped for spool files.
560 Observed some $tls_peerdn in wild which contained \n, which resulted
561 in spool file corruption.
563 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
564 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
565 or write after TLS renegotiation, which otherwise led to messages
568 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
569 as a tracking header (ie: a signed header comes before the signature).
570 Patch from Wolfgang Breyha.
572 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
573 comma-sep list; embedded commas doubled.
575 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
577 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
579 Report and patch from Dmitry Banschikov.
581 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
582 Removed SSL_clear() after SSL_new() which led to protocol negotiation
583 failures. We appear to now support TLS1.1+ with Exim.
585 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
586 lets Exim select keys and certificates based upon TLS SNI from client.
587 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
588 before an outbound SMTP session. New log_selector, +tls_sni.
590 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
591 NULL dereference. Report and patch from Alun Jones.
593 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
594 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
595 Not seeing resolver debug output on NetBSD, but suspect this is a
596 resolver implementation change.
598 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
599 Left warnings. Added "eximon gdb" invocation mode.
601 PP/21 Defaulting "accept_8bitmime" to true, not false.
603 PP/22 Added -bw for inetd wait mode support.
605 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
606 locate the relevant includes and libraries. Made this the default.
608 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
609 Bugzilla 1246, report and most of solution from Tomasz Kusy.
611 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
612 This may cause build issues on older platforms.
614 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
615 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
616 gnutls_require_mac & gnutls_require_protocols (no longer supported).
617 Added SNI support via GnuTLS too.
618 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
620 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
622 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
623 Applied second patch from Janne, implementing suggestion to default
624 multiple-strings-in-record handling to match SPF spec.
626 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
628 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
629 read-only, out of scope).
630 Patch from Wolfgang Breyha, report from Stuart Northfield.
632 PP/29 Fix three issues highlighted by clang analyser static analysis.
633 Only crash-plausible issue would require the Cambridge-specific
634 iplookup router and a misconfiguration.
635 Report from Marcin Mirosław.
637 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
639 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
640 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
641 As part of this, removing so much warning spew let me fix some minor
642 real issues in debug logging.
644 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
645 assignment on my part. Fixed.
647 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
648 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
649 Janne Snabb (who went above and beyond: thank you).
651 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
652 string otherwise requires a connection and a bunch more work and it's
653 relatively easy to get wrong. Should also expose TLS library linkage
656 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
657 64-bit ${eval} (JH/03).
659 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
660 GNU libc to support some of the 64-bit stuff, should not lead to
661 conflicts. Defined before os.h is pulled in, so if a given platform
662 needs to override this, it can.
664 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
665 protection layer was required, which is not implemented.
666 Bugzilla 1254, patch from Wolfgang Breyha.
668 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
669 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
670 tls_dhparam take prime identifiers. Also unbreak combination of
671 OpenSSL+DH_params+TLSSNI.
673 PP/39 Disable SSLv2 by default in OpenSSL support.
679 PP/01 Solaris build fix for Oracle's LDAP libraries.
680 Bugzilla 1109, patch from Stephen Usher.
682 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
684 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
687 TF/02 Fix a couple more cases where we did not log the error message
688 when unlink() failed. See also change 4.74-TF/03.
690 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
691 lock up or crash if it happened to be inside a call to libc when it
692 got a SIGUSR1 from exiwhat.
694 The SIGUSR1 handler appends the current process status to the process
695 log which is later printed by exiwhat. It used to use the general
696 purpose logging code to do this, but several functions it calls are
697 not safe for signals.
699 The new output code in the SIGUSR1 handler is specific to the process
700 log, and simple enough that it's easy to inspect for signal safety.
701 Removing some special cases also simplifies the general logging code.
702 Removing the spurious timestamps from the process log simplifies
705 TF/04 Improved ratelimit ACL condition.
707 The /noupdate option has been deprecated in favour of /readonly which
708 has clearer semantics. The /leaky, /strict, and /readonly update modes
709 are mutually exclusive. The update mode is no longer included in the
710 database key; it just determines when the database is updated. (This
711 means that when you upgrde Exim will forget old rate measurements.)
713 Exim now checks that the per_* options are used with an update mode that
714 makes sense for the current ACL. For example, when Exim is processing a
715 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
716 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
717 must specify per_mail/readonly. If you omit the update mode it defaults to
718 /leaky where that makes sense (as before) or /readonly where required.
720 The /noupdate option is now undocumented but still supported for
721 backwards compatibility. It is equivalent to /readonly except that in
722 ACLs where /readonly is required you may specify /leaky/noupdate or
723 /strict/noupdate which are treated the same as /readonly.
725 A useful new feature is the /count= option. This is a generalization
726 of the per_byte option, so that you can measure the throughput of other
727 aggregate values. For example, the per_byte option is now equivalent
728 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
730 The per_rcpt option has been generalized using the /count= mechanism
731 (though it's more complicated than the per_byte equivalence). When it is
732 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
733 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
734 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
735 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
736 Note that using per_rcpt with a non-readonly update mode in more than
737 one ACL will cause the recipients to be double-counted. (The per_mail
738 and per_byte options don't have this problem.)
740 The handling of very low rates has changed slightly. If the computed rate
741 is less than the event's count (usually one) then this event is the first
742 after a long gap. In this case the rate is set to the same as this event's
743 count, so that the first message of a spam run is counted properly.
745 The major new feature is a mechanism for counting the rate of unique
746 events. The new per_addr option counts the number of different
747 recipients that someone has sent messages to in the last time period. It
748 behaves like per_rcpt if all the recipient addresses are different, but
749 duplicate recipient addresses do not increase the measured rate. Like
750 the /count= option this is a general mechanism, so the per_addr option
751 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
752 example, measure the rate that a client uses different sender addresses
753 with the options per_mail/unique=$sender_address. There are further
754 details in the main documentation.
756 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
758 TF/06 Removed a few PCRE remnants.
760 TF/07 Automatically extract Exim's version number from tags in the git
761 repository when doing development or release builds.
763 PP/02 Raise smtp_cmd_buffer_size to 16kB.
764 Bugzilla 879. Patch from Paul Fisher.
766 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
767 Heavily based on revision 40f9a89a from Simon Arlott's tree.
770 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
772 PP/05 Variable $av_failed, true if the AV scanner deferred.
773 Bugzilla 1078. Patch from John Horne.
775 PP/06 Stop make process more reliably on build failure.
776 Bugzilla 1087. Patch from Heiko Schlittermann.
778 PP/07 Make maildir_use_size_file an _expandable_ boolean.
779 Bugzilla 1089. Patch from Heiko Schlittermann.
781 PP/08 Handle ${run} returning more data than OS pipe buffer size.
782 Bugzilla 1131. Patch from Holger Weiß.
784 PP/09 Handle IPv6 addresses with SPF.
785 Bugzilla 860. Patch from Wolfgang Breyha.
787 PP/10 GnuTLS: support TLS 1.2 & 1.1.
789 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
792 PP/11 match_* no longer expand right-hand-side by default.
793 New compile-time build option, EXPAND_LISTMATCH_RHS.
794 New expansion conditions, "inlist", "inlisti".
796 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
798 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
800 PP/14 fix log_write() format string regression from TF/03.
801 Bugzilla 1152. Patch from Dmitry Isaikin.
807 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
809 PP/02 Harmonised TLS library version reporting; only show if debugging.
810 Layout now matches that introduced for other libraries in 4.74 PP/03.
812 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
814 PP/04 New "dns_use_edns0" global option.
816 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
819 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
820 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
822 TK/01 Updated PolarSSL code to 0.14.2.
823 Bugzilla 1097. Patch from Andreas Metzler.
825 PP/07 Catch divide-by-zero in ${eval:...}.
828 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
831 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
832 format-string attack -- SECURITY: remote arbitrary code execution.
834 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
835 time unintentionally subject to list matching rules, letting the header
836 cause arbitrary Exim lookups (of items which can occur in lists, *not*
837 arbitrary string expansion). This allowed for information disclosure.
839 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
840 INT_MIN/-1 -- value coerced to INT_MAX.
846 NM/01 Workround for PCRE version dependency in version reporting
849 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
850 This fixes portability to compilers other than gcc, notably
851 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
853 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
854 makefiles for portability to HP-UX and POSIX correctness.
856 PP/01 Permit LOOKUP_foo enabling on the make command-line.
857 Also via indented variable definition in the Makefile.
858 (Debugging by Oliver Heesakkers).
860 PP/02 Restore caching of spamd results with expanded spamd_address.
861 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
863 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
864 Improves build reliability. Fix from: Frank Elsner
866 NM/02 Fix wide character breakage in the rfc2047 coding
867 Fixes bug 1064. Patch from Andrey N. Oktyabrski
869 NM/03 Allow underscore in dnslist lookups
870 Fixes bug 1026. Patch from Graeme Fowler
872 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
873 Code patches from Adam Ciarcinski of NetBSD.
875 NM/04 Fixed exiqgrep to cope with mailq missing size issue
878 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
879 is logged, to avoid truncation. Patch from John Horne.
881 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
882 Patch from Jakob Hirsch.
884 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
885 SQL string expansion failure details.
886 Patch from Andrey Oktyabrski.
888 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
889 Patch from Simon Arlott.
891 PP/09 New lookups functionality failed to compile on old gcc which rejects
892 extern declarations in function scope.
893 Patch from Oliver Fleischmann
895 PP/10 Use sig_atomic_t for flags set from signal handlers.
896 Check getgroups() return and improve debugging.
897 Fixed developed for diagnosis in bug 927 (which turned out to be
900 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
901 Patch from Mark Zealey.
903 PP/12 Bugzilla 1056: Improved spamd server selection.
904 Patch from Mark Zealey.
906 PP/13 Bugzilla 1086: Deal with maildir quota file races.
907 Based on patch from Heiko Schlittermann.
909 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
910 Patch from Uwe Doering, sign-off by Michael Haardt.
912 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
913 variable declaration deep within a block. Bug and patch from
916 PP/15 lookups-Makefile IRIX compatibilty coercion.
918 PP/16 Make DISABLE_DKIM build knob functional.
920 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
921 Patch by Simon Arlott
923 TF/03 Fix valgrind.h portability to C89 compilers that do not support
924 variable argument macros. Our copy now differs from upstream.
930 TF/01 Failure to get a lock on a hints database can have serious
931 consequences so log it to the panic log.
933 TF/02 Log LMTP confirmation messages in the same way as SMTP,
934 controlled using the smtp_confirmation log selector.
936 TF/03 Include the error message when we fail to unlink a spool file.
938 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
939 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
940 for maintaining out-of-tree patches for some time.
942 PP/01 Bugzilla 139: Documentation and portability issues.
943 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
944 Handle per-OS dynamic-module compilation flags.
946 PP/02 Let /dev/null have normal permissions.
947 The 4.73 fixes were a little too stringent and complained about the
948 permissions on /dev/null. Exempt it from some checks.
949 Reported by Andreas M. Kirchwitz.
951 PP/03 Report version information for many libraries, including
952 Exim version information for dynamically loaded libraries. Created
953 version.h, now support a version extension string for distributors
954 who patch heavily. Dynamic module ABI change.
956 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
957 privilege escalation vulnerability whereby the Exim run-time user
958 can cause root to append content of the attacker's choosing to
961 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
964 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
965 If dropping privileges for untrusted macros, we disabled normal logging
966 on the basis that it would fail; for the Exim run-time user, this is not
967 the case, and it resulted in successful deliveries going unlogged.
968 Fixed. Reported by Andreas Metzler.
974 PP/01 Date: & Message-Id: revert to normally being appended to a message,
975 only prepend for the Resent-* case. Fixes regression introduced in
976 Exim 4.70 by NM/22 for Bugzilla 607.
978 PP/02 Include check_rfc2047_length in configure.default because we're seeing
979 increasing numbers of administrators be bitten by this.
981 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
983 PP/03 Bugzilla 994: added openssl_options main configuration option.
985 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
987 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
989 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
991 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
992 without a peer certificate, leading to a segfault because of an
993 assumption that peers always have certificates. Be a little more
994 paranoid. Problem reported by Martin Tscholak.
996 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
997 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
998 NB: ClamAV planning to remove STREAM in "middle of 2010".
999 CL also introduces -bmalware, various -d+acl logging additions and
1000 more caution in buffer sizes.
1002 PP/09 Implemented reverse_ip expansion operator.
1004 PP/10 Bugzilla 937: provide a "debug" ACL control.
1006 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1008 PP/12 Bugzilla 973: Implement --version.
1010 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1012 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1014 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1016 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1017 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1018 ignore trailing whitespace.
1020 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1022 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1025 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1026 Notification from Dr Andrew Aitchison.
1028 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1029 ExtendedDetectionInfo response format.
1030 Notification from John Horne.
1032 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1035 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1036 XSL and documented dependency on system catalogs, with examples of how
1039 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1042 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1043 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1044 configuration file which is writeable by the Exim user or group.
1046 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1047 of configuration files to cover files specified with the -C option if
1048 they are going to be used with root privileges, not just the default
1051 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1052 option (effectively making it always true).
1054 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1055 files to be used while preserving root privileges.
1057 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1058 that rogue child processes cannot use them.
1060 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1061 run-time user, instead of root.
1063 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1064 Exim run-time user without dropping privileges.
1066 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1067 result string, instead of calling string_vformat() twice with the same
1070 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1071 for other users. Others should always drop root privileges if they use
1072 -C on the command line, even for a whitelisted configure file.
1074 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1076 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1082 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1083 $data_path, and $header_path variables; fixed documentation bugs and
1086 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1087 exipick to access non-standard spools, including the "frozen" queue
1090 NM/01 Bugzilla 965: Support mysql stored procedures.
1091 Patch from Alain Williams
1093 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1095 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1096 Patch from Andreas Metzler
1098 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1099 Patch from Kirill Miazine
1101 NM/05 Bugzilla 671: Added umask to procmail example.
1103 JJ/03 installed exipick 20100323.0, fixing doc bug
1105 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1106 directory. Notification and patch from Dan Rosenberg.
1108 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1110 TK/02 Improve log output when DKIM signing operation fails.
1112 MH/01 Treat the transport option dkim_domain as a colon separated
1113 list, not as a single string, and sign the message with each element,
1114 omitting multiple occurences of the same signer.
1116 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1117 Bugzilla 985, 986. Patch by Simon Arlott
1119 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1120 Patch by Simon Arlott
1122 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1123 MBX locking. Notification from Dan Rosenberg.
1129 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1131 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1133 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1135 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1137 NM/04 Bugzilla 915: Flag broken perl installation during build.
1143 TK/01 Added patch by Johannes Berg that expands the main option
1144 "spamd_address" if it starts with a dollar sign.
1146 TK/02 Write list of recipients to X-Envelope-Sender header when building
1147 the mbox-format spool file for content scanning (suggested by Jakob
1150 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1151 (http://www.dcc-servers.net/) support via dccifd. Activated by
1152 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1154 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1155 by Mark Daniel Reidel <mr@df.eu>.
1157 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1158 When building exim an external PCRE library is now needed -
1159 PCRE is a system library on the majority of modern systems.
1160 See entry on PCRE_LIBS in EDITME file.
1162 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1163 conversation. Added nologin parameter to request.
1164 Patch contributed by Kirill Miazine.
1166 TF/01 Do not log submission mode rewrites if they do not change the address.
1168 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1170 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1171 log files in place. Contributed by Roberto Lima.
1173 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1175 TF/03 Bugzilla 615: When checking the local_parts router precondition
1176 after a local_part_suffix or local_part_prefix option, Exim now
1177 does not use the address's named list lookup cache, since this
1178 contains cached lookups for the whole local part.
1180 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1181 Robert Millan. Documentation is in experimental-spec.txt.
1183 TF/04 Bugzilla 668: Fix parallel build (make -j).
1185 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1187 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1188 Patch provided by Jan Srzednicki.
1190 TF/05 Leading white space used to be stripped from $spam_report which
1191 wrecked the formatting. Now it is preserved.
1193 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1194 that they are available at delivery time.
1196 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1198 TF/08 TLS error reporting now respects the incoming_interface and
1199 incoming_port log selectors.
1201 TF/09 Produce a more useful error message if an SMTP transport's hosts
1202 setting expands to an empty string.
1204 NM/06 Bugzilla 744: EXPN did not work under TLS.
1205 Patch provided by Phil Pennock.
1207 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1208 Patch provided by Richard Godbee.
1210 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1211 acl_smtp_notquit, added index entry.
1213 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1214 Patch provided by Eugene Bujak.
1216 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1217 accept(). Patch provided by Maxim Dounin.
1219 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1220 Patch provided by Phil Pennock.
1222 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1224 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1225 Patch provided by Brad "anomie" Jorsch.
1227 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1228 Patch provided by Dean Brooks.
1230 TK/05 Add native DKIM support (does not depend on external libraries).
1232 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1233 Patch provided by Graeme Fowler.
1235 NM/16 Bugzilla 851: Documentation example syntax fix.
1237 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1239 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1242 NM/19 Bugzilla 745: TLS version reporting.
1243 Patch provided by Phil Pennock.
1245 NM/20 Bugzilla 167: bool: condition support.
1246 Patch provided by Phil Pennock.
1248 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1249 clients. Patch provided by Phil Pennock.
1251 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1252 Patch provided by Brad "anomie" Jorsch.
1254 NM/23 Bugzilla 687: Fix misparses in eximstats.
1255 Patch provided by Heiko Schlittermann.
1257 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1258 Patch provided by Heiko Schlittermann.
1260 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1261 plus update to original patch.
1263 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1265 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1266 Patch provided by David Brownlee.
1268 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1270 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1272 NM/30 Bugzilla 888: TLS documentation bugfixes.
1274 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1276 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1277 Unlike the original bugzilla I have changed all shell scripts in src tree.
1279 NM/33 Bugzilla 898: Transport filter timeout fix.
1280 Patch by Todd Rinaldo.
1282 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1283 Patch by Serge Demonchaux.
1285 NM/35 Bugzilla 39: Base64 decode bug fixes.
1286 Patch by Jakob Hirsch.
1288 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1290 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1292 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1294 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1300 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1301 ALT-N's libdkim that I have put here:
1302 http://duncanthrax.net/exim-experimental/
1304 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1305 were called 'true' and it seems that C99 defines that as a reserved
1306 keyword to be used with 'bool' variable types. That means you could
1307 not include C99-style headers which use bools without triggering
1308 build errors in sieve.c.
1310 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1311 as mailq or other aliases. Changed the --help handling significantly
1312 to do whats expected. exim_usage() emits usage/help information.
1314 SC/01 Added the -bylocaldomain option to eximstats.
1316 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1318 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1320 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1326 PH/01 Another patch from the Sieve maintainer.
1328 PH/02 When an IPv6 address is converted to a string for single-key lookup
1329 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1330 dots are used instead of colons so that keys in lsearch files need not
1331 contain colons. This was done some time before quoting was made available
1332 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1333 (notated using the quote facility) so as to distinguish them from IPv4
1334 keys. This meant that lookups for IP addresses in host lists did not work
1335 for iplsearch lookups.
1337 This has been fixed by arranging for IPv6 addresses to be expressed with
1338 colons if the lookup type is iplsearch. This is not incompatible, because
1339 previously such lookups could never work.
1341 The situation is now rather anomolous, since one *can* have colons in
1342 ordinary lsearch keys. However, making the change in all cases is
1343 incompatible and would probably break a number of configurations.
1345 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1348 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1349 conversion specification without a maximum field width, thereby enabling
1350 a rogue spamd server to cause a buffer overflow. While nobody in their
1351 right mind would setup Exim to query an untrusted spamd server, an
1352 attacker that gains access to a server running spamd could potentially
1353 exploit this vulnerability to run arbitrary code as the Exim user.
1355 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1356 $primary_hostname instead of what libspf2 thinks the hosts name is.
1358 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1359 a directory entry by the name of the lookup key. Previously, if a
1360 symlink pointed to a non-existing file or a file in a directory that
1361 Exim lacked permissions to read, a lookup for a key matching that
1362 symlink would fail. Now it is enough that a matching directory entry
1363 exists, symlink or not. (Bugzilla 503.)
1365 PH/03 The body_linecount and body_zerocount variables are now exported in the
1368 PH/04 Added the $dnslist_matched variable.
1370 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1371 This means they are set thereafter only if the connection becomes
1374 PH/06 Added the client_condition to authenticators so that some can be skipped
1375 by clients under certain conditions.
1377 PH/07 The error message for a badly-placed control=no_multiline_responses left
1378 "_responses" off the end of the name.
1380 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1382 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1383 (without spaces) instead of just copying the configuration text.
1385 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1387 PH/11 Added $max_received_linelength.
1389 PH/12 Added +ignore_defer and +include_defer to host lists.
1391 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1392 way in which PCRE > 7.0 is built.
1394 PH/14 Implemented queue_only_load_latch.
1396 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1397 MAIL command. The effect was to mangle the value on 64-bit systems.
1399 PH/16 Another patch from the Sieve maintainer.
1401 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1403 PH/18 If a system quota error occurred while trying to create the file for
1404 a maildir delivery, the message "Mailbox is full" was not appended to the
1405 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1406 only to a quota excession during the actual writing of the file.
1408 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1409 characters?) which causes problems in log lines. The DN values are now
1410 passed through string_printing() before being added to log lines.
1412 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1413 and InterBase are left for another time.)
1415 PH/21 Added message_body_newlines option.
1417 PH/22 Guard against possible overflow in moan_check_errorcopy().
1419 PH/23 POSIX allows open() to be a macro; guard against that.
1421 PH/24 If the recipient of an error message contained an @ in the local part
1422 (suitably quoted, of course), incorrect values were put in $domain and
1423 $local_part during the evaluation of errors_copy.
1429 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1430 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1433 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1434 issue a MAIL command.
1436 PH/02 In an ACL statement such as
1438 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1440 if a client was not listed at all, or was listed with a value other than
1441 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1442 the condition was not true (as it should be), so access was not denied.
1443 The bug was that the ! inversion was incorrectly passed on to the second
1444 item. This has been fixed.
1446 PH/03 Added additional dnslists conditions == and =& which are different from
1447 = and & when the dns lookup returns more than one IP address.
1449 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1450 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1452 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1453 FSYNC, which compiles an option called disable_fsync that allows for
1454 bypassing fsync(). The documentation is heavily laced with warnings.
1456 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1458 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1459 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1460 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1461 including adding "make clean"; (3) Added -fPIC when compiling the test
1462 dynamically loaded module, to get rid of a warning.
1464 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1465 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1466 = 0s. The bug is otherwise harmless.
1468 PH/07 There was a bug in the dovecot authenticator such that the value of
1469 $auth1 could be overwritten, and so not correctly preserved, after a
1470 successful authentication. This usually meant that the value preserved by
1471 the server_setid option was incorrect.
1473 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1475 PH/09 Installed PCRE release 7.0.
1477 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1478 run for batched SMTP input. It is now run at the start of every message
1479 in the batch. While fixing this I discovered that the process information
1480 (output by running exiwhat) was not always getting set for -bs and -bS
1481 input. This is fixed, and it now also says "batched" for BSMTP.
1483 PH/11 Added control=no_pipelining.
1485 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1486 patch, slightly modified), and move the expansion of helo_data till after
1487 the connection is made in the smtp transport (so it can use these
1490 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1492 PH/14 Added log_selector = +pid.
1494 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1496 PH/16 Add ${if forany and ${if forall.
1498 PH/17 Added dsn_from option to vary the From: line in DSNs.
1500 PH/18 Flush SMTP output before performing a callout, unless control =
1501 no_callout_flush is set.
1503 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1504 was true (the default) a successful delivery failed to delete the retry
1505 item, thus causing premature timeout of the address. The bug is now
1508 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1510 PH/21 Long custom messages for fakedefer and fakereject are now split up
1511 into multiline reponses in the same way that messages for "deny" and
1512 other ACL rejections are.
1514 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1515 with slight modification.
1517 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1518 draft, changing the syntax and factoring some duplicate code".
1520 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1521 for deliveries of the second and subsequent messages over the same SMTP
1524 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1525 ${reduce, with only minor "tidies".
1527 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1529 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1530 expansion side effects.
1532 PH/27 When a message times out after an over-quota error from an Exim-imposed
1533 quota, the bounce message says "mailbox is full". This message was not
1534 being given when it was a system quota that was exceeded. It now should
1537 MH/03 Made $recipients available in local_scan(). local_scan() already has
1538 better access to the recipient list through recipients_list[], but
1539 $recipients can be useful in postmaster-provided expansion strings.
1541 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1542 in the case of a MAIL command with additional options following the
1543 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1544 were accidentally chopped off.
1546 PH/29 SMTP synchronization checks are implemented when a command is read -
1547 there is a check that no more input is waiting when there shouldn't be
1548 any. However, for some commands, a delay in an ACL can mean that it is
1549 some time before the response is written. In this time, more input might
1550 arrive, invalidly. So now there are extra checks after an ACL has run for
1551 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1552 pipelining has not been advertised.
1554 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1556 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1557 returned in the SMTP response when smtp_return_error_details was false.
1558 This has been fixed.
1560 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1561 instead of the C I/O that was originally supplied, because problems were
1562 reported on Solaris.
1564 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1565 Exim which did not show up earlier: it was assuming that a call to
1566 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1567 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1568 was a macro that became an assignment, so it seemed to work. This has
1569 changed to a proper function call with a void return, hence the compile
1570 error. Exim's code has been fixed.
1572 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1575 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1577 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1579 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1580 because it turns out to be too big or there is a timeout) there is no
1581 "Completed" line in the log. When some messages of this type were
1582 selected by exigrep, they were listed as "not completed". Others were
1583 picked up by some special patterns. I have improved the selection
1584 criteria to be more general.
1586 PH/38 The host_find_failed option in the manualroute router can now be set
1587 to "ignore", to completely ignore a host whose IP address cannot be
1588 found. If all hosts are ignored, the behaviour is controlled by the new
1589 host_all_ignored option.
1591 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1592 homing or because of multiple MX records with /mx) generated more than
1593 one IP address, and the following item turned out to be the local host,
1594 all the secondary addresses of the first item were incorrectly removed
1595 from the list, along with the local host and any following hosts (which
1596 is what is supposed to happen).
1598 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1599 whoever called Exim into the -H file. In the case of the daemon it was
1600 behaving confusingly. When first started, it used values for whoever
1601 started the daemon, but after a SIGHUP it used the Exim user (because it
1602 calls itself on a restart). I have changed the code so that it now always
1605 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1606 message are rejected with the same error (e.g. no authentication or bad
1607 sender address), and a DATA command is nevertheless sent (as can happen
1608 with PIPELINING or a stupid MUA), the error message that was given to the
1609 RCPT commands is included in the rejection of the DATA command. This is
1610 intended to be helpful for MUAs that show only the final error to their
1613 PH/42 Another patch from the Sieve maintainer.
1615 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1616 Eximstats - Fixed some broken HTML links and added missing column headers
1618 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1619 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1621 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1623 PH/43 Yet another patch from the Sieve maintainer.
1625 PH/44 I found a way to check for a TCP/IP connection going away before sending
1626 the response to the final '.' that terminates a message, but only in the
1627 case where the client has not sent further data following the '.'
1628 (unfortunately, this is allowed). However, in many cases there won't be
1629 any further data because there won't be any more messages to send. A call
1630 to select() can be used: if it shows that the input is "ready", there is
1631 either input waiting, or the socket has been closed. An attempt to read
1632 the next input character can distinguish the two cases. Previously, Exim
1633 would have sent an OK response which the client would never have see.
1634 This could lead to message repetition. This fix should cure that, at
1635 least in a lot of common cases.
1637 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1638 advertised in response to EHLO.
1644 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1645 fixed by 4.65/MH/01 (is this a record?) are fixed:
1647 (i) An empty string was always treated as zero by the numeric comparison
1648 operators. This behaviour has been restored.
1650 (ii) It is documented that the numeric comparison operators always treat
1651 their arguments as decimal numbers. This was broken in that numbers
1652 starting with 0 were being interpreted as octal.
1654 While fixing these problems I realized that there was another issue that
1655 hadn't been noticed. Values of message_size_limit (both the global option
1656 and the transport option) were treated as octal if they started with 0.
1657 The documentation was vague. These values are now always treated as
1658 decimal, and I will make that clear in the documentation.
1664 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1665 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1668 MH/01 Don't check that the operands of numeric comparison operators are
1669 integers when their expansion is in "skipping" mode (fixes bug
1670 introduced by 4.64-PH/07).
1672 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1673 child addresses, Exim now panics and dies. Previously, because the count
1674 is held in a short int, deliveries were likely to be lost. As such a
1675 large number of recipients for a single message is ridiculous
1676 (performance will be very, very poor), I have chosen to impose a limit
1677 rather than extend the field.
1683 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1684 leftover -K file (the existence of which was triggered by #402).
1685 While we were at it, introduced process PID as part of the -K
1686 filename. This should rule out race conditions when creating
1689 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1690 processing considerably. Previous code took too long for large mails,
1691 triggering a timeout which in turn triggers #401.
1693 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1694 in the DK code in transports.c. sendfile() is not really portable,
1695 hence the _LINUX specificness.
1697 TF/01 In the add_headers option to the mail command in an Exim filter,
1698 there was a bug that Exim would claim a syntax error in any
1699 header after the first one which had an odd number of characters
1702 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1703 callout verification, Exim cached a "reject" for the entire domain. This
1704 is correct for most verifications, but it is not correct for a recipient
1705 verification with use_sender or use_postmaster set, because in that case
1706 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1707 case of MAIL FROM:<> rejection from other early rejections (e.g.
1708 rejection of HELO). When verifying a recipient using a non-null MAIL
1709 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1710 Whatever the result of the callout, the value of the domain cache is
1711 left unchanged (for any other kind of callout, getting as far as trying
1712 RCPT means that the domain itself is ok).
1714 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1717 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1718 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1719 session, the daemon ignores ECONNECT errors and logs others; it now
1720 ignores EPIPE as well.
1722 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1723 (quoted-printable decoding).
1725 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1726 later the small subsequent patch to fix an introduced bug.
1728 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1730 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1732 PH/08 An error is now given if message_size_limit is specified negative.
1734 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1735 to be given (somewhat) arbitrary names.
1737 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1740 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1741 miscellaneous code fixes
1743 PH/10 Added the log_reject_target ACL modifier to specify where to log
1746 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1747 hostname. This is wrong, because it relates to the incoming message (and
1748 probably the interface on which it is arriving) and not to the outgoing
1749 callout (which could be using a different interface). This has been
1750 changed to use the value of the helo_data option from the smtp transport
1751 instead - this is what is used when a message is actually being sent. If
1752 there is no remote transport (possible with a router that sets up host
1753 addresses), $smtp_active_hostname is used.
1755 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1756 tweaks were necessary in order to get it to work (see also 21 below):
1757 (a) The code assumed that strncpy() returns a negative number on buffer
1758 overflow, which isn't the case. Replaced with Exim's string_format()
1760 (b) There were several signed/unsigned issues. I just did the minimum
1761 hacking in of casts. There is scope for a larger refactoring.
1762 (c) The code used strcasecmp() which is not a standard C function.
1763 Replaced with Exim's strcmpic() function.
1764 (d) The code set only $1; it now sets $auth1 as well.
1765 (e) A simple test gave the error "authentication client didn't specify
1766 service in request". It would seem that Dovecot has changed its
1767 interface. Fortunately there's a specification; I followed it and
1768 changed what the client sends and it appears to be working now.
1770 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1773 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1774 address A is aliased to B and C, where B exists and C does not. Without
1775 -v the output is "A verified" because verification stops after a
1776 successful redirection if more than one address is generated. However,
1777 with -v the child addresses are also verified. Exim was outputting "A
1778 failed to verify" and then showing the successful verification for C,
1779 with its parentage. It now outputs "B failed to verify", showing B's
1780 parentage before showing the successful verification of C.
1782 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1783 look up a TXT record in a specific list after matching in a combined
1786 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1787 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1788 they consult the DNS. I had assumed they would set it the way they
1789 wanted; and indeed my experiments on Linux seem to show that in some
1790 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1791 To be on the safe side, however, I have now made the interface to
1792 host_find_byname() similar to host_find_bydns(), with an argument
1793 containing the DNS resolver options. The host_find_byname() function now
1794 sets these options at its start, just as host_find_bydns() does. The smtp
1795 transport options dns_qualify_single and dns_search_parents are passed to
1796 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1797 of host_find_byname() use the default settings of RES_DEFNAMES
1798 (qualify_single) but not RES_DNSRCH (search_parents).
1800 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1801 spool_read_header() do less string testing, by means of a preliminary
1802 switch on the second character of optional "-foo" lines. (This is
1803 overdue, caused by the large number of possibilities that now exist.
1804 Originally there were few.) While I was there, I also converted the
1805 str(n)cmp tests so they don't re-test the leading "-" and the first
1806 character, in the hope this might squeeze out yet more improvement.
1808 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1809 flag allowing group syntax was set by the header_syntax check but not
1810 turned off, possible causing trouble later; (2) The flag was not being
1811 set at all for the header_verify test, causing "group"-style headers to
1812 be rejected. I have now set it in this case, and also caused header_
1813 verify to ignore an empty address taken from a group. While doing this, I
1814 came across some other cases where the code for allowing group syntax
1815 while scanning a header line wasn't quite right (mostly, not resetting
1816 the flag correctly in the right place). These bugs could have caused
1817 trouble for malformed header lines. I hope it is now all correct.
1819 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1820 with the "reply" argument non-NULL. The code, however (which originally
1821 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1822 but it didn't always do it. This confused somebody who was copying the
1823 code for some other use. I have removed all the tests.
1825 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1826 feature that was used to support insecure browsers during the U.S. crypto
1827 embargo. It requires special client support, and Exim is probably the
1828 only MTA that supported it -- and would never use it because real RSA is
1829 always available. This code has been removed, because it had the bad
1830 effect of slowing Exim down by computing (never used) parameters for the
1831 RSA_EXPORT functionality.
1833 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1834 authenticator to fail if there's a tab character in the incoming data
1835 (there should never be unless someone is messing about, as it's supposed
1836 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1837 if the connection is using TLS or if the remote IP is the same as the
1838 local IP, and the "valid-client-cert option" if a client certificate has
1841 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1842 authenticators. This can be used for authorization after authentication
1843 succeeds. (In the case of plaintext, it servers for both authentication
1846 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1847 if any retry times were supplied.
1849 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1850 connection, where there is no client IP address to check. In this
1851 situation, the verify now always succeeds.
1853 PH/25 Applied John Jetmore's -Mset patch.
1855 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1857 PH/27 In a string expansion for a processed (not raw) header when multiple
1858 headers of the same name were present, leading whitespace was being
1859 removed from all of them, but trailing whitespace was being removed only
1860 from the last one. Now trailing whitespace is removed from each header
1861 before concatenation. Completely empty headers in a concatenation (as
1862 before) are ignored.
1864 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1865 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1867 PH/29 [Removed. This was a change that I later backed out, and forgot to
1868 correct the ChangeLog entry (that I had efficiently created) before
1869 committing the later change.]
1871 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1872 address errors (4xx response to RCPT) over the same connection as other
1873 messages routed to the same hosts. Such deliveries are always "forced",
1874 so retry times are not inspected. This resulted in far too many retries
1875 for the affected addresses. The effect occurred only when there were more
1876 hosts than the hosts_max_try setting in the smtp transport when it had
1877 the 4xx errors. Those hosts that it had tried were not added to the list
1878 of hosts for which the message was waiting, so if all were tried, there
1879 was no problem. Two fixes have been applied:
1881 (i) If there are any address or message errors in an SMTP delivery, none
1882 of the hosts (tried or untried) are now added to the list of hosts
1883 for which the message is waiting, so the message should not be a
1884 candidate for sending over the same connection that was used for a
1885 successful delivery of some other message. This seems entirely
1886 reasonable: after all the message is NOT "waiting for some host".
1887 This is so "obvious" that I'm not sure why it wasn't done
1888 previously. Hope I haven't missed anything, but it can't do any
1889 harm, as the worst effect is to miss an optimization.
1891 (ii) If, despite (i), such a delivery is accidentally attempted, the
1892 routing retry time is respected, so at least it doesn't keep
1893 hammering the server.
1895 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1896 in ${readsocket because some servers need this prod.
1898 PH/32 Added some extra debug output when updating a wait-xxx database.
1900 PH/33 The hint "could be header name not terminated by colon", which has been
1901 given for certain expansion errors for a long time, was not being given
1902 for the ${if def:h_colon_omitted{... case.
1904 PH/34 The spec says: "With one important exception, whenever a domain list is
1905 being scanned, $domain contains the subject domain." There was at least
1906 one case where this was not true.
1908 PH/35 The error "getsockname() failed: connection reset by peer" was being
1909 written to the panic log as well as the main log, but it isn't really
1910 panic-worthy as it just means the connection died rather early on. I have
1911 removed the panic log writing for the ECONNRESET error when getsockname()
1914 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1915 runs only) independently of the message's sender address. This meant
1916 that, if the 4xx error was in fact related to the sender, a different
1917 message to the same recipient with a different sender could confuse
1918 things. In particualar, this can happen when sending to a greylisting
1919 server, but other circumstances could also provoke similar problems.
1920 I have changed the default so that the retry time for these errors is now
1921 based a combination of the sender and recipient addresses. This change
1922 can be overridden by setting address_retry_include_sender=false in the
1925 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1926 remote server are returned as part of bounce messages. This was not
1927 happening for LMTP over a pipe (the lmtp transport), but now it is the
1928 same for both kinds of LMTP.
1930 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1931 in header lines that were in fact CNAMEs. This is no longer the case.
1933 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1934 and queue runs started by the daemon processed all messages. This has
1935 been fixed so that -R and -S can now usefully be given with -q<time>.
1937 PH/40 Import PCRE release 6.7 (fixes some bugs).
1939 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1941 PH/42 Give an error if -q is specified more than once.
1943 PH/43 Renamed the variables $interface_address and $interface_port as
1944 $received_ip_address and $received_port, to make it clear that these
1945 values apply to message reception, and not to the outgoing interface when
1946 a message is delivered. (The old names remain recognized, of course.)
1948 PH/44 There was no timeout on the connect() call when using a Unix domain
1949 socket in the ${readsocket expansion. There now is.
1951 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1952 be meaningful with "accept".
1954 SC/01 Eximstats V1.43
1955 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1957 SC/02 Eximstats V1.44
1958 Use a glob alias rather than an array ref in the generated
1959 parser. This improves both readability and performance.
1961 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1962 Collect SpamAssassin and rejection statistics.
1963 Don't display local sender or destination tables unless
1964 there is data to show.
1965 Added average volumes into the top table text output.
1967 SC/04 Eximstats V1.46
1968 Collect data on the number of addresses (recipients)
1969 as well as the number of messages.
1971 SC/05 Eximstats V1.47
1972 Added 'Message too big' to the list of mail rejection
1973 reasons (thanks to Marco Gaiarin).
1975 SC/06 Eximstats V1.48
1976 Mainlog lines which have GMT offsets and are too short to
1977 have a flag are now skipped.
1979 SC/07 Eximstats V1.49 (Alain Williams)
1980 Added the -emptyok flag.
1982 SC/08 Eximstats V1.50
1983 Fixes for obtaining the IP address from reject messages.
1985 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1986 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1987 whitesspace changes from 4.64-PH/27
1989 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1992 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1993 are found, allow negative numbers in numeric criteria)
1995 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1997 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1998 to match changes made in 4.64-PH/43
2000 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2002 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2003 "domain" is actually an IP address, and force a failure. This locks out
2004 those revolvers/nameservers that support "A-for-A" lookups, in
2005 contravention of the specifications.
2007 PH/48 When a host name was looked up from an IP address, and the subsequent
2008 forward lookup of the name timed out, the host name was left in
2009 $sender_host_name, contrary to the specification.
2011 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2012 restricted to single-key lookups, Exim was not diagnosing an error if
2013 * or *@ was used with a query-style lookup.
2015 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2017 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2018 long ago, but noone interested enough thought of it. Let's just say that
2019 the "1.1" means that there are some new functions that weren't there at
2020 some point in the past.
2022 PH/51 Error processing for expansion failure of helo_data from an smtp
2023 transport during callout processing was broken.
2025 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2026 tested/used via the -bh/-bhc/-bs options.
2028 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2029 bug, fixed in subsequent PCRE releases).
2031 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2032 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2034 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2040 SC/01 Use a glob alias rather than an array ref in eximstats generated
2041 parser. This improves both readability and performance.
2043 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2044 Don't display local sender or destination tables in eximstats unless
2045 there is data to show.
2046 Added average volumes into the eximstats top table text output.
2048 SC/03 Collect data on the number of addresses (recipients) as well
2049 as the number of messages in eximstats.
2051 TF/01 Correct an error in the documentation for the redirect router. Exim
2052 does (usually) call initgroups() when daemonizing.
2054 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2055 with consistent privilege compared to when running as a daemon.
2057 TF/03 Note in the spec that $authenticated_id is not set for local
2058 submissions from trusted users.
2060 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2061 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2063 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2064 by adding some example configuration directives to the default
2065 configuration file. A little bit of work is required to uncomment the
2066 directives and define how usernames and passwords are checked, but
2067 there is now a framework to start from.
2069 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2070 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2071 without this. I don't know how relevant this is to other LDAP libraries.
2073 PH/02 Add the verb name to the "unknown ACL verb" error.
2075 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2077 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2079 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2080 directory not expanded when it should be if an expanded home directory
2081 was set for the address (which is overridden by the transport).
2083 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2086 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2087 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2088 because it is too late at that time, and has no effect.
2090 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2091 security issue with \' (bugzilla #107). I could not use the
2092 PQescapeStringConn() function, because it needs a PGconn value as one of
2095 PH/08 When testing addresses using -bt, indicate those final addresses that
2096 are duplicates that would not cause an additional delivery. At least one
2097 person was confused, thinking that -bt output corresponded to deliveries.
2098 (Suppressing duplicates isn't a good idea as you lose the information
2099 about possibly different redirections that led to the duplicates.)
2101 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2102 systems where poll() doesn't work, in particular OS X.
2104 PH/10 Added more information to debugging output for retry time not reached.
2106 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2107 operations in malware.c.
2109 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2112 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2113 syslog_duplication set false, log lines that would normally be written
2114 both the the main log and to the reject log were not written to syslog at
2117 PH/14 In the default configuration, change the use of "message" in ACL warn
2118 statements to "add_header".
2120 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2121 not followed by a command (e.g. "seen endif").
2123 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2124 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2127 PH/17 Added extra conditions to the default value of delay_warning_condition
2131 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2132 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2133 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2136 The Auto-Submitted: and various List- headers are standardised, whereas I
2137 don't think Precedence: ever was.
2139 PH/18 Refactored debugging code in route_finduser() to show more information,
2140 in particular, the error code if getpwnam() issues one.
2142 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2143 This is apparently needed in addition to the PH/07 change above to avoid
2144 any possible encoding problems.
2146 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2147 but not after initializing Perl.
2149 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2150 output them only if debugging. By default they are written stderr,
2151 apparently, which is not desirable.
2153 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2156 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2159 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2161 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2162 authenticated or an ident call has been made. Suppress the default
2163 values for $authenticated_id and $authenticated_sender (but permit -oMai
2164 and -oMas) when testing with -bh.
2166 PH/24 Re-jigged the order of the tests in the default configuration so that the
2167 tests for valid domains and recipients precede the DNS black list and CSA
2168 tests, on the grounds that those ones are more expensive.
2170 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2171 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2172 This bug exists in every version of Exim that I still have, right back to
2175 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2176 However, an attempt to turn on case-sensitivity in a regex key by
2177 including (?-i) didn't work because the subject string was already
2178 lowercased, and the effects were non-intuitive. It turns out that a
2179 one-line patch can be used to allow (?-i) to work as expected.
2185 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2186 other effects) broke the use of negated acl sub-conditions.
2188 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2191 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2192 "Deny" causes Exim to reject the incoming connection with a 554 error.
2193 Unfortunately, if there is a major crisis, such as a disk failure,
2194 tcp-wrappers gives "deny", whereas what one would like would be some
2195 kind of temporary error. A kludge has been added to help with this.
2196 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2197 554 error is used if errno is still zero or contains ENOENT (which occurs
2198 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2201 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2203 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2204 errors. Otherwise a message that provokes a temporary error (when other
2205 messages do not) can cause a whole host to time out.
2207 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2208 addresses were routed directly to files or pipes from a redirect router.
2209 File deliveries just didn't batch; pipe deliveries might have suffered
2212 PH/06 A failure to get a lock for a hints database would erroneously always say
2213 "Failed to get write lock", even when it was really a read lock.
2215 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2216 of 0600. This has been changed to use the value of the lockfile_mode
2217 option (which defaults to 0600).
2219 PH/08 Applied small patch from the Sieve maintainer.
2221 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2222 folder from quota calculations, a direct delivery into this folder messed
2223 up the contents of the maildirsize file. This was because the regex was
2224 used only to exclude .Trash (or whatever) when the size of the mailbox
2225 was calculated. There was no check that a delivery was happening into an
2226 excluded directory. This bug has been fixed by ignoring all quota
2227 processing for deliveries into excluded directories.
2229 PH/10 Added the maildirfolder_create_regex option to appendfile.
2235 PH/01 The code for finding all the local interface addresses on a FreeBSD
2236 system running IPv6 was broken. This may well have applied to all BSD
2237 systems, as well as to others that have similar system calls. The broken
2238 code found IPv4 interfaces correctly, but gave incorrect values for the
2239 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2240 that it would not match correctly against @[] and not recognize the IPv6
2243 PH/02 The ipliteral router was not recognizing addresses of the form user@
2244 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2246 PH/03 Added disable_ipv6.
2248 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2249 decoded form, because it is most often used to construct To: headers
2250 lines in autoreplies, and the decoded form may well be syntactically
2251 invalid. However, $reply_address has leading white space removed, and all
2252 newlines turned into spaces so that the autoreply transport does not
2255 PH/05 If group was specified without a user on a router, and no group or user
2256 was specified on a transport, the group from the router was ignored.
2258 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2259 for visible compile-time settings that can be used to change these
2260 numbers, for those that want even more. Backwards compatibility with old
2261 spool files has been maintained. However, going back to a previous Exim
2262 release will lost any variables that are in spool files.
2264 PH/07 Two small changes when running in the test harness: increase delay when
2265 passing a TCP/IP connection to a new process, in case the original
2266 process has to generate a bounce, and remove special handling of
2267 127.0.0.2 (sic), which is no longer necessary.
2269 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2270 be the same on different OS.
2272 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2275 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2276 whether --show-vars was specified or not
2278 JJ/02 exipick: Added support for new ACL variable spool format introduced
2281 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2282 syntactically invalid From: or Reply-to: line, and a filter used this to
2283 generate an autoreply, and therefore failed to obtain an address for the
2284 autoreply, Exim could try to deliver to a non-existent relative file
2285 name, causing unrelated and misleading errors. What now happens is that
2286 it logs this as a hard delivery error, but does not attempt to create a
2289 PH/11 The exinext utility has a -C option for testing purposes, but although
2290 the given file was scanned by exinext itself; it wasn't being passed on
2291 when Exim was called.
2293 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2294 an end-of-file indication when reading a command response.
2296 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2297 compiled. In many other places in Exim, IPv6 addresses are always
2298 recognized, so I have changed this. It also means that IPv4 domain
2299 literals of the form [IPV4:n.n.n.n] are now always recognized.
2301 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2302 used if the router is not running as root, for example, when verifying at
2303 ACL time, or when using -bh. The debugging output from this situation was
2304 non-existent - all you got was a failure to exec. I have made two
2307 (a) Failures to set uid/gid, the current directory, or a process leader
2308 in a subprocess such as that created by queryprogram now generate
2309 suitable debugging ouput when -d is set.
2311 (b) The queryprogram router detects when it is not running as root,
2312 outputs suitable debugging information if -d is set, and then runs
2313 the subprocess without attempting to change uid/gid.
2315 PH/15 Minor change to Makefile for building test_host (undocumented testing
2318 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2319 additional section of a DNS packet that returns MX or SRV records.
2320 Instead, it always explicitly searches for A/AAAA records. This avoids
2321 major problems that occur when a DNS server includes only records of one
2322 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2323 fixed another bug: if SRV records were looked up and the corresponding
2324 address records were *not* found in the additional section, the port
2325 values from the SRV records were lost.
2327 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2328 using the correct key (the original address) when searching the retry
2329 rules in order to find which one to use for generating the retry hint.
2331 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2332 adding the default one. Similarly, if it contains a Reply-To: header, the
2333 errors_reply_to option, if set, is not used.
2335 PH/19 When calculating a retry time, Exim used to measure the "time since
2336 failure" by looking at the "first failed" field in the retry record. Now
2337 it does not use this if it is later than than the arrival time of the
2338 message. Instead it uses the arrival time. This makes for better
2339 behaviour in cases where some deliveries succeed, thus re-setting the
2340 "first failed" field. An example is a quota failure for a huge message
2341 when small messages continue to be delivered. Without this change, the
2342 "time since failure" will always be short, possible causing more frequent
2343 delivery attempts for the huge message than are intended.
2344 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2346 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2347 $1, $2, $3) because the numerical variables can be reset during some
2348 expansion items (e.g. "match"), thereby losing the authentication data.
2350 PH/21 Make -bV show the size of off_t variables so that the test suite can
2351 decide whether to run tests for quotas > 2G.
2353 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2354 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2355 filecount value is greater than 2G or if a quota value is greater than 2G
2356 on a system where the size of off_t is not greater than 4, a panic error
2359 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2360 never match. The debug and -bh output now contains an explicit error
2361 message indicating a malformed IPv4 address or mask.
2363 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2364 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2365 PH/23 above applies.
2367 PH/25 Do not write to syslog when running in the test harness. The only
2368 occasion when this arises is a failure to open the main or panic logs
2369 (for which there is an explicit test).
2371 PH/26 Added the /no_tell option to "control=freeze".
2373 PH/27 If a host name lookup failed very early in a connection, for example, if
2374 the IP address matched host_lookup and the reverse lookup yielded a name
2375 that did not have a forward lookup, an error message of the form "no IP
2376 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2377 could be logged. Now it outputs the IP address instead of "NULL".
2379 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2380 allows the sender and the authenticated sender to be set when
2381 submitting a message from within Exim. Since child_open_exim() is
2382 documented for local_scan(), the new function should be too.
2384 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2385 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2386 results in an empty string is now treated as unset.
2388 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2390 PH/31 Added sender_verify_fail logging option.
2392 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2393 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2394 tidied the source and removed it altogether.
2396 PH/33 When a queue run was abandoned because the load average was too high, a
2397 log line was always written; now it is written only if the queue_run log
2398 selector is set. In addition, the log line for abandonment now contains
2399 information about the queue run such as the pid. This is always present
2400 in "start" and "stop" lines but was omitted from the "abandon" line.
2402 PH/34 Omit spaces between a header name and the colon in the error message that
2403 is given when verify = headers_syntax fails (if there are lots of them,
2404 the message gets confusing).
2406 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2407 names, as there are now some PTR records that contain slashes. This check
2408 is only to protect against broken name servers that fall over on strange
2409 characters, so the fact that it applies to all lookups doesn't matter.
2411 PH/36 Now that the new test suite is complete, we can remove some of the
2412 special code in Exim that was needed for the old test suite. For example,
2413 sorting DNS records because real resolvers return them in an arbitrary
2414 order. The new test suite's fake resolver always returns records in the
2417 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2418 bounces) except when queue_only is set, to avoid logging races between
2419 the different processes.
2421 PH/38 Panic-die if .include specifies a non-absolute path.
2423 PH/39 A tweak to the "H" retry rule from its user.
2425 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2426 a label. They prevented compilation on older perls.
2428 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2429 a warning to be raised on newish perls.
2431 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2432 on queue. Changes to match documented behaviour of showing count of
2433 messages matching specified criteria.
2435 PH/40 Changed the default ident timeout from 30s to 5s.
2437 PH/41 Added support for the use of login_cap features, on those BSD systems
2438 that have them, for controlling the resources used by pipe deliveries.
2440 PH/42 The content-scanning code uses fopen() to create files in which to put
2441 message data. Previously it was not paying any attention to the mode of
2442 the files. Exim runs with umask(0) because the rest of the code creates
2443 files with open(), and sets the required mode explicitly. Thus, these
2444 files were ending up world-writeable. This was not a big issue, because,
2445 being within the spool directory, they were not world-accessible. I have
2446 created a function called modefopen, which takes an additional mode
2447 argument. It sets umask(777), creates the file, chmods it to the required
2448 mode, then resets the umask. All the relevant calls to fopen() in the
2449 content scanning code have been changed to use this function.
2451 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2452 to 24 hours. This avoids potential overflow problems when processing G
2453 and H retry rules. I suspect nobody ever tinkers with this value.
2455 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2457 PH/45 When the plaintext authenticator is running as a client, the server's
2458 challenges are checked to ensure they are valid base64 strings. By
2459 default, the authentication attempt is cancelled if an invalid string is
2460 received. Setting client_ignore_invalid_base64 true ignores these errors.
2461 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2462 they are received. Thus, the responses can be made to depend on the
2463 challenges. If an invalid string is ignored, an empty string is placed in
2466 PH/46 Messages that are created by the autoreply transport now contains a
2467 References: header, in accordance with RFCs 2822 and 3834.
2469 PH/47 Added authenticated_sender_force to the smtp transport.
2471 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2473 PH/49 Installed latest patch from the Sieve maintainer.
2475 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2476 was also set, the appendfile transport was unnecessarily scanning a
2477 directory of message files (e.g. for maildir delivery) to find the count
2478 of files (along with the size), even though it did not need this
2479 information. It now does the scan only if it needs to find either the
2480 size of the count of files.
2482 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2484 PH/52 Two bugs concerned with error handling when the smtp transport is
2487 (i) Exim was not creating retry information for temporary errors given
2488 for individual recipients after the DATA command when the smtp transport
2489 was used in LMTP mode. This meant that they could be retried too
2490 frequently, and not timed out correctly.
2492 (ii) Exim was setting the flag that allows error details to be returned
2493 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2494 recipients that were returned after the DATA command.
2496 PH/53 This is related to PH/52, but is more general: for any failing address,
2497 when detailed error information was permitted to be returned to the
2498 sender, but the error was temporary, then after the final timeout, only
2499 "retry timeout exceeded" was returned. Now it returns the full error as
2500 well as "retry timeout exceeded".
2502 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2503 do this, and (what is worse) MTAs that accept it.
2505 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2506 will now be deprecated.
2508 PH/56 New os.c-cygwin from the Cygwin maintainer.
2510 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2511 formats (previously only available in exim formats via -bpr, -bpru,
2512 and -bpra. Now also available in native and exiqgrep formats)
2514 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2515 with very large, slow to parse queues
2517 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2519 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2521 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2522 responses to authentication challenges, though it was showing the
2523 challenges; (ii) I've removed the CR characters from the debug output for
2526 PH/58 Allow for the insertion of a newline as well as a space when a string
2527 is turned into more than one encoded-word during RFC 2047 encoding. The
2528 Sieve code now uses this.
2530 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2531 data_4xx, lost_connection, tls_required.
2533 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2534 message was being sent as an SMTP response.
2536 PH/61 Add -l and -k options to exicyclog.
2538 PH/62 When verifying, if an address was redirected to one new address, so that
2539 verification continued, and the new address failed or deferred after
2540 having set something in $address_data, the value of $address_data was not
2541 passed back to the ACL. This was different to the case when no
2542 redirection occurred. The value is now passed back in both cases.
2544 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2545 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2546 use login_cap.h, so on its own it isn't the distinguishing feature. The
2547 new name refers directly to the setclassresources() function.
2549 PH/65 Added configuration files for NetBSD3.
2551 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2553 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2554 is preferred over IPv4.
2556 PH/68 The bounce_return_message and bounce_return_body options were not being
2557 honoured for bounces generated during the reception of non-SMTP messages.
2558 In particular, this applied to messages rejected by the ACL. This bug has
2559 been fixed. However, if bounce_return_message is true and bounce_return_
2560 body is false, the headers that are returned for a non-SMTP message
2561 include only those that have been read before the error was detected.
2562 (In the case of an ACL rejection, they have all been read.)
2564 PH/69 The HTML version of the specification is now built in a directory called
2565 spec_html instead of spec.html, because the latter looks like a path with
2566 a MIME-type, and this confuses some software.
2568 PH/70 Catch two compiler warnings in sieve.c.
2570 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2571 function verify_get_ident() calls ip_connect() to connect a socket, but
2572 if the "connect()" function timed out, ip_connect() used to close the
2573 socket. However, verify_get_ident() also closes the socket later, and in
2574 between Exim writes to the log, which may get opened at this point. When
2575 the socket was closed in ip_connect(), the log could get the same file
2576 descriptor number as the socket. This naturally causes chaos. The fix is
2577 not to close the socket in ip_connect(); the socket should be closed by
2578 the function that creates it. There was only one place in the code where
2579 this was missing, in the iplookup router, which I don't think anybody now
2580 uses, but I've fixed it anyway.
2582 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2583 well as to direct DNS lookups. Otherwise the handling of names in host
2584 lists is inconsistent and therefore confusing.
2590 PH/01 Two changes to the default runtime configuration:
2592 (1) Move the checks for relay_from_hosts and authenticated clients from
2593 after to before the (commented out) DNS black list checks.
2595 (2) Add control=submission to the relay_from_hosts and authenticated
2596 clients checks, on the grounds that messages accepted by these
2597 statements are most likely to be submissions.
2599 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2601 (1) Generate an error if the third argument for the ${prvs expansion is
2604 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2607 (3) Reset the variables that are obtained from the first argument of
2608 ${prvscheck and used in the second argument before leaving the code,
2609 because their memory is reclaimed, so using them afterwards may do
2612 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2613 one (it's much easier than Tom thought :-).
2615 (5) Because of (4), we can now allow for the use of $prvscheck_result
2616 inside the third argument.
2618 PH/03 For some reason, the default setting of PATH when running a command from
2619 a pipe transport was just "/usr/bin". I have changed it to
2622 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2623 anything to be listed in the output from -bV.
2625 PH/05 When a filter generated an autoreply, the entire To: header line was
2626 quoted in the delivery log line, like this:
2628 => >A.N.Other <ano@some.domain> <original@ddress> ...
2630 This has been changed so that it extracts the operative address. There
2631 may be more than one such address. If so, they are comma-separated, like
2634 => >ano@some.domain,ona@other.domain <original@ddress> ...
2636 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2637 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2638 not being looked up in the rDNS to get a host name, Exim was showing the
2639 IP address twice in Received: lines, even though the IP addresses were
2640 identical. For example:
2642 Received: from [1.2.3.4] (helo=[1.2.3.4])
2644 However, if the real host name was known, it was omitting the HELO data
2645 if it matched the actual IP address. This has been tidied up so that it
2646 doesn't show the same IP address twice.
2648 PH/07 When both +timestamp and +memory debugging was on, the value given by
2649 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2650 called by the debug printing, thereby overwriting the timestamp buffer.
2651 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2653 PH/08 When the original message was included in an autoreply transport, it
2654 always said "this is a copy of the message, including all the headers",
2655 even if body_only or headers_only was set. It now gives an appropriate
2658 PH/09 Applied a patch from the Sieve maintainer which:
2660 o fixes some comments
2661 o adds the (disabled) notify extension core
2662 o adds some debug output for the result of if/elsif tests
2663 o points to the current vacation draft in the documentation
2664 and documents the missing references header update
2668 o fixes a bug in processing the envelope test (when testing
2669 multiple envelope elements, the last element determinted the
2672 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2673 Electronic Mail") by including:
2675 Auto-submitted: auto-generated
2677 in the messages that it generates (bounce messages and others, such as
2678 warnings). In the case of bounce messages for non-SMTP mesages, there was
2679 also a typo: it was using "Auto_submitted" (underscore instead of
2680 hyphen). Since every message generated by Exim is necessarily in response
2681 to another message, thes have all been changed to:
2683 Auto-Submitted: auto-replied
2685 in accordance with these statements in the RFC:
2687 The auto-replied keyword:
2689 - SHOULD be used on messages sent in direct response to another
2690 message by an automatic process,
2692 - MUST NOT be used on manually-generated messages,
2694 - MAY be used on Delivery Status Notifications (DSNs) and Message
2695 Disposition Notifications (MDNs),
2697 - MUST NOT be used on messages generated by automatic or periodic
2698 processes, except for messages which are automatic responses to
2701 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2702 to the default Received: header definition.
2704 PH/12 Added log selector acl_warn_skipped (default on).
2706 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2707 variables because (a) they are in the wrong storage pool and (b) even if
2708 they were copied, it wouldn't work properly because of the caching.
2710 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2711 checking when decoding. Apparently there are clients that generate
2712 overlong encoded strings. Why am I not surprised?
2714 PH/15 If the first argument of "${if match_address" was not empty, but did not
2715 contain an "@" character, Exim crashed. Now it writes a panic log message
2716 and treats the condition as false.
2718 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2720 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2721 extension "envelope-auth". The code is finished and in agreement with
2722 other implementations, but there is no documentation so far and in fact,
2723 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2724 not changing the active code.
2726 Print executed "if" and "elsif" statements when debugging is used. This
2727 helps a great deal to understand what a filter does.
2729 Document more things not specified clearly in RFC3028. I had all this
2730 sorted out, when out of a sudden new issues came to my mind. Oops."
2732 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2735 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2736 canonical form (as documented). However, after a host name lookup from
2737 the IP address, check_host() was doing a simple string comparison with
2738 addresses acquired from the DNS when checking that the found name did
2739 have the original IP as one of its addresses. Since any found IPv6
2740 addresses are likely to be in abbreviated form, the comparison could
2741 fail. Luckily, there already exists a function for doing the comparison
2742 by converting both addresses to binary, so now that is used instead of
2743 the text comparison.
2745 PH/20 There was another similar case to PH/19, when a complete host name was
2746 given in a host list; looking up its IP address could give an abbreviated
2747 form, whereas the current host's name might or might not be abbreviated.
2748 The same fix has been applied.
2754 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2755 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2758 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2759 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2761 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2763 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2764 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2765 Also, for "Auto-Submitted", treat anything other than "no" as
2766 non-personal, in accordance with RFC 3834. (Previously it treated
2767 anything starting "auto-" as non-personal.)
2769 TF/01 The control=submission/name=... option had a problem with syntax
2770 errors if the name included a slash character. The /name= option
2771 now slurps the rest of the string, so it can include any characters
2772 but it must come last in the list of options (after /sender_retain
2775 PH/05 Some modifications to the interface to the fake nameserver for the new
2783 TK/01 Added the "success_on_redirect" address verification option. See
2784 NewStuff for rationale and an example.
2786 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2788 PH/02 Patch to exigrep to allow it to work on syslog lines.
2790 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2791 fread() to skip over the body file's header line, because in Cygwin the
2792 header line is locked and is inaccessible.
2794 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2795 co-exist for some time) to make it clear that it is the Exim ID that is
2796 referenced, not the Message-ID: header line.
2798 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2799 string_format() function, because snprintf() does not exist on all
2802 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2803 ${stat: expansion item.
2805 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2806 protocol synchronization error", to keep the pedants happy.
2808 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2809 well as for IRIX systems, when gcc is being used. See the host.c source
2812 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2814 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2817 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2818 transport and to the smtp transport in LMTP mode.
2820 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2822 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2823 run a filter in a subprocess. This could lead to confusion in subsequent
2824 lookups in the parent process. There should also be a search_tidyup() at
2825 the end of the subprocess.
2827 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2828 only if the host matched helo_try_verify_hosts, which caused the
2829 verification to occur when the EHLO/HELO command was issued. The ACL just
2830 tested the remembered result. Now, if a previous verification attempt has
2831 not happened, "verify = helo" does it there and then.
2833 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2835 TK/03 Fix log output including CR from clamd.
2837 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2838 exist provoked a memory error which could cause a segfault.
2840 PH/15 Installed PCRE 6.2
2842 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2844 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2845 of the problem. Specifically, suggested +O2 rather than +O1 for the
2848 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2850 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2851 record, the port was not correctly logged when the outgoing_port log
2852 selector was set (it logged the transort's default port).
2854 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2855 fallback_hosts, and "hosts" in the smtp transport.
2857 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2858 host errors such as "Connection refused".
2860 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2861 authentication with radiusclient 0.4.9:
2863 - Error returned from rc_read_config was caught wrongly
2864 - Username/password not passed on to radius server due to wrong length.
2866 The presumption is that some radiusclient API changes for 4.51/PH/17
2867 were not taken care of correctly. The code is still untested by me (my
2868 Linux distribution still has 0.3.2 of radiusclient), but it was
2869 contributed by a Radius user.
2871 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2872 expanding the "port" option of the smtp transport.
2874 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2875 while reading a MIME header. Thanks to Tom Hughes for a patch.
2877 PH/24 Include config.h inside local_scan.h so that configuration settings are
2880 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2881 that in an ACL for RCPT (for example), you can examine exactly what was
2884 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2885 commands, but it was not correctly comparing the address with the actual
2886 client host address. Thus, it would show the EHLO address in Received:
2887 header lines when this was not necessary.
2889 PH/27 Added the % operator to ${eval:}.
2891 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2892 it should be ignoring failures (because with -C, for example, it has lost
2893 privilege). It wasn't ignoring creation failures other than "already
2896 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2899 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2900 because an input line was too long, either on its own, or by virtue of
2901 too many continuations, the temporary file was not being removed, and the
2902 return code was incorrect.
2904 PH/31 Missing "BOOL" in function definition in filtertest.c.
2906 PH/32 Applied Sieve patches from the maintainer.
2908 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2910 PH/33 Added "verify = not_blind".
2912 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2913 Local/Makefile (with some defaults set). These are used in built scripts
2914 such as exicyclog, but they have never been used in the exim_install
2915 script (though there are many overriding facilities there). I have
2916 arranged that the exim_install script now takes note of these two
2919 PH/35 Installed configuration files for Dragonfly.
2921 PH/36 When a locally submitted message by a trusted user did not contain a
2922 From: header, and the sender address was obtained from -f or from an SMTP
2923 MAIL command, and the trusted user did not use -F to supply a sender
2924 name, $originator_name was incorrectly used when constructing a From:
2925 header. Furthermore, $originator_name was used for submission mode
2926 messages from external hosts without From: headers in a similar way,
2927 which is clearly wrong.
2929 PH/37 Added control=suppress_local_fixups.
2931 PH/38 When log_selector = +received_sender was set, and the addition of the
2932 sender made the log line's construction buffer exactly full, or one byte
2933 less than full, an overflow happened when the terminating "\n" was
2936 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2937 when the result of a list match is failure because a DNS lookup failed.
2939 PH/40 RM_COMMAND is now used in the building process.
2941 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2942 the "build-* directories that it finds.
2944 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2945 address in a domain literal was a prefix of an interface address.
2947 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2948 when verifying a sender address, unless rewrite_headers is false.
2950 PH/44 Wrote a long comment about why errors_to addresses are verified as
2951 recipients, not senders.
2953 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2954 the ratelimit ACL was added.
2956 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2958 PH/46 Added extra information about PostgreSQL errors to the error string.
2960 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2961 avoiding the need to install special zones in a real server. This is
2962 backwards compatible; if it can't find the fake resolver, it drops back.
2963 Thus, both old and new test suites can be run.
2965 TF/02 Added util/ratelimit.pl
2967 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2968 clock is set back in time.
2970 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2971 Candler <B.Candler@pobox.com>.
2973 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2974 OK for addresses that are the result of redirections.
2976 PH/48 A number of further additions for the benefit of the new test suite,
2977 including a fake gethostbyname() that interfaces to the fake DNS resolver
2980 TF/06 The fix for widen_domains has also been applied to qualify_single and
2981 search_parents which are the other dnslookup options that can cause
2984 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2987 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2989 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2990 service" if it encountered an MX record with an empty target hostname.
2991 The message is now "an MX or SRV record indicated no SMTP service".
2993 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2994 if the DNS of the sending site is misconfigured. This is quite a
2995 common situation. This change restores the behaviour of treating a
2996 helo verification defer as a failure.
2998 PH/51 If self=fail was set on a router, the bounce message did not include the
2999 actual error message.
3005 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3007 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3008 command itself ended in error, the underlying message about the transport
3009 filter timeout was being overwritten with the pipe command error. Now the
3010 underlying error message should be appended to the second error message.
3012 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3014 PH/02 Reduce the amount of output that "make" produces by default. Full output
3015 can still be requested.
3017 PH/03 The warning log line about a condition test deferring for a "warn" verb
3018 was being output only once per connection, rather than after each
3019 occurrence (because it was using the same function as for successful
3020 "warn" verbs). This seems wrong, so I have changed it.
3022 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3023 it should not have, which might have caused a crash in the right
3024 circumstances, but probably never did.
3026 PH/04 Installed a modified version of Tony Finch's patch to make submission
3027 mode fix the return path as well as the Sender: header line, and to
3028 add a /name= option so that you can make the user's friendly name appear
3031 TF/03 Added the control = fakedefer ACL modifier.
3033 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3034 Mark Lowes for thorough testing.
3036 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3038 TK/03 Merged latest SRS patch from Miles Wilton.
3040 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3041 with the definition in sysexits.h (which is #included earlier).
3042 Fortunately, Exim does not actually use EX_OK. The code used to try to
3043 preserve the sysexits.h value, by assumimg that macro definitions were
3044 scanned for macro replacements. I have been disabused of this notion,
3045 so now the code just undefines EX_OK before #including unistd.h.
3047 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3048 in the smtp transport. When a block could not be written in a single
3049 write() function, the timeout was being re-applied to each part-write.
3050 This seems wrong - if the receiver was accepting one byte at a time it
3051 would take for ever. The timeout is now adjusted when this happens. It
3052 doesn't have to be particularly precise.
3054 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3055 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3057 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3058 without a domain if the check to <postmaster@domain> fails.
3060 SC/01 Eximstats: added -xls and the ability to specify output files
3061 (patch written by Frank Heydlauf).
3063 SC/02 Eximstats: use FileHandles for outputing results.
3065 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3067 SC/04 Eximstats: fixed display of large numbers with -nvr option
3069 SC/05 Eximstats: fixed merging of reports with empty tables.
3071 SC/06 Eximstats: added the -include_original_destination flag
3073 SC/07 Eximstats: removed tabs and trailing whitespace.
3075 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3077 TK/06 MBOX spool code: Add real "From " MBOX separator line
3078 so the .eml file is really in mbox format (even though
3079 most programs do not really care). Patch from Alex Miller.
3081 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3082 The latter is generated from $received_to and is only set if the
3083 message has one envelope recipient. SA can use these headers,
3084 obviously out-of-the-box. Patch from Alex Miller.
3086 PH/08 The ${def test on a variable was returning false if the variable's
3087 value was "0", contrary to what the specification has always said!
3088 The result should be true unless the variable is empty.
3090 PH/09 The syntax error of a character other than { following "${if
3091 def:variable_name" (after optional whitespace) was not being diagnosed.
3092 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3093 accidental colon was present, for example, could give incorrect results.
3095 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3096 result is used (not including appendfile, where other changes are about
3099 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3100 This involved changing a lot of size variables from int to off_t. It
3101 should work with maildirs and everything.
3103 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3104 spamd dying while we are connected to it.
3106 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3109 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3110 function for BDB 4.3.
3112 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3114 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3115 the sources for PCRE have been re-arranged and more files are now
3118 PH/15 The code I had for printing potentially long long variables in PH/11
3119 above was not the best (it lost precision). The length of off_t variables
3120 is now inspected at build time, and an appropriate printing format (%ld
3121 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3122 to be "long long int" or "long int". This is needed for the internal
3123 formatting function string_vformat().
3125 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3126 the configuration file to be ":syslog", then the script "guesses" where
3127 the logs files are, rather than using the compiled in default. In our
3128 case the guess is not the same as the compiled default, so the script
3129 suddenly stopped working when I started to use syslog. The patch checks
3130 to see if log_file_path is "". If so, it attempts to read it from exim
3131 with no configuration file to get the compiled in version, before it
3132 falls back to the previous guessing code."
3134 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3135 implementing BATV in an Exim configuration. See NewStuff for the gory
3138 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3139 Makefile that are specific to HP-UX.
3141 PH/18 If the "use_postmaster" option was set for a recipient callout together
3142 with the "random" option, the postmaster address was used as the MAIL
3143 FROM address for the random test, but not for the subsequent recipient
3144 test. It is now used for both.
3146 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3147 patch removes a few documentation additions to RFC 3028, because the
3148 latest draft now contains them. It adds the new en;ascii-case comparator
3149 and a new error check for 8bit text in MIME parts. Comparator and
3150 require names are now matched exactly. I enabled the subaddress
3151 extension, but it is not well tested yet (read: it works for me)."
3153 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3154 rework some of the code of TK/09 above to avoid the hardwired use of
3155 "%lld" and "long long". Replaced the call to snprintf() with a call to
3158 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3159 records point to non-existent hosts", "retry timeout exceeded", and
3160 "retry time not reached for any host after a long failure period".
3162 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3163 experimental DomainKeys support:
3165 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3166 (2) On an error such as an illegally used "control", the wrong name for
3167 the control was given.
3169 These problems did NOT occur unless DomainKeys support was compiled.
3171 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3173 PH/24 Added ${if match_ip condition.
3175 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3176 databases so that it will be absolutely obvious if a crash occurs in the
3177 DB library. This is a regular occurrence (often caused by mis-matched
3180 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3181 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3182 picked up on a user's system that detects such things. There doesn't seem
3183 to be a gcc warning option for this - only an attribute that has to be
3184 put on the function's prototype. It seems that in Fedora Core 4 they have
3185 set this on a number of new functions. No doubt there will be more in due
3188 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3189 specify a transport. However, if an address that was verified by such a
3190 router was the subject of a callout, Exim crashed because it tried to
3191 read the rcpt_include_affixes from the non-existent transport. Now it
3192 just assumes that the setting of that option is false. This bug was
3193 introduced by 4.51/PH/31.
3195 PH/28 Changed -d+all to exclude +memory, because that information is very
3196 rarely of interest, but it makes the output a lot bigger. People tend to
3197 do -d+all out of habit.
3199 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3200 code in os-type was giving problems when libc.so lives in lib64, like on
3203 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3204 aren't the modern standard, and it seems that some systems' include files
3205 don't always have them. Exim was already checking for some of the newer
3206 ones like T_AAAA, and defining it itself. I've added checks for all the
3207 record types that Exim uses.
3209 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3210 not automatically generating a new one, as it is supposed to. This
3211 prevented TLS from working. If the file did exist, but contained invalid
3212 data, a new version was generated, as expected. It was only the case of a
3213 non-existent file that was broken.
3215 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3216 with a change in libdomainkeys > 0.64.
3218 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3219 from DNS. If the selector record carries the flag, it now has
3220 precedence over the domain-wide flag.
3222 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3224 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3225 the use of a port name that isn't defined in /etc/services) occurred, the
3226 message was deferred as in a normal delivery, and thus remained on the
3227 spool, instead of being failed because of the mua_wrapper setting. This
3228 is now fixed, and I tidied up some of the mua_wrapper messages at the
3231 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3232 the messages in a hash of arrays rather than using individual hashes.
3233 This is a bit cleaner and results in dramatic memory savings, albeit
3234 at a slight CPU cost.
3236 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3237 as requested by Marc Sherman.
3239 SC/10 Eximstats: added histograms for user specified patterns as requested
3242 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3244 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3245 fopen() in the content-scanning modules that did not already have it.
3251 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3252 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3254 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3256 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3258 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3259 to test Sieve filters that use "vacation".
3261 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3262 that changes the way the GnuTLS parameters are stored in the cache file.
3263 The new format can be generated externally. For backward compatibility,
3264 if the data in the cache doesn't make sense, Exim assumes it has read an
3265 old-format file, and it generates new data and writes a new file. This
3266 means that you can't go back to an older release without removing the
3269 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3270 work if there are any delivery delays because "one_time" forces the
3271 parent to be marked "delivered", so its unseen clone is never tried
3272 again. For this reason, Exim now forbids the simultaneous setting of
3275 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3276 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3277 entry if you want to know the details. The fix, however, neglected to
3278 consider the case when local delivery batching is involved. The test for
3279 "previously delivered" was not happening when checking to see if an
3280 address could be batched with a previous (undelivered) one; under
3281 certain circumstances this could lead to multiple deliveries to the same
3284 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3285 in its include files, and this causes problems building Exim.
3287 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3288 header_syntax) but Exim was just ignoring anything given after a slash.
3289 In particular, this caused confusion with an attempt to use "verify =
3290 reverse_host_lookup/defer_ok". An error is now given when options are
3291 supplied for verify items that do not have them. (Maybe reverse_host_
3292 lookup should have a defer_ok option, but that's a different point.)
3294 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3295 defined by RFC 821) to 2048, because there were problems with some AUTH
3296 commands, and RFC 1869 says the size should be increased for extended
3297 SMTP commands that take arguments.
3299 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3302 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3303 "unknown" error; now it says that the functionality isn't in the binary.
3305 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3306 an address' error message when a string expansion fails (syntax or
3307 whatever). Otherwise the password may appear in the log. Following change
3308 PH/42 below, there is no longer a chance of it appearing in a bounce
3311 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3313 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3314 of its addresses was ever tried. (Bugzilla bug #2.)
3316 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3317 the result incorrectly in the debug output. (It correctly added a newline
3318 to what was transported.)
3320 TF/01 Added $received_time.
3322 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3323 commented out examples of how to interface to a virus scanner and to
3324 SpamAssassin. Also added commented examples of av_scanner and
3325 spamd_address settings.
3327 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3328 and controls are allowed in which ACLs. There were a couple of minor
3329 errors. Some of the entries in the conditions table (which is a table of
3330 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3331 negation of where the condition IS allowed.
3333 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3335 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3336 header file does not have a version number, so I've had to invent a new
3337 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3338 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3339 radiusclient), but it was contributed by a Radius user.
3341 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3342 files or directories, for OpenSSL.
3344 PH/19 When an Exim process that is running as root has to create an Exim log
3345 file, it does so in a subprocess that runs as exim:exim so as to get the
3346 ownership right at creation (otherwise, other Exim processes might see
3347 the file with the wrong ownership). There was no test for failure of this
3348 fork() call, which would lead to the process getting stuck as it waited
3349 for a non-existent subprocess. Forks do occasionally fail when resources
3350 run out. I reviewed all the other calls to fork(); they all seem to check
3353 PH/20 When checking for unexpected SMTP input at connect time (before writing
3354 the banner), Exim was not dealing correctly with a non-positive return
3355 from the read() function. If the client had disconnected by this time,
3356 the result was a log entry for a synchronization error with an empty
3357 string after "input=" when read() returned zero. If read() returned -1
3358 (an event I could not check), uninitialized data bytes were printed.
3359 There were reports of junk text (parts of files, etc) appearing after
3362 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3364 PH/22 Added support for macro redefinition, and (re)definition in between
3365 driver and ACL definitions.
3367 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3368 forgetting to use the resulting value; it was using the unexpanded value.
3370 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3371 hadn't been configured. The fix is from Juergen Kreileder, who
3372 understands it better than I do:
3374 "Here's what I see happening with three configured cyrus_sasl
3375 authenticators configured (plain, login, cram-md5):
3377 On startup auth_cyrus_sasl_init() gets called for each of these.
3378 This means three calls to sasl_listmech() without a specified mech_list.
3379 => SASL tests which mechs of all available mechs actually work
3380 => three warnings about OTP not working
3381 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3383 With the patch, sasl_listmech() also gets called three times. But now
3384 SASL's mech_list option is set to the server_mech specified in the the
3385 authenticator. Or in other words, the answer from sasl_listmech()
3386 gets limited to just the mech you're testing for (which is different
3388 => the return list contains just 'plain' or 'login', 'cram-md5' or
3389 nothing depending on the value of ob->server_mech.
3391 I've just tested the patch: Authentication still works fine,
3392 unavailable mechs specified in the exim configuration are still
3393 caught, and the auth.log warnings about OTP are gone."
3395 PH/25 When debugging is enabled, the contents of the command line are added
3396 to the debugging output, even when log_selector=+arguments is not
3399 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3400 answer is "GNU", and only if the return is "GNU/something" is the answer
3403 PH/27 $acl_verify_message is now set immediately after the failure of a
3404 verification in an ACL, and so is available in subsequent modifiers. In
3405 particular, the message can be preserved by coding like this:
3407 warn !verify = sender
3408 set acl_m0 = $acl_verify_message
3410 Previously, $acl_verify_message was set only while expanding "message"
3411 and "log_message" when a very denied access.
3413 PH/28 Modified OS/os.c-Linux with
3415 -#ifndef OS_LOAD_AVERAGE
3416 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3418 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3419 nomenclature these days.)
3421 PH/29 Installed patch from the Sieve maintainer that adds the options
3422 sieve_useraddress and sieve_subaddress to the redirect router.
3424 PH/30 In these circumstances:
3425 . Two addresses routed to the same list of hosts;
3426 . First host does not offer TLS;
3427 . First host accepts first address;
3428 . First host gives temporary error to second address;
3429 . Second host offers TLS and a TLS session is established;
3430 . Second host accepts second address.
3431 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3432 and peerdn, if requested) that were in fact used only for the second
3435 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3436 attention to any local part prefix or suffix that was matched by the
3437 router that accepted the address. It now behaves in the same way as it
3438 does for delivery: the affixes are removed from the local part unless
3439 rcpt_include_affixes is set on the transport.
3441 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3442 timeout during the DATA phase of an incoming message.
3444 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3445 applied a patch sanctioned by the Sieve maintainer.
3447 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3448 the uid or gid is negative. A case of a negative gid caused this to be
3449 noticed. The fix allows for either to be negative.
3451 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3452 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3455 PH/36 The change PH/12 above was broken. Fixed it.
3457 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3458 the grounds that routing the same address twice would always produce the
3459 same answer. This might have been true once, but it is certainly no
3460 longer true now. Routing a child address may depend on the previous
3461 routing that produced that child. Some complicated redirection strategies
3462 went wrong when messages had multiple recipients, and made Exim's
3463 behaviour dependent on the order in which the addresses were given.
3465 I have moved the duplicate checking until after the routing is complete.
3466 Exim scans the addresses that are assigned to local and remote
3467 transports, and removes any duplicates. This means that more work will be
3468 done, as duplicates will always all be routed, but duplicates are
3469 presumably rare, so I don't expect this is of any significance.
3471 For deliveries to pipes, files, and autoreplies, the duplicate checking
3472 still happens during the routing process, since they are not going to be
3475 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3476 It corrects a timeout issue with spamd. This is Ian's comment: "The
3477 background is that sometimes spamd either never reads data from a
3478 connection it has accepted, or it never writes response data. The exiscan
3479 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3480 blindly assumes that writes won't block so it may never time out."
3482 PH/39 Allow G after quota size as well as K and M.
3484 PH/40 The value set for $authenticated_id in an authenticator may not contain
3485 binary zeroes or newlines because the value is written to log lines and
3486 to spool files. There was no check on this. Now the value is run through
3487 the string_printing() function so that such characters are converted to
3488 printable escape sequences.
3490 PH/41 $message_linecount is a new variable that contains the total number of
3491 lines in the message. Compare $body_linecount, which is the count for the
3494 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3495 bounce and delay warning messages, except in certain special cases, which
3498 (a) An SMTP error message from a remote host;
3499 (b) A message specified in a :fail: redirection;
3500 (c) A message specified in a "fail" command in a system filter;
3501 (d) A message specified in a FAIL return from the queryprogram router;
3502 (e) A message specified by the cannot_route_message router option.
3504 In these cases only, Exim does include the error details in bounce and
3505 warning messages. There are also a few cases where bland messages such
3506 as "unrouteable address" or "local delivery error" are given.
3508 PH/43 $value is now also set for the "else" part of a ${run expansion.
3510 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3511 being worked on, but at least Exim now implements the latest version to
3514 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3515 process to complete was treated as a delivery failure, a timeout while
3516 writing the message to the pipe was logged, but erroneously treated as a
3517 successful delivery. Such timeouts include transport filter timeouts. For
3518 consistency with the overall process timeout, these timeouts are now
3519 treated as errors, giving rise to delivery failures by default. However,
3520 there is now a new Boolean option for the pipe transport called
3521 timeout_defer, which, if set TRUE, converts the failures into defers for
3522 both kinds of timeout. A transport filter timeout is now identified in
3525 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3526 systems where "make" and "gmake" are different, calling "gmake" at top
3527 level broke things. I've arranged for the value of $(MAKE) to be passed
3528 from the Makefile to this script so that it can call the same version of
3532 A note about Exim versions 4.44 and 4.50
3533 ----------------------------------------
3535 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3536 changes of various kinds. As a consequence, a big documentation update was
3537 needed. This delayed the release for rather longer than seemed good, especially
3538 in the light of a couple of (minor) security issues. Therefore, the changes
3539 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3540 release. So 4.44 and 4.50 are in effect two different branches that both start
3543 I have left the 4.50 change log unchanged; it contains all the changes since
3544 4.43. The change log for 4.44 is below; many of its items are identical to
3545 those for 4.50. This seems to be the most sensible way to preserve the
3546 historical information.
3552 1. Minor wording change to the doc/README.SIEVE file.
3554 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3555 computation of the current number of files was incorrect.
3557 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3558 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3561 4. Give more explanation in the error message when the command for a transport
3562 filter fails to execute.
3564 5. There are several places where Exim runs a non-Exim command in a
3565 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3566 was being done only for the command run by the queryprogram router. It is
3567 now done for all such subprocesses. The other cases are: ${run, transport
3568 filters, and the commands run by the lmtp and pipe transports.
3570 6. Added CONFIGURE_GROUP build-time option.
3572 7. Some older OS have a limit of 256 on the maximum number of file
3573 descriptors. Exim was using setrlimit() to set 1000 as a large value
3574 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3575 systems. I've change it so that if it can't get 1000, it tries for 256.
3577 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3578 was an oversight, and furthermore, ever since the addition of extra
3579 controls (e.g. 4.43/32), the checks on when to allow different forms of
3580 "control" were broken. There should now be diagnostics for all cases when a
3581 control that does not make sense is encountered.
3583 9. Added the /retain_sender option to "control=submission".
3585 10. $recipients is now available in the predata ACL (oversight).
3587 11. Tidy the search cache before the fork to do a delivery from a message
3588 received from the command line. Otherwise the child will trigger a lookup
3589 failure and thereby defer the delivery if it tries to use (for example) a
3590 cached ldap connection that the parent has called unbind on.
3592 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3593 of $address_data from the recipient verification was clobbered by the
3594 sender verification.
3596 13. The value of address_data from a sender verification is now available in
3597 $sender_address_data in subsequent conditions in the ACL statement.
3599 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3601 15. Added a new option "connect=<time>" to callout options, to set a different
3604 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3605 was its contents. (It was OK if the option was not defined at all.)
3607 17. A "Completed" log line is now written for messages that are removed from
3608 the spool by the -Mrm option.
3610 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3611 information about exactly what failed.
3613 19. Added -dd to debug only the daemon process.
3615 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3616 handles timeouts, both on the server side and network timeouts. Renamed the
3617 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3619 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3620 It is now set to "smtps".
3622 22. $host_address is now set to the target address during the checking of
3623 ignore_target_hosts.
3625 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3626 being passed; this would have caused $sender_host_name to have been used if
3627 matching the list had actually called for a host name (not very likely,
3628 since this list is usually IP addresses). A host name is now passed as
3631 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3632 code that specifies a non-restarting handler (typically sigaction() in
3633 modern systems) in an attempt to fix a rare and obscure crash bug.
3635 25. Narrowed the window for a race in the daemon that could cause it to ignore
3636 SIGCHLD signals. This is not a major problem, because they are used only to
3637 wake it up if nothing else does.
3639 26. A malformed maildirsize file could cause Exim to calculate negative values
3640 for the mailbox size or file count. Odd effects could occur as a result.
3641 The maildirsize information is now recalculated if the size or filecount
3644 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3645 support for a long time. Removed HAVE_SYS_VFS_H.
3647 28. Installed the latest version of exipick from John Jetmore.
3649 29. In an address list, if the pattern was not a regular expression, an empty
3650 subject address (from a bounce message) matched only if the pattern was an
3651 empty string. Non-empty patterns were not even tested. This was the wrong
3652 because it is perfectly reasonable to use an empty address as part of a
3653 database query. An empty address is now tested by patterns that are
3654 lookups. However, all the other forms of pattern expect the subject to
3655 contain a local part and a domain, and therefore, for them, an empty
3656 address still always fails if the pattern is not itself empty.
3658 30. Exim went into a mad DNS loop when attempting to do a callout where the
3659 host was specified on an smtp transport, and looking it up yielded more
3660 than one IP address.
3662 31. Re-factored the code for checking spool and log partition space into a
3663 function that finds that data and another that does the check. The former
3664 is then used to implement four new variables: $spool_space, $log_space,
3665 $spool_inodes, and $log_inodes.
3667 32. The RFC2047 encoding function was originally intended for short strings
3668 such as real names; it was not keeping to the 75-character limit for
3669 encoded words that the RFC imposes. It now respects the limit, and
3670 generates multiple encoded words if necessary. To be on the safe side, I
3671 have increased the buffer size for the ${rfc2047: expansion operator from
3674 33. It is now permitted to omit both strings after an "if" condition; if the
3675 condition is true, the result is "true". As before, when the second string
3676 is omitted, a false condition yields an empty string. This makes it less
3677 cumbersome to write custom ACL and router conditions.
3679 34. Failure to deliver a bounce message always caused it to be frozen, even if
3680 there was an errors_to setting on the router. The errors_to setting is now
3683 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3684 canonical form (fully expanded) before being placed in
3685 $sender_host_address.
3687 36. The table in the code that translates DNS record types into text (T_A to
3688 "A" for instance) was missing entries for NS and CNAME. It is just possible
3689 that this could have caused confusion if both these types were looked up
3690 for the same domain, because the text type is used as part of Exim's
3691 per-process caching. But the chance of anyone hitting this buglet seems
3694 37. The dnsdb lookup has been extended in a number of ways.
3696 (1) There is a new type, "zns", which walks up the domain tree until it
3697 finds some nameserver records. It should be used with care.
3699 (2) There is a new type, "mxh", which is like "mx" except that it returns
3700 just the host names, not the priorities.
3702 (3) It is now possible to give a list of domains (or IP addresses) to be
3703 looked up. The behaviour when one of the lookups defers can be
3704 controlled by a keyword.
3706 (4) It is now possible to specify the separator character for use when
3707 multiple records are returned.
3709 38. The dnslists ACL condition has been extended: it is now possible to supply
3710 a list of IP addresses and/or domains to be looked up in a particular DNS
3713 39. Added log_selector=+queue_time_overall.
3715 40. When running the queue in the test harness, wait just a tad after forking a
3716 delivery process, to get repeatability of debugging output.
3718 41. Include certificate and key file names in error message when GnuTLS fails
3719 to set them up, because the GnuTLS error message doesn't include the name
3720 of the failing file when there is a problem reading it.
3722 42. Allow both -bf and -bF in the same test run.
3724 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3726 44. The "Exiscan patch" is now merged into the mainline Exim source.
3728 45. Sometimes the final signoff response after QUIT could fail to get
3729 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3730 < 0 before doing a fflush(). This bug looks as though it goes back to the
3731 introduction of TLS in release 3.20, but "sometimes" must have been rare
3732 because the tests only now provoked it.
3734 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3735 (this can affect the format of dates).
3737 47. exim_tidydb, when checking for the continued existence of a message for
3738 which it has found a message-specific retry record, was not finding
3739 messages that were in split spool directories. Consequently, it was
3740 deleting retry records that should have stayed in existence.
3742 48. Steve fixed some bugs in eximstats.
3744 49. The SPA authentication driver was not abandoning authentication and moving
3745 on to the next authenticator when an expansion was forced to fail,
3746 contradicting the general specification for all authenticators. Instead it
3747 was generating a temporary error. It now behaves as specified.
3749 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3750 (the order specifies the preference for clients). The order is now AES256,
3751 AES128, 3DES, ARCFOUR128.
3753 51. Small patch to Sieve code - explicitly set From: when generating an
3756 52. Exim crashed if a remote delivery caused a very long error message to be
3757 recorded - for instance if somebody sent an entire SpamAssassin report back
3758 as a large number of 550 error lines. This bug was coincidentally fixed by
3759 increasing the size of one of Exim's internal buffers (big_buffer) that
3760 happened as part of the Exiscan merge. However, to be on the safe side, I
3761 have made the code more robust (and fixed the comments that describe what
3764 53. Now that there can be additional text after "Completed" in log lines (if
3765 the queue_time_overall log selector is set), a one-byte patch to exigrep
3766 was needed to allow it to recognize "Completed" as not the last thing in
3769 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3770 patch that reportedly fixes this has been added. I am not expert enough to
3771 create a test for it. This is what the patch creator wrote:
3773 "I found a little strange behaviour of ldap code when working with
3774 Windows 2003 AD Domain, where users was placed in more than one
3775 Organization Units. When I tried to give exim partial DN, the exit code
3776 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3777 But simultaneously result of request was absolutely normal ldap result,
3778 so I produce this patch..."
3780 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3781 REFERENCE, so I have modified the code to exclude the patch when that macro
3784 55. Some experimental protocols are using DNS PTR records for new purposes. The
3785 keys for these records are domain names, not reversed IP addresses. The
3786 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3787 leaves it alone. Component reversal etc. now happens only for IP addresses.
3790 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3792 57. Double the size of the debug message buffer (to 2048) so that more of very
3793 long debug lines gets shown.
3795 58. The exicyclog utility now does better if the number of log files to keep
3796 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3798 59. Two changes related to the smtp_active_hostname option:
3800 (1) $smtp_active_hostname is now available as a variable.
3801 (2) The default for smtp_banner uses $smtp_active_hostname instead
3802 of $primary_hostname.
3804 60. The host_aton() function is supposed to be passed a string that is known
3805 to be a valid IP address. However, in the case of IPv6 addresses, it was
3806 not checking this. This is a hostage to fortune. Exim now panics and dies
3807 if the condition is not met. A case was found where this could be provoked
3808 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3809 components; fortuitously, this particular loophole had already been fixed
3810 by change 4.50/55 above.
3812 If there are any other similar loopholes, the new check in host_aton()
3813 itself should stop them being exploited. The report I received stated that
3814 data on the command line could provoke the exploit when Exim was running as
3815 exim, but did not say which command line option was involved. All I could
3816 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3817 running as the user.
3820 61. There was a buffer overflow vulnerability in the SPA authentication code
3821 (which came originally from the Samba project). I have added a test to the
3822 spa_base64_to_bits() function which I hope fixes it.
3825 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3826 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3828 63. The daemon start-up calls getloadavg() while still root for those OS that
3829 need the first call to be done as root, but it missed one case: when
3830 deliver_queue_load_max is set with deliver_drop_privilege. This is
3831 necessary for the benefit of the queue runner, because there is no re-exec
3832 when deliver_drop_privilege is set.
3834 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3835 This has been fixed.
3837 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3838 was in use, was not putting the data itself into the right store pool;
3839 consequently, it could be overwritten for a subsequent message in the same
3840 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3843 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3845 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3846 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3847 the function were treating the return as a boolean value, which happened to
3848 work because 0=false and not-0=true, but is not correct code.
3850 68. The host_aton() function was not handling scoped IPv6 addresses (those
3851 with, for example, "%eth0" on the end) correctly.
3853 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3854 negated items (that is, ~something) in unsigned ints. Some compilers
3855 apparently mutter when there is no cast.
3857 70. If an address verification called from an ACL failed, and did not produce a
3858 user-specific message (i.e. there was only a "system" message), nothing was
3859 put in $acl_verify_message. In this situation, it now puts the system
3862 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3863 session; change 4.31/43 added the unwanted input to the log line - except
3864 that it did not do this in the start of session case. It now does.
3866 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3867 This is wrong and can cause the other end to generate a synchronization
3868 error if it is another Exim or anything else that does the synchronization
3869 check. A QUIT command is no longer sent after a timeout.
3871 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3872 during host lookups.
3874 74. The defer_ok option of callout verification was not working if it was used
3875 when verifying addresses in header lines, that is, for this case:
3877 verify = header_sender/callout=defer_ok
3879 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3880 those file descriptors could be used for SMTP connections. If anything
3881 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3882 could be sent to the SMTP client, causing chaos. The daemon now opens
3883 stdin, stdout, and stderr to /dev/null when it puts itself into the
3886 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3887 log by default. The user can override this with suitable Perl magic.
3889 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3890 the log message when discard triggers, was not working for the DATA ACL or
3891 for the non-SMTP ACL.
3893 78. Error message wording change in sieve.c.
3895 79. If smtp_accept_max_per_host was set, the number of connections could be
3896 restricted to fewer than expected, because the daemon was trying to set up
3897 a new connection before checking whether the processes handling previous
3898 connections had finished. The check for completed processes is now done
3899 earlier. On busy systems, this bug wouldn't be noticed because something
3900 else would have woken the daemon, and it would have reaped the completed
3903 80. If a message was submitted locally by a user whose login name contained one
3904 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3905 It caused a spool format error. I have fixed the spool reading code. A
3906 related problem was that the "from" clause in the Received: line became
3907 illegal because of the space(s). It is now covered by ${quote_local_part.
3909 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3912 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3914 83. Patch from Sieve maintainer to fix unterminated string problem in
3915 "vacation" handling.
3917 84. Some minor changes to the Linux configuration files to help with other
3918 OS variants using glibc.
3920 85. One more patch for Sieve to update vacation handling to latest spec.
3923 ----------------------------------------------------
3924 See the note above about the 4.44 and 4.50 releases.
3925 ----------------------------------------------------
3931 1. Change 4.43/35 introduced a bug that caused file counts to be
3932 incorrectly computed when quota_filecount was set in an appendfile
3935 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3936 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3939 3. Give more explanation in the error message when the command for a transport
3940 filter fails to execute.
3942 4. There are several places where Exim runs a non-Exim command in a
3943 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3944 was being done only for the command run by the queryprogram router. It is
3945 now done for all such subprocesses. The other cases are: ${run, transport
3946 filters, and the commands run by the lmtp and pipe transports.
3948 5. Some older OS have a limit of 256 on the maximum number of file
3949 descriptors. Exim was using setrlimit() to set 1000 as a large value
3950 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3951 systems. I've change it so that if it can't get 1000, it tries for 256.
3953 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3954 was an oversight, and furthermore, ever since the addition of extra
3955 controls (e.g. 4.43/32), the checks on when to allow different forms of
3956 "control" were broken. There should now be diagnostics for all cases when a
3957 control that does not make sense is encountered.
3959 7. $recipients is now available in the predata ACL (oversight).
3961 8. Tidy the search cache before the fork to do a delivery from a message
3962 received from the command line. Otherwise the child will trigger a lookup
3963 failure and thereby defer the delivery if it tries to use (for example) a
3964 cached ldap connection that the parent has called unbind on.
3966 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3967 of $address_data from the recipient verification was clobbered by the
3968 sender verification.
3970 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3971 was its contents. (It was OK if the option was not defined at all.)
3973 11. A "Completed" log line is now written for messages that are removed from
3974 the spool by the -Mrm option.
3976 12. $host_address is now set to the target address during the checking of
3977 ignore_target_hosts.
3979 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3980 being passed; this would have caused $sender_host_name to have been used if
3981 matching the list had actually called for a host name (not very likely,
3982 since this list is usually IP addresses). A host name is now passed as
3985 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3986 code that specifies a non-restarting handler (typically sigaction() in
3987 modern systems) in an attempt to fix a rare and obscure crash bug.
3989 15. Narrowed the window for a race in the daemon that could cause it to ignore
3990 SIGCHLD signals. This is not a major problem, because they are used only to
3991 wake it up if nothing else does.
3993 16. A malformed maildirsize file could cause Exim to calculate negative values
3994 for the mailbox size or file count. Odd effects could occur as a result.
3995 The maildirsize information is now recalculated if the size or filecount
3998 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3999 support for a long time. Removed HAVE_SYS_VFS_H.
4001 18. Updated exipick to current release from John Jetmore.
4003 19. Allow an empty sender to be matched against a lookup in an address list.
4004 Previously the only cases considered were a regular expression, or an
4007 20. Exim went into a mad DNS lookup loop when doing a callout where the
4008 host was specified on the transport, if the DNS lookup yielded more than
4011 21. The RFC2047 encoding function was originally intended for short strings
4012 such as real names; it was not keeping to the 75-character limit for
4013 encoded words that the RFC imposes. It now respects the limit, and
4014 generates multiple encoded words if necessary. To be on the safe side, I
4015 have increased the buffer size for the ${rfc2047: expansion operator from
4018 22. Failure to deliver a bounce message always caused it to be frozen, even if
4019 there was an errors_to setting on the router. The errors_to setting is now
4022 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4023 canonical form (fully expanded) before being placed in
4024 $sender_host_address.
4026 24. Updated eximstats to version 1.33
4028 25. Include certificate and key file names in error message when GnuTLS fails
4029 to set them up, because the GnuTLS error message doesn't include the name
4030 of the failing file when there is a problem reading it.
4032 26. Expand error message when OpenSSL has problems setting up cert/key files.
4035 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4036 (this can affect the format of dates).
4038 28. exim_tidydb, when checking for the continued existence of a message for
4039 which it has found a message-specific retry record, was not finding
4040 messages that were in split spool directories. Consequently, it was
4041 deleting retry records that should have stayed in existence.
4043 29. eximstats updated to version 1.35
4044 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4045 1.35 - bugfix such that pie charts by volume are generated correctly
4047 30. The SPA authentication driver was not abandoning authentication and moving
4048 on to the next authenticator when an expansion was forced to fail,
4049 contradicting the general specification for all authenticators. Instead it
4050 was generating a temporary error. It now behaves as specified.
4052 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4053 (the order specifies the preference for clients). The order is now AES256,
4054 AES128, 3DES, ARCFOUR128.
4056 31. Small patch to Sieve code - explicitly set From: when generating an
4059 32. Exim crashed if a remote delivery caused a very long error message to be
4060 recorded - for instance if somebody sent an entire SpamAssassin report back
4061 as a large number of 550 error lines. This bug was coincidentally fixed by
4062 increasing the size of one of Exim's internal buffers (big_buffer) that
4063 happened as part of the Exiscan merge. However, to be on the safe side, I
4064 have made the code more robust (and fixed the comments that describe what
4067 33. Some experimental protocols are using DNS PTR records for new purposes. The
4068 keys for these records are domain names, not reversed IP addresses. The
4069 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4070 leaves it alone. Component reversal etc. now happens only for IP addresses.
4073 34. The host_aton() function is supposed to be passed a string that is known
4074 to be a valid IP address. However, in the case of IPv6 addresses, it was
4075 not checking this. This is a hostage to fortune. Exim now panics and dies
4076 if the condition is not met. A case was found where this could be provoked
4077 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4078 components; fortuitously, this particular loophole had already been fixed
4079 by change 4.50/55 or 4.44/33 above.
4081 If there are any other similar loopholes, the new check in host_aton()
4082 itself should stop them being exploited. The report I received stated that
4083 data on the command line could provoke the exploit when Exim was running as
4084 exim, but did not say which command line option was involved. All I could
4085 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4086 running as the user.
4089 35. There was a buffer overflow vulnerability in the SPA authentication code
4090 (which came originally from the Samba project). I have added a test to the
4091 spa_base64_to_bits() function which I hope fixes it.
4094 36. The daemon start-up calls getloadavg() while still root for those OS that
4095 need the first call to be done as root, but it missed one case: when
4096 deliver_queue_load_max is set with deliver_drop_privilege. This is
4097 necessary for the benefit of the queue runner, because there is no re-exec
4098 when deliver_drop_privilege is set.
4100 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4101 was in use, was not putting the data itself into the right store pool;
4102 consequently, it could be overwritten for a subsequent message in the same
4103 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4106 38. Sometimes the final signoff response after QUIT could fail to get
4107 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4108 < 0 before doing a fflush(). This bug looks as though it goes back to the
4109 introduction of TLS in release 3.20, but "sometimes" must have been rare
4110 because the tests only now provoked it.
4116 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4117 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4118 Now it returns an integer. A number of calls were still expecting a T/F
4119 return. Fortuitously, in all cases, the tests worked in OK situations,
4120 which is the norm. However, things would have gone wrong on any write
4121 failures on the smtp file descriptor. This function is used when sending
4122 messages over SMTP and also when doing verify callouts.
4124 2. When Exim is called to do synchronous delivery of a locally submitted
4125 message (the -odf or -odi options), it no longer closes stderr before doing
4128 3. Implemented the mua_wrapper option.
4130 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4132 5. Implemented the functions header_remove(), header_testname(),
4133 header_add_at_position(), and receive_remove_recipient(), and exported them
4136 6. If an ACL "warn" statement specified the addition of headers, Exim already
4137 inserted X-ACL-Warn: at the start if there was no header name. However, it
4138 was not making this test for the second and subsequent header lines if
4139 there were newlines in the string. This meant that an invalid header could
4140 be inserted if Exim was badly configured.
4142 7. Allow an ACL "warn" statement to add header lines at the start or after all
4143 the Received: headers, as well as at the end.
4145 8. Added the rcpt_4xx retry error code.
4147 9. Added postmaster_mailfrom=xxx to callout verification option.
4149 10. Added mailfrom=xxxx to the callout verification option, for verify=
4152 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4153 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4155 12. Inserted some casts to stop certain compilers warning when using pointer
4156 differences as field lengths or precisions in printf-type calls (mostly
4157 affecting debugging statements).
4159 13. Added optional readline() support for -be (dynamically loaded).
4161 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4162 same clock tick as a message's arrival, so that its received time was the
4163 same as the "first fail" time on the retry record, and that message
4164 remained on the queue past the ultimate address timeout, every queue runner
4165 would try a delivery (because it was past the ultimate address timeout) but
4166 after another failure, the ultimate address timeout, which should have then
4167 bounced the address, did not kick in. This was a "< instead of <=" error;
4168 in most cases the first failure would have been in the next clock tick
4169 after the received time, and all would be well.
4171 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4172 being recognized when the domain list was tested by the match_domain
4173 condition in an expansion string.
4175 16. Added the ${str2b64: operator.
4177 17. Exim was always calling setrlimit() to set a large limit for the number of
4178 processes, without checking whether the existing limit was already
4179 adequate. (It did check for the limit on file descriptors.) Furthermore,
4180 errors from getrlimit() and setrlimit() were being ignored. Now they are
4181 logged to the main and panic logs, but Exim does carry on, to try to do its
4182 job under whatever limits there are.
4184 18. Imported PCRE 5.0.
4186 19. Trivial typo in log message " temporarily refused connection" (the leading
4189 20. If the log selector return_path_on_delivery was set and an address was
4190 redirected to /dev/null, the delivery process crashed because it assumed
4191 that a return path would always be set for a "successful" delivery. In this
4192 case, the whole delivery is bypassed as an optimization, and therefore no
4195 21. Internal re-arrangement: the function for sending a challenge and reading
4196 a response while authentication was assuming a zero-terminated challenge
4197 string. It's now changed to take a pointer and a length, to allow for
4198 binary data in such strings.
4200 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4202 23. Exim was not respecting finduser_retries when seeking the login of the
4203 uid under which it was called; it was always trying 10 times. (The default
4204 setting of finduser_retries is zero.) Also, it was sleeping after the final
4205 failure, which is pointless.
4207 24. Implemented tls_on_connect_ports.
4209 25. Implemented acl_smtp_predata.
4211 26. If the domain in control=submission is set empty, Exim assumes that the
4212 authenticated id is a complete email address when it generates From: or
4213 Sender: header lines.
4215 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4216 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4217 chown and chgrp in /bin and hostname in /usr/bin.
4219 28. Exim was keeping the "process log" file open after each use, just as it
4220 does for the main log. This opens the possibility of it remaining open for
4221 long periods when the USR1 signal hits a daemon. Occasional processlog
4222 errors were reported, that could have been caused by this. Anyway, it seems
4223 much more sensible not to leave this file open at all, so that is what now
4226 29. The long-running daemon process does not normally write to the log once it
4227 has entered its main loop, and it closes the log before doing so. This is
4228 so that log files can straightforwardly be renamed and moved. However,
4229 there are a couple of unusual error situations where the daemon does write
4230 log entries, and I had neglected to close the log afterwards.
4232 30. The text of an SMTP error response that was received during a remote
4233 delivery was being truncated at 512 bytes. This is too short for some of
4234 the long messages that one sometimes sees. I've increased the limit to
4237 31. It is now possible to make retry rules that apply only when a message has a
4238 specific sender, in particular, an empty sender.
4240 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4241 it possible to be selective about when SMTP synchronization is enforced.
4243 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4245 32. Implemented hosts_connection_nolog.
4247 33. Added an ACL for QUIT.
4249 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4252 35. Added mailbox_size and mailbox_filecount to appendfile.
4254 36. Added control = no_multiline_responses to ACLs.
4256 37. There was a bug in the logic of the code that waits for the clock to tick
4257 in the case where the clock went backwards by a substantial amount such
4258 that the microsecond fraction of "now" was more than the microsecond
4259 fraction of "then" (but the whole seconds number was less).
4261 38. Added support for the libradius Radius client library this is found on
4262 FreeBSD (previously only the radiusclient library was supported).
4268 1. When certain lookups returned multiple values in the form name=value, the
4269 quoting of the values was not always being done properly. Specifically:
4270 (a) If the value started with a double quote, but contained no whitespace,
4272 (b) If the value contained whitespace other than a space character (i.e.
4273 tabs or newlines or carriage returns) it was not quoted.
4274 This fix has been applied to the mysql and pgsql lookups by writing a
4275 separate quoting function and calling it from the lookup code. The fix
4276 should probably also be applied to nisplus, ibase and oracle lookups, but
4277 since I cannot test any of those, I have not disturbed their existing code.
4279 2. A hit in the callout cache for a specific address caused a log line with no
4280 reason for rejecting RCPT. Now it says "Previous (cached) callout
4281 verification failure".
4283 3. There was an off-by-one bug in the queryprogram router. An over-long
4284 return line was truncated at 256 instead of 255 characters, thereby
4285 overflowing its buffer with the terminating zero. As well as fixing this, I
4286 have increased the buffer size to 1024 (and made a note to document this).
4288 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4289 when Exim is waiting for an SMTP response from a remote server, Exim
4290 restarts its select() call on the socket, thereby resetting its timeout.
4291 This is not a problem when such interrupts are rare. Somebody set up a cron
4292 job to run exiwhat every 2 minutes, which is less than the normal select()
4293 timeout (5 or 10 minutes). This meant that the select() timeout never
4294 kicked in because it was always reset. I have fixed this by comparing the
4295 time when an interrupt arrives with the time at the start of the first call
4296 to select(). If more time than the timeout has elapsed, the interrupt is
4297 treated as a timeout.
4299 5. Some internal re-factoring in preparation for the addition of Sieve
4300 extensions (by MH). In particular, the "personal" test is moved to a
4301 separate function, and given an option for scanning Cc: and Bcc: (which is
4302 not set for Exim filters).
4304 6. When Exim created an email address using the login of the caller as the
4305 local part (e.g. when creating a From: or Sender: header line), it was not
4306 quoting the local part when it contained special characters such as @.
4308 7. Installed new OpenBSD configuration files.
4310 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4311 try to make them clearer.
4313 9. Callout options, other than the timeout value, were being ignored when
4314 verifying sender addresses in header lines. For example, when using
4316 verify = header_sender/callout=no_cache
4318 the cache was (incorrectly) being used.
4320 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4321 only the Cygwin environment.
4323 11. When return_path_on_delivery was set as a log selector, if different remote
4324 addresses in the same message used different return paths and parallel
4325 remote delivery occurred, the wrong values would sometimes be logged.
4326 (Whenever a remote delivery process finished, the return path value from
4327 the most recently started remote delivery process was logged.)
4329 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4330 lines when AUTH and/or TLS are in use. This is the "received protocol"
4331 field. Exim used to use "asmtp" for authenticated SMTP, without any
4332 indication (in the protocol name) for TLS use. Now it follows the RFC and
4333 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4334 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4335 names appear in log lines as well as in Received: header lines.
4337 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4338 extensions, and comparison tests, and to fix some bugs.
4340 14. Changes to the "personal" filter test:
4342 (1) The test was buggy in that it was just doing the equivalent of
4343 "contains" tests on header lines. For example, if a user's address was
4344 anne@some.where, the "personal" test would incorrectly be true for
4346 To: susanne@some.where
4348 This test is now done by extracting each address from the header in turn,
4349 and checking the entire address. Other tests that are part of "personal"
4350 are now done using regular expressions (for example, to check local parts
4351 of addresses in From: header lines).
4353 (2) The list of non-personal local parts in From: addresses has been
4354 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4355 taken from the Sieve specification recommendations.
4357 (3) If the message contains any header line starting with "List-" it is
4358 treated as non-personal.
4360 (4) The test for "circular" in the Subject: header line has been removed
4361 because it now seems ill-conceived.
4363 15. Minor typos in src/EDITME comments corrected.
4365 16. Installed latest exipick from John Jetmore.
4367 17. If headers_add on a router specified a text string that was too long for
4368 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4369 of string_sprintf() is now avoided.
4371 18. $message_body_size was not set (it was always zero) when running the DATA
4372 ACL and the local_scan() function.
4374 19. For the "mail" command in an Exim filter, no default was being set for
4375 the once_repeat time, causing a random time value to be used if "once" was
4376 specified. (If the value happened to be <= 0, no repeat happened.) The
4377 default is now 0s, meaning "never repeat". The "vacation" command was OK
4378 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4379 (I found it when inspecting the code).
4381 20. There is now an overall timeout for performing a callout verification. It
4382 defaults to 4 times the callout timeout, which applies to individual SMTP
4383 commands during the callout. The overall timeout applies when there is more
4384 than one host that can be tried. The timeout is checked before trying the
4385 next host. This prevents very long delays if there are a large number of
4386 hosts and all are timing out (e.g. when the network connections are timing
4387 out). The value of the overall timeout can be changed by specifying an
4388 additional sub-option for "callout", called "maxwait". For example:
4390 verify = sender/callout=5s,maxwait=20s
4392 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4393 to the end before writing, but this should make it even safer).
4395 22. Exim was forgetting that it had advertised PIPELINING for the second and
4396 subsequent messages on an SMTP connection. It was also not resetting its
4397 memory on STARTTLS and an internal HELO.
4399 23. When Exim logs an SMTP synchronization error within a session, it now
4400 records whether PIPELINING has been advertised or not.
4402 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4403 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4404 rather than long int.
4406 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4408 26. Added the never_mail option to autoreply.
4414 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4415 crash if the getsockname() call failed; this can happen if a connection is
4416 closed very soon after it is established. The problem was simply in the
4417 order in which certain operations were done, causing Exim to try to write
4418 to the SMTP stream before it had set up the file descriptor. The bug has
4419 been fixed by making things happen in the correct order.
4425 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4426 before the connection was closed, thus losing the rejection response.
4428 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4429 some early Solaris releases, but causes trouble in current releases where
4430 socklen_t is defined.
4432 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4435 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4438 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4439 that releasing the top of it at the end releases what was used for sub-
4440 expansions (unless the block got too big). However, discard this block if
4441 the first thing is a variable or header, so that we can use its block when
4442 it is dynamic (useful for very large $message_headers, for example).
4444 6. Lookups now cache *every* query, not just the most recent. A new, separate
4445 store pool is used for this. It can be recovered when all lookup caches are
4446 flushed. Lookups now release memory at the end of their result strings.
4447 This has involved some general refactoring of the lookup sources.
4449 7. Some code has been added to the store_xxx() functions to reduce the amount
4450 of flapping under certain conditions.
4452 8. log_incoming_interface used to affect only the <= reception log lines. Now
4453 it causes the local interface and port to be added to several more SMTP log
4454 lines, for example "SMTP connection from", and rejection lines.
4456 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4458 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4460 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4461 could be overwritten at the end of the current message (or the start of a
4462 new message if it was set in a HELO ACL). The value is now preserved for
4463 the duration of the SMTP connection.
4465 12. If a transport had a headers_rewrite setting, and a matching header line
4466 contained an unqualified address, that address was qualified, even if it
4467 did not match any rewriting rules. The underlying bug was that the values
4468 of the flags that permit the existence of unqualified sender and recipient
4469 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4470 non-local messages, and by -bnq for local messages) were not being
4471 preserved with the message after it was received.
4473 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4474 "next input=" as part of the text comprising the host identity instead of
4475 the correct text. The code was using the same buffer for two different
4476 strings. However, depending on which order the printing function evaluated
4477 its arguments, the bug did not always show up. Under Linux, for example, my
4478 test suite worked just fine.
4480 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4481 doesn't work with some older versions of Perl. It has been changed to "my",
4482 which in any case is probably the better facility to use.
4484 15. A really picky compiler found some instances of statements for creating
4485 error messages that either had too many or two few arguments for the format
4488 16. The size of the buffer for calls to the DNS resolver has been increased
4489 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4490 for addresses that have a lot of PTR records. This alleviates a problem; it
4491 does not fully solve it.
4493 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4494 buffer now truncates the list and logs the incident, which is the same
4495 action as happens when Exim is looking up a host name and its aliases.
4496 Previously in this situation something unpredictable would happen;
4497 sometimes it was "internal error: store_reset failed".
4499 18. If a server dropped the connection unexpectedly when an Exim client was
4500 using GnuTLS and trying to read a response, the client delivery process
4501 crashed while trying to generate an error log message.
4503 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4504 string, for example:
4506 warn message = H1: something\nH2: something
4508 the text was added as a single header line from Exim's point of view
4509 though it ended up OK in the delivered message. However, searching for the
4510 second and subsequent header lines using $h_h2: did not work. This has been
4511 fixed. Similarly, if a system filter added multiple headers in this way,
4512 the routers could not see them.
4514 20. Expanded the error message when iplsearch is called with an invalid key to
4515 suggest using net-iplsearch in a host list.
4517 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4518 ACLs are no longer actually imposed (and a message to that effect is
4521 22. If a "gecos" field in a passwd entry contained escaped characters, in
4522 particular, if it contained a \" sequence, Exim got it wrong when building
4523 a From: or a Sender: header from that name. A second bug also caused
4524 incorrect handling when an unquoted " was present following a character
4525 that needed quoting.
4527 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4528 was not being matched caselessly.
4530 24. Arranged for all hyphens in the exim.8 source to be escaped with
4533 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4534 properly. Recipient callout cache records were still being keyed to include
4535 the sender, even when use_sender was set false. This led to far more
4536 callouts that were necessary. The sender is no longer included in the key
4537 when use_sender is false.
4539 26. Added "control = submission" modifier to ACLs.
4541 27. Added the ${base62d: operator to decode base 62 numbers.
4543 28. dnsdb lookups can now access SRV records.
4545 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4546 the configuration file.
4548 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4549 (-v) mode. This makes the output for a verbose queue run more intelligible.
4551 31. Added a use_postmaster feature to recipient callouts.
4553 32. Added the $body_zerocount variable, containing the number of binary zero
4554 bytes in the message body.
4556 33. The time of last modification of the "new" subdirectory is now used as the
4557 "mailbox time last read" when there is a quota error for a maildir
4560 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4562 35. Added +ignore_unknown as a special item in host lists.
4564 36. Code for decoding IPv6 addresses in host lists is now included, even if
4565 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4566 address was recognized as an IP address, but was then not correctly decoded
4567 into binary, causing unexpected and incorrect effects when compared with
4574 1. Very minor rewording of debugging text in manualroute to say "list of
4575 hosts" instead of "hostlist".
4577 2. If verify=header_syntax was set, and a header line with an unqualified
4578 address (no domain) and a large number of spaces between the end of the
4579 name and the colon was received, the reception process suffered a buffer
4580 overflow, and (when I tested it) crashed. This was caused by some obsolete
4581 code that should have been removed. The fix is to remove it!
4583 3. When running in the test harness, delay a bit after writing a bounce
4584 message to get a bit more predictability in the log output.
4586 4. Added a call to search_tidyup() just before forking a reception process. In
4587 theory, someone could use a lookup in the expansion of smtp_accept_max_
4588 per_host which, without the tidyup, could leave open a database connection.
4590 5. Added the variables $recipient_data and $sender_data which get set from a
4591 lookup success in an ACL "recipients" or "senders" condition, or a router
4592 "senders" option, similar to $domain_data and $local_part_data.
4594 6. Moved the writing of debug_print from before to after the "senders" test
4597 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4598 problems for message scanning, either using a data ACL, or using
4599 local_scan() because the Received: header was not generated till after they
4600 were called (in order to set the time as the time of reception completion).
4601 I have revised the way this works. The header is now generated after the
4602 body is received, but before the ACL or local_scan() are called. After they
4603 are run, the timestamp in the header is updated.
4609 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4610 before starting a queue runner without re-exec. This happened only when
4611 deliver_drop_privilege was set or when the Exim user was set to root. The
4612 effect of the bug was that timeouts during subsequent deliveries caused
4613 crashes instead of being properly handled. The handler is now left at its
4614 default (and expected) setting.
4616 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4617 message, again when deliver_drop_privilege is set or Exim is run as root.
4618 The bug described in (1) was not present in this case, but the tidying up
4619 of the other signals was missing. I have made the two cases consistent.
4621 3. The ignore_target_hosts setting on a manualroute router was being ignored
4622 for hosts that were looked up using the /MX notation.
4624 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4627 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4628 operated on the sender address. After changing the $sender_address to <>
4629 for the sender address verify, Exim was re-instated it as the original
4630 (before rewriting) address, but remembering that it had rewritten it, so it
4631 wasn't rewriting it again. This bug also had the effect of breaking the
4632 sender address verification caching when the sender address was rewritten.
4634 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4635 This has been changed so that if the ip literal address matches
4636 ignore_target_hosts, the router declines.
4638 7. Added expansion conditions match_domain, match_address, and match_local_
4639 part (NOT match_host).
4641 8. The placeholder for the Received: header didn't have a length field set.
4643 9. Added code to Exim itself and to exim_lock to test for a specific race
4644 condition that could lead to file corruption when using MBX delivery. The
4645 issue is with the lockfile that is created in /tmp. If this file is removed
4646 after a process has opened it but before that process has acquired a lock,
4647 there is the potential for a second process to recreate the file and also
4648 acquire a lock. This could lead to two Exim processes writing to the file
4649 at the same time. The added code performs the same test as UW imapd; it
4650 checks after acquiring the lock that its file descriptor still refers to
4651 the same named file.
4653 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4654 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4657 11. Added the smtp_active_hostname option. If used, this will typically be made
4658 to depend on the incoming interface address. Because $interface_address is
4659 not set up until the daemon has forked a reception process, error responses
4660 that can happen earlier (such as "too many connections") no longer contain
4663 12. If an expansion in a condition on a "warn" statement fails because a lookup
4664 defers, the "warn" statement is abandoned, and the next ACL statement is
4665 processed. Previously this caused the whole ACL to be aborted.
4667 13. Added the iplsearch lookup type.
4669 14. Added ident_timeout as a log selector.
4671 15. Added tls_certificate_verified as a log selector.
4673 16. Added a global option tls_require_ciphers (compare the smtp transport
4674 option of the same name). This controls incoming TLS connections.
4676 17. I finally figured out how to make tls_require_ciphers do a similar thing
4677 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4678 before starting the TLS session.
4680 18. Tabs are now shown as \t in -bP output.
4682 19. If the log selector return_path_on_delivery was set, Exim crashed when
4683 bouncing a message because it had too many Received: header lines.
4685 20. If two routers both had headers_remove settings, and the first one included
4686 a superfluous trailing colon, the final name in the first list and the
4687 first name in the second list were incorrectly joined into one item (with a
4688 colon in the middle).
4694 1. Added -C and -D options to the exinext utility, mainly to make it easier
4695 to include in the automated testing, but these could be helpful when
4696 multiple configurations are in use.
4698 2. The exinext utility was not formatting the output nicely when there was
4699 an alternate port involved in the retry record key, nor when there was a
4700 message id as well (for retries that were specific to a specific message
4701 and a specific host). It was also confused by IPv6 addresses, because of
4702 the additional colons they contain. I have fixed the IPv4 problem, and
4703 patched it up to do a reasonable job for IPv6.
4705 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4706 delivery, the log line now contains "pipelined" if PIPELINING was used.
4708 4. An SMTP transport process used to panic and die if the bind() call to set
4709 an explicit outgoing interface failed. This has been changed; it is now
4710 treated in the same way as a connect() failure.
4712 5. A reference to $sender_host_name in the part of a conditional expansion
4713 that was being skipped was still causing a DNS lookup. This no longer
4716 6. The def: expansion condition was not recognizing references to header lines
4717 that used bh_ and bheader_.
4719 7. Added the _cache feature to named lists.
4721 8. The code for checking quota_filecount in the appendfile transport was
4722 allowing one more file than it should have been.
4724 9. For compatibility with Sendmail, the command line option
4732 and sets the incoming protocol and host name (for trusted callers). The
4733 host name and its colon can be omitted when only the protocol is to be set.
4734 Note the Exim already has two private options, -pd and -ps, that refer to
4735 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4736 "s", but I don't think that's a major issue.
4738 10. A number of refactoring changes to the code, none of which should affect
4741 (a) The number of logging options was getting close to filling up the
4742 32-bit word that was used as a bit map. I have split them into two classes:
4743 those that are passed in the argument to log_write(), and those that are
4744 only ever tested independently outside of that function. These are now in
4745 separate 32-bit words, so there is plenty of room for expansion again.
4746 There is no change in the user interface or the logging behaviour.
4748 (b) When building, for example, log lines, the code previously used a
4749 macro that called string_cat() twice, in order to add two strings. This is
4750 not really sufficiently general. Furthermore, there was one instance where
4751 it was actually wrong because one of the argument was used twice, and in
4752 one call a function was used. (As it happened, calling the function twice
4753 did not affect the overall behaviour.) The macro has been replaced by a
4754 function that can join an arbitrary number of extra strings onto a growing
4757 (c) The code for expansion conditions now uses a table and a binary chop
4758 instead of a serial search (which was left over from when there were very
4759 few conditions). Also, it now recognizes conditions like "pam" even when
4760 the relevant support is not compiled in: a suitably worded error message is
4761 given if an attempt is made to use such a condition.
4763 11. Added ${time_interval:xxxxx}.
4765 12. A bug was causing one of the ddress fields not to be passed back correctly
4766 from remote delivery subprocesses. The field in question was not being
4767 subsequently used, so this caused to problems in practice.
4769 13. Added new log selectors queue_time and deliver_time.
4771 14. Might have fixed a bug in maildirsizefile handling that threw up
4772 "unexpected character" debug warnings, and recalculated the data
4773 unnecessarily. In any case, I expanded the warning message to give more
4776 15. Added the message "Restricted characters in address" to the statements in
4777 the default ACL that block characters like @ and % in local parts.
4779 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4780 Three changes have been made:
4782 (a) There was a serious bug; a negative response to MAIL caused the whole
4783 recipient domain to be cached as invalid, thereby blocking all messages
4784 to all local parts at the same domain, from all senders. This bug has
4785 been fixed. The domain is no longer cached after a negative response to
4786 MAIL if the sender used is not empty.
4788 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4791 (c) A new callout option, "use_sender" has been added for people who want
4792 the modified behaviour.
4798 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4801 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4802 indeed breaks things for older releases.
4804 3. Added additional logging to the case where there is a problem reading data
4805 from a filter that is running in a subprocess using a pipe, in order to
4806 try to track down a specific problem.
4808 4. Testing facility fudge: when running in the test harness and attempting
4809 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4810 getting "No route to host". Convert this to a timeout.
4812 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4815 6. Some OS don't have socklen_t but use size_t instead. This affects the
4816 fifth argument of getsockopt() amongst other things. This is now
4817 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4818 can be set for individual OS. I have set it for SunOS5, OSF1, and
4819 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4820 some earlier ones do not.
4822 7. Change 4.30/15 was not doing the test caselessly.
4824 8. The standard form for an IPv6 address literal was being rejected by address
4825 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4826 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4827 this, as well as the form without the "IPv6" on the front (but only when
4828 address literals are enabled, of course).
4830 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4832 10. Exim crashed if a message with an empty sender address specified by -f
4833 encountered a router with an errors_to setting. This could be provoked only
4834 by a command such as
4838 where an empty string was supplied; "<>" did not hit this bug.
4840 11. Installed PCRE release 4.5.
4842 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4843 remained set. It is now erased.
4845 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4846 times from message ids (which are base 36 rather than the normal 62).
4848 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4849 were being counted as actual protocol errors, and logged if the log
4850 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4851 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4852 and DATA following a set of rejected RCPTs do not count as protocol errors.
4853 In other words, Exim assumes they were pipelined, though this may not
4854 actually be the case. Of course, in all cases the client gets an
4855 appropriate error code.
4857 15. If a lookup fails in an ACL condition, a message about the failure may
4858 be available; it is used if testing the ACL cannot continue, because most
4859 such messages specify what the cause of the deferral is. However, some
4860 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4861 that caused an old message to be retained and used if a later statement
4862 caused a defer, replacing the real cause of the deferral.
4864 16. If an IP address had so many PTR records that the DNS lookup buffer
4865 was not large enough to hold them, Exim could crash while trying to process
4866 the truncated data. It now detects and logs this case.
4868 17. Further to 4.21/58, another change has been made: if (and only if) the
4869 first line of a message (the first header line) ends with CRLF, a bare LF
4870 in a subsequent header line has a space inserted after it, so as not to
4871 terminate the header.
4873 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4874 unnecessarily, used atoi() instead of strtol(), and didn't check the
4875 termination when getting file sizes from file names by regex.
4877 19. Completely re-implemented the support for maildirsize files, in the light
4878 of a number of problems with the previous contributed implementation
4879 (4.30/29). In particular:
4881 . If the quota is zero, the maildirsize file is maintained, but no quota is
4884 . If the maildir directory does not exist, it is created before any attempt
4885 to write a maildirsize file.
4887 . The quota value in the file is just a cache; if the quota is changed in
4888 the transport, the new value overrides.
4890 . A regular expression is available for excluding directories from the
4893 20. The autoreply transport checks the characters in options that define the
4894 message's headers; it allows continued headers, but it was checking with
4895 isspace() after an embedded newline instead of explicitly looking for a
4898 21. If all the "regular" hosts to which an address was routed had passed their
4899 expiry times, and had not reached their retry times, the address was
4900 bounced, even if fallback hosts were defined. Now Exim should go on to try
4903 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4904 equivalent code in the SMTP transport. Some hosts send humungous responses
4905 to HELO/EHLO, more than 1024 it seems.
4907 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4908 gives compiler warnings in some environments. I've now done it "properly",
4911 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4912 (because of problems with the built-in one) was declared to return uschar *
4913 instead of char *, causing compiler failure.
4915 25. Fixed a file descriptor leak when processing alias/forward files.
4917 26. Fixed a minor format string issue in dbfn.c.
4919 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4921 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4922 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4925 29. When a sender address is verified, it is cached, to save repeating the test
4926 when there is more than one recipient in a message. However, when the
4927 verification involves a callout, it is possible for different callout
4928 options to be set for different recipients. It is too complicated to keep
4929 track of this in the cache, so now Exim always runs a verification when a
4930 callout is required, relying on the callout cache for the optimization.
4931 The overhead is duplication of the address routing, but this should not be
4934 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4935 to be verified with callout=postmaster, and the main callout worked but the
4936 postmaster check failed, the verification correctly failed. However, if a
4937 subsequent RCPT command asked for sender verification *without* the
4938 postmaster check, incorrect caching caused this verification also to fail,
4941 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4942 it was not caching the DNS options (qualify_single, search_parents) that
4943 were used when the lookup failed. A subsequent lookup with different
4944 options therefore always gave the same answer, though there were cases
4945 where it should not have. (Example: a "domains = !$mx_any" option on a
4946 dnslookup router: the "domains" option is always processed without any
4947 widening, but the router might have qualify_single set.) Now Exim uses the
4948 cached value only when the same options are set.
4950 32. Added John Jetmore's "exipick" utility to the distribution.
4952 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4953 than a timeout (e.g. a certificate is required, and is not provided), an
4954 Exim server now closes the connection immediately. Previously it waited for
4955 the client to close - but if the client is SSL, it seems that they each
4956 wait for each other, leading to a delay before one of them times out.
4958 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4959 maintained 0.8.x compatibility because I don't think many are using it, and
4960 it is clearly obsolete.
4962 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4965 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4966 client certificate was expired. A simple patch fixes this, though I don't
4967 understand the full logic of why the verify callback is called multiple
4970 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4971 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4972 which causes problems with some clients (such as the Certicom SSL Plus
4973 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4974 disables the coutermeasure allowing Eudora to connect."
4976 38. Exim was not checking that a write() to a log file succeeded. This could
4977 lead to Bad Things if a log got too big, in particular if it hit a file
4978 size limit. Exim now panics and dies if it cannot write to a log file, just
4979 as it does if it cannot open a log file.
4981 39. Modified OS/Makefile-Linux so that it now contains
4983 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4985 The two -D definitions ensure that Exim is compiled with large file
4986 support, which makes it possible to handle log files that are bigger than
4989 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4990 instance) a domain was checked against a named list that involved a lookup,
4991 causing $domain_data to be set, then another domain was checked against the
4992 same list, then the first domain was re-checked, the value of $domain_data
4993 after the final check could be wrong. In particular, if the second check
4994 failed, it could be set empty. This bug probably also applied to
4997 41. The strip_trailing_dot option was not being applied to the address given
4998 with the -f command-line option.
5000 42. The code for reading a message's header from the spool was incrementing
5001 $received_count, but never initializing it. This meant that the value was
5002 incorrect (doubled) while delivering a message in the same process in which
5003 it was received. In the most common configuration of Exim, this never
5004 happens - a fresh exec is done - but it can happen when
5005 deliver_drop_privilege is set.
5007 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5008 it now includes up to 150 characters of the unexpected data in the log
5011 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5012 and building data strings. The size of both of these buffers was 10 000
5013 bytes - far larger than anybody would *ever* want, thought I. Needless to
5014 say, somebody hit the limit. I have increased the maximum line length to
5015 20 000 and the maximum data length of concatenated lines to 100 000. I have
5016 also fixed two bugs, because there was no checking on these buffers. Tsk,
5017 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5018 buffer is too small.
5020 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5021 lsearch lookups. Now it does.
5023 46. When parsing a route_list item in a manualroute router, a fixed-length
5024 buffer was used for the list of hosts. I made this 1024 bytes long,
5025 thinking that nobody would ever have a list of hosts that long. Wrong.
5026 Somebody had a whole pile of complicated expansion conditions, and the
5027 string was silently truncated, leading to an expansion error. It turns out
5028 that it is easier to change to an unlimited length (owing to other changes
5029 that have happened since this code was originally written) than to build
5030 structure for giving a limitation error. The length of the item that
5031 expands into the list of hosts is now unlimited.
5033 47. The lsearch lookup could not handle data where the length of text line was
5034 more than 4095 characters. Such lines were truncated, leading to shortened
5035 data being returned. It should now handle lines of any length.
5037 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5038 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5041 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5042 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5043 no longer happens in comment lines. A list of replacements is now placed
5044 at the head of all of the source files, except those whose only change is
5045 to replace PERL_COMMAND in the very first #! line.
5047 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5048 messages on the queue, with a bottom-up merge sort, using code contributed
5049 by Michael Haardt. This should make operations like -bp somewhat faster on
5050 large queues. It won't affect queue runners, except when queue_run_in_order
5053 51. Installed eximstats 1.31 in the distribution.
5055 52. Added support for SRV lookups to the dnslookup router.
5057 53. If an ACL referred to $message_body or $message_body_end, the value was not
5058 reset for any messages that followed in the same SMTP session.
5060 54. The store-handling optimization for building very long strings was not
5061 differentiating between the different store pools. I don't think this
5062 actually made any difference in practice, but I've tidied it.
5064 55. While running the routers to verify a sender address, $sender_address
5065 was still set to the sender address. This is wrong, because when routing to
5066 send a bounce to the sender, it would be empty. Therefore, I have changed
5067 it so that, while verifying a sender address, $sender_address is set to <>.
5068 (There is no change to what happens when verifying a recipient address.)
5070 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5071 target A or AAAA records (if not already returned) without resetting the
5072 qualify_single or search_parents options of the DNS resolver. These are
5073 inappropriate in this case because the targets of MX and SRV records must
5074 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5075 target that, when qualified, matched something in the local domain. These
5076 two options are now turned off when doing these lookups.
5078 57. It seems that at least some releases of Reiserfs (which does not have the
5079 concept of a fixed number of inodes) returns zero and not -1 for the
5080 number of available inodes. This interacted badly with check_spool_inodes,
5081 which assumed that -1 was the "no such thing" setting. What I have done is
5082 to check that the total number of inodes is greater than zero before doing
5083 the test of how many are available.
5085 58. When a "warn" ACL statement has a log_message modifier, the message is
5086 remembered, and not repeated. This is to avoid a lot of repetition when a
5087 message has many recipients that cause the same warning to be written.
5088 Howewer, Exim was preserving the list of already written lines for an
5089 entire SMTP session, which doesn't seem right. The memory is now reset if a
5090 new message is started.
5092 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5093 debugging output unless log_rewrite was also set.
5095 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5096 of (int)(handle) when we know that handle contains (void *)(-1).
5098 61. The Exim daemon panic-logs an error return when it closes the incoming
5099 connection. However "connection reset by peer" seems to be common, and
5100 isn't really an error worthy of noting specially, so that particular error
5103 62. When Exim is trying to find all the local interfaces, it used to panic and
5104 die if the ioctl to get the interface flags failed. However, it seems that
5105 on at least one OS (Solaris 9) it is possible to have an interface that is
5106 included in the list of interfaces, but for which you get a failure error
5107 for this call. This happens when the interface is not "plumbed" into a
5108 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5109 failure of the "get flags" call assumes that the interface is down.
5111 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5112 makes life easier for people who are doing arithmetic on fields extracted
5113 from dates, where you often get leading zeros that should not be
5114 interpreted as octal.
5116 64. Added qualify_domain to the redirect router, to override the global
5119 65. If a pathologically long header line contained very many addresses (the
5120 report of this problem mentioned 10 000) and each of them was rewritten,
5121 Exim could use up a very large amount of memory. (It kept on making new
5122 copies of the header line as it rewrote, and never released the old ones.)
5123 At the expense of a bit more processing, the header rewriting function has
5124 been changed so that it no longer eats memory in this way.
5126 66. The generation of the Received: header has been moved from the time that a
5127 message starts to be received, to the time that it finishes. The timestamp
5128 in the Received: header should now be very close to that of the <= log
5129 line. There are two side-effects of this change:
5131 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5132 logged header lines no longer include the local Received: line, because
5133 it has not yet been created. The same applies to a copy of the message
5134 that is returned to a non-SMTP sender when a message is rejected.
5136 (b) When a filter file is tested using -bf, no additional Received: header
5137 is added to the test message. After some thought, I decided that this
5140 This change does not affect the value of $received_for. It is still set
5141 after address rewriting, but before local_scan() is called.
5143 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5145 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5146 gave an unhelpful panic error message, and a defer error. I have managed to
5147 change this behaviour so that it now rejects any supplied certificate,
5148 which seems right, as the list of acceptable certificates is empty.
5150 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5151 gave an unhelpful defer error. I have not managed to make this reject any
5152 supplied certificates, but the error message it gives is "no certificate
5153 supplied", which is not helpful.
5155 70. exigrep's output now also includes lines that are not associated with any
5156 message, but which match the given pattern. Implemented by a patch from
5157 Martin Sluka, which also tidied up the Perl a bit.
5159 71. Recipient callout verification, like sender verification, was using <> in
5160 the MAIL FROM command. This isn't really the right thing, since the actual
5161 sender may affect whether the remote host accepts the recipient or not. I
5162 have changed it to use the actual sender in the callout; this means that
5163 the cache record is now keyed on a recipient/sender pair, not just the
5164 recipient address. There doesn't seem to be a real danger of callout loops,
5165 since a callout by the remote host to check the sender would use <>.
5166 [SEE ABOVE: changed after hitting problems.]
5168 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5169 temporary errors. However, in the case of such a code being given after
5170 the end of a data transmission (i.e. after ".") Exim was failing to write
5171 a retry record for the message. (Yes, there was some broken host that was
5172 actually sending 8xx at this point.)
5174 73. An unknown lookup type in a host list could cause Exim to panic-die when
5175 the list was checked. (An example that provoked this was putting <; in the
5176 middle of a list instead of at the start.) If this happened during a DATA
5177 ACL check, a -D file could be left lying around. This kind of configuration
5178 error no longer causes Exim to die; instead it causes a defer errror. The
5179 incident is still logged to the main and panic logs.
5181 74. Buglet left over from Exim 3 conversion. The message "too many messages
5182 in one connection" was written to the rejectlog but not the mainlog, except
5183 when address rewriting (yes!) was being logged.
5185 75. Added write_rejectlog option.
5187 76. When a system filter was run not as root (that is, when system_filter_user
5188 was set), the values of the $n variables were not being returned to the
5189 main process; thus, they were not subsequently available in the $sn
5192 77. Added +return_path_on_delivery log selector.
5194 78. A connection timeout was being treated differently from recipients deferred
5195 when testing hosts_max_try with a message that was older than the host's
5196 retry timeout. (The host should not be counted, thus allowing all hosts to
5197 be tried at least once before bouncing.) This may have been the cause of an
5198 occasionally reported bug whereby a message would remain on the queue
5199 longer than the retry timeout, but would be bounced if a delivery was
5200 forced. I say "may" because I never totally pinned down the problem;
5201 setting up timeout/retry tests is difficult. See also the next item.
5203 79. The ultimate address timeout was not being applied to errors that involved
5204 a combination of host plus message (for example, a timeout on a MAIL
5205 command). When an address resolved to a number of possible hosts, and they
5206 were not all tried for each delivery (e.g. because of hosts_max_try), a
5207 message could remain on the queue longer than the retry timeout.
5209 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5212 81. Fixed an obscure SMTP outgoing bug which required at least the following
5213 conditions: (a) there was another message waiting for the same server;
5214 (b) the server returned 5xx to all RCPT commands in the first message so
5215 that the message was not completed; (c) the server dropped the connection
5216 or gave a negative response to the RSET that Exim sends to abort the
5217 transaction. The observed case was a dropped connection after DATA that had
5218 been sent in pipelining mode. That is, the server had advertised PIPELINING
5219 but was not implementing it correctly. The effect of the bug was incorrect
5220 behaviour, such as trying another host, and this could lead to a crash.
5226 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5227 and daemon.c were passed as pointers to ints; they should have been
5228 pointers to socklen_t variables (which are typically unsigned ints).
5230 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5233 3. Fixed a really odd bug that affected only the testing scheme; patching a
5234 certain fixed string in the binary changed the value of another string that
5235 happened to be identical to the end of the original first string.
5237 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5238 name", it returns that address as the IP address. On some operating
5239 systems (e.g. Solaris), it also passes back the IP address string as the
5240 "host name". However, on others (e.g. Linux), it passes back an empty
5241 string. Exim wasn't checking for this, and was changing the host name to an
5242 empty string, assuming it had been canonicized.
5244 5. Although rare, it is permitted to have more than one PTR record for a given
5245 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5246 all the names associated with an address, because they do in Solaris.
5247 However, it seems that they do not in Linux for data that comes from the
5248 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5249 I found this out when I moved to a new Linux workstation and tried to run
5250 the Exim test suite.
5252 To get round this problem I have changed the code so that it now does its
5253 own call to the DNS to look up PTR records when searching for a host name.
5254 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5255 addresses that are only in /etc/hosts are still found.
5257 This behaviour is, however, controlled by an option called host_lookup_
5258 order, which defaults to "bydns:byaddr". If people want to use the other
5259 order, or indeed, just use one or the other means of lookup, they can
5260 specify it in this variable.
5262 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5263 some operating systems, this comes back from gethostbyaddr() as an empty
5264 string, and this is what Exim used to test for. However, it seems that in
5265 other systems, "." is yielded. Exim now tests for this case too.
5267 7. The values of check_spool_space and check_log_space are now held internally
5268 as a number of kilobytes instead of an absolute number of bytes. If a
5269 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5270 kilobyte. This means that much larger values can be stored.
5272 8. Exim monitor: an attempt to get the action menu when not actually pointing
5273 at a message produces an empty menu entitled "No message selected". This
5274 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5275 no entries in it ("Shell widget menu has zero width and/or height"). So I
5276 have added a single, blank menu entry in this case.
5278 9. Added ${quote_local_part.
5280 10. MIME decoding is now applied to the contents of Subject: header lines when
5283 11. Now that a reference to $sender_host_address automatically causes a reverse
5284 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5285 host lookup before query-style lookups in lists that might use this
5286 variable. This has therefore been abolished, and the "net-" prefix is no
5287 longer necessary for query-style lookups.
5289 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5290 have been a typo for LFLAGS, so it has been changed.
5292 13. The install script calls Exim with "-C /dev/null" in order to find the
5293 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5294 to be output. Howeve, since Exim outputs its version number before the
5295 error, it didn't break the script. It just looked ugly. I fixed this by
5296 always allowing "-C /dev/null" if the caller is root.
5298 14. Ignore overlarge ACL variable number when reading spool file - insurance
5299 against a later release with more variables having written the file.
5301 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5302 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5303 this, as well as the form without the "IPv6" on the front.
5305 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5306 OS/Makefile-Darwin file.
5308 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5309 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5311 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5312 variable. However, some OS use ints for inodes. I've added cast to long int
5313 to get rid of the compiler warning.
5315 19. I had forgotten to lock out "/../" in configuration file names when
5316 ALT_CONFIG_PREFIX was set.
5318 20. Routers used for verification do not need to specify transports. However,
5319 if such a router generated a host list, and callout was configured, Exim
5320 crashed, because it could not find a port number from the (non-existent)
5321 transport. It now assumes port 25 in this circumstance.
5323 21. Added the -t option to exigrep.
5325 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5326 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5327 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5328 them is set, LOOKUP_LSEARCH is forced.
5330 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5332 24. Added sender and host information to the "rejected by local_scan()" log
5333 line; previously there was no indication of these.
5335 25. Added .include_if_exists.
5337 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5338 Linux. It turns out that not all file systems support this. Apparently some
5339 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5340 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5341 not supported on the file descriptor, is now ignored when Exim is trying to
5342 sync a directory. This applies only to Linux.
5344 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5346 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5347 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5350 29. Added support for maildirsize files from supplied patch (modified a bit).
5352 30. The use of :fail: followed by an empty string could lead Exim to respond to
5353 sender verification failures with (e.g.):
5355 550 Verification failed for <xxx>
5356 550 Sender verify failed
5358 where the first response line was missing the '-' that indicates it is not
5359 the final line of the response.
5361 31. The loop for finding the name of the user that called Exim had a hardwired
5362 limit of 10; it now uses the value of finduser_retries, which is used for
5363 all other user lookups.
5365 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5368 33. Exim was neglecting to zero errno before one call of strtol() when
5369 expanding a string and expecting an integer value. On some systems this
5370 resulted in spurious "integer overflow" errors. Also, it was casting the
5371 result into an int without checking.
5373 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5374 did not work. The code looks as if it has *never* worked, though it appears
5375 to have been documented since at least releast 1.62. I have made it work.
5377 35. The "timeout_DNS" error in retry rules, also documented since at least
5378 1.62, also never worked. As it isn't clear exactly what this means, and
5379 clearly it isn't a major issue, I have abolished the feature by treating it
5380 as "timeout", and writing a warning to the main and panic logs.
5382 36. The display of retry rules for -brt wasn't always showing the error code
5385 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5386 timeout_connect_A, timeout_connect_MX.
5388 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5389 to the empty sender.
5391 39. The daemon was not analysing the content of -oX till after it had closed
5392 stderr and disconnected from the controlling terminal. This meant that any
5393 syntax errors were only noted on the panic log, and the return code from
5394 the command was 0. By re-arranging the code a little, I've made the
5395 decoding happen first, so such errors now appear on stderr, and the return
5396 code is 1. However, the actual setting up of the sockets still happens in
5397 the disconnected process, so errors there are still only recorded on the
5400 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5401 connections (as happens on some IP stacks) was logged at start up time as
5402 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5403 it from "IPv6 and IPv4", which means that two separate sockets are being
5406 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5407 says whether AF_INET or AF_INET6 was passed as an argument.
5409 42. Exiwhat output was messed up when time zones were included in log
5412 43. Exiwhat now gives more information about the daemon's listening ports,
5413 and whether -tls-on-connect was used.
5415 44. The "port" option of the smtp transport is now expanded.
5417 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5418 silently ignored. Now an error message is written to the main and panic
5421 46. There's a new ACL modifier called "logwrite" which writes to a log file
5422 as soon as it is encountered.
5424 47. Added $local_user_uid and $local_user_gid at routing time.
5426 48. Exim crashed when trying to verify a sender address that was being
5429 49. Exim was recognizing only a space character after ".include". It now also
5430 recognizes a tab character.
5432 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5433 extracting the relevant information from the specification. The man page no
5434 longer contains scrambled data for the -d option, and I've added a section
5435 at the front about calling Exim under different names.
5437 51. Added "extra_headers" argument to the "mail" command in filter files.
5439 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5442 53. Installed eximstats 1.29.
5444 54. Added transport_filter_timeout as a generic transport option.
5446 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5447 Cc: header lines. This was required by RFC 822, but it not required by RFC
5450 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5451 incoming messages that did not have them. Now it does so only if the
5452 message originates locally, that is, if there is no associated remote host
5453 address. When Resent- header lines are present, this applies to the Resent-
5454 lines rather than the non-Resent- lines.
5456 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5457 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5459 58. Messages for configuration errors now include the name of the main
5460 configuration file - useful now that there may be more than one file in a
5461 list (.included file names were always shown).
5463 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5464 for those rare installations that do not start the daemon as root or run it
5465 setuid root. I've cut out the call to initgroups() if the daemon is not
5468 60. The Exim user and group can now be bound into the binary as text strings
5469 that are looked up at the start of Exim's processing.
5471 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5473 62. Added $mailstore_basename variable.
5475 63. Installed patch to sieve.c from Michael Haardt.
5477 64. When Exim failed to open the panic log after failing to open the main log,
5478 the original message it was trying to log was written to stderr and debug
5479 output, but if they were not available (the usual case in production), it
5480 was lost. Now it is written to syslog before the two lines that record the
5481 failures to open the logs.
5483 65. Users' Exim filters run in subprocesses under the user's uid. It is
5484 possible for a "deliver" command or an alias in a "personal" command to
5485 provoke an address rewrite. If logging of address rewriting is configured,
5486 this fails because the process is not running as root or exim. There may be
5487 a better way of dealing with this, but for the moment (because 4.30 needs
5488 to be released), I have disabled address rewrite logging when running a
5489 filter in a non-root, non-exim process.
5495 1. The buildconfig auxiliary program wasn't quoting the value set for
5496 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5497 not defined. This bug was masked in 4.22 by the effect that was fixed in
5500 2. Some messages that were rejected after a message id was allocated were
5501 shown as "incomplete" by exigrep. It no longer does this for messages that
5502 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5504 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5505 have allow_domain_literals set, the ID did not get logged in the <= line.
5506 Domain literals are now always recognized in Message-ID: header lines.
5508 4. The first argument for a ${extract expansion item is the key name or field
5509 number. Leading and trailing spaces in this item were not being ignored,
5510 causing some misleading effects.
5512 5. When deliver_drop_privilege was set, single queue runner processes started
5513 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5514 same command in the process it spins off) were not dropping privilege.
5516 6. When the daemon running as "exim" started a queue runner, it always
5517 re-executed Exim in the spun-off process. This is a waste of effort when
5518 deliver_drop_privilege is set. The new process now just calls the
5519 queue-runner function directly.
5525 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5528 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5529 ignored. Though the use of -oP was forcing the writing of a pid file, it
5530 was always written to the default place.
5532 3. If the message "no IP address found for host xxxx" is generated during
5533 incoming verification, it is now followed by identification of the incoming
5534 connection (so you can more easily find what provoked it).
5536 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5538 5. Added some features to "harden" Exim a bit more against certain attacks:
5540 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5541 be put in Local/Makefile. This is like the never_users runtime option,
5542 but it cannot be overridden. The default setting is "root".
5544 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5545 prefix string with which any file named in a -C command line option
5548 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5549 is retained for -C and -D only if the caller of Exim is root. Without
5550 it, the exim user may also use -C and -D and retain privilege.
5552 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5553 command line option is disabled.
5555 6. Macro names set by the -D option must start with an upper case letter, just
5556 like macro names defined in the configuration file.
5558 7. Added "dereference=" facility to LDAP.
5560 8. Two instances of the typo "uknown" in the source files are fixed.
5562 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5563 the Configure-Makefile script screwed up while processing it.
5565 10. Incorporated PCRE 4.4.
5567 11. The SMTP synchronization check was not operating right at the start of an
5568 SMTP session. For example, it could not catch a HELO sent before the client
5569 waited for the greeting. There is now a check for outstanding input at the
5570 point when the greeting is written. Because of the duplex, asynchronous
5571 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5572 way, but not yet received, when the check is performed.
5574 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5575 on TCP/IP sockets, because this apparently causes some broken clients to
5578 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5579 unchanged) from the Cygwin maintainer.
5581 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5582 was supported instead of testing for mbx. Effectively a typo.
5584 15. The spa authenticator server code was not checking that the input it
5585 received was valid base64.
5587 16. The debug output line for the "set" modifier in ACLs was not showing the
5588 name of the variable that was being set.
5590 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5592 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5593 lookup on its own. Something else was needed to trigger the lookup. For
5594 example, a match in host_lookup or the need for a host name in a host list.
5595 Now, if $sender_host_name is referenced and the host name has not yet been
5596 looked up, a lookup is performed. If the lookup fails, the variable remains
5597 empty, and $host_lookup_failed is set to "1".
5599 19. Added "eqi" as a case-independent comparison operator.
5601 20. The saslauthd authentication condition could segfault if neither service
5602 nor realm was specified.
5604 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5605 error message that was logged was misleading, and incoming SMTP
5606 connections were dropped. The message is now more accurate, and temporary
5607 errors are given to SMTP connections.
5609 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5610 (except RSET) with a 421 error, until QUIT is received. However, it was
5611 failing to send a response to QUIT.
5613 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5614 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5615 ACL is now run after the helo_try_verify_hosts code.
5617 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5618 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5619 case-independent, so other case variants are also recognized). Apparently
5620 some systems use these upper case variants.
5622 25. If more than two messages were waiting for the same host, and a transport
5623 filter was specified for the transport, Exim sent two messages over the
5624 same TCP/IP connection, and then failed with "socket operation on non-
5625 socket" when it tried to send the third.
5627 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5629 27. The extern definition of crypt16() in expand.c was not being excluded when
5630 the OS had its own crypt16() function.
5632 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5633 as a preferred synonym for return_size_limit, both as an option and as an
5636 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5638 30. Changed the default configuration ACL to relax the local part checking rule
5639 for addresses that are not in any local domains. For these addresses,
5640 slashes and pipe symbols are allowed within local parts, but the sequence
5641 /../ is explicitly forbidden.
5643 31. SPA server authentication was not clearing the challenge buffer before
5646 32. log_message in a "warn" ACL statement was writing to the reject log as
5647 well as to the main log, which contradicts the documentation and doesn't
5648 seem right (because no rejection is happening). So I have stopped it.
5650 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5651 However, I am unable to do any testing of this.
5653 34. Fixed an infelicity in the appendfile transport. When checking directories
5654 for a mailbox, to see if any needed to be created, it was accidentally
5655 using path names with one or more superfluous leading slashes; tracing
5656 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5658 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5659 contents are added to the log line that is written for every discarded
5660 recipient. (Previously a log_message setting was ignored.)
5662 36. The ${quote: operator now quotes the string if it is empty.
5664 37. The install script runs exim in order to find its version number. If for
5665 some reason other than non-existence or emptiness, which it checks, it
5666 could not run './exim', it was installing it with an empty version number,
5667 i.e. as "exim-". This error state is now caught, and the installation is
5670 38. An argument was missing from the function that creates an error message
5671 when Exim fails to connect to the socket for saslauthd authentication.
5672 This could cause Exim to crash, or give a corrupted message.
5674 39. Added isip, isip4, and isip6 to ${if conditions.
5676 40. The ACL variables $acl_xx are now saved with the message, and can be
5677 accessed later in routers, transports, and filters.
5679 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5680 strings in the file are not string-expanded.
5682 42. If a MAIL command specified a SIZE value that was too large to fit into an
5683 int variable, the check against message_size_limit failed. Such values are
5684 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5685 day this will have to be increased, but I don't think I want to be around
5686 when emails are that large.
5693 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5694 iconv() is not standard in FreeBSD.
5696 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5697 IPv6 enabled. The observed symptom was a segmentation fault on return from
5698 the function os_common_find_running_interfaces() in src/os.c.
5700 3. In the check_special_case() function in daemon.c I had used "errno" as an
5701 argument name, which causes warnings on some systems. This was basically a
5702 typo, since it was named "eno" in the comments!
5704 4. The code that waits for the clock to tick (at a resolution of some fraction
5705 of a second) so as to ensure message-id uniqueness was always waiting for
5706 at least one whole tick, when it could have waited for less. [This is
5707 almost certainly not relevant at current processor speeds, where it is
5708 unlikely to ever wait at all. But we try to future-proof.]
5710 5. The function that sleeps for a time interval that includes fractions of a
5711 second contained a race. It did not block SIGALRM between setting the
5712 timer, and suspending (a couple of lines later). If the interval was short
5713 and the sigsuspend() was delayed until after it had expired, the suspension
5714 never ended. On busy systems this could lead to processes getting stuck for
5717 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5718 process, before it forks any delivery processes. The open lookup caching
5719 mechanism meant that the open file or database connection was passed into
5720 the delivery process. The problem was that delivery processes always tidy
5721 up cached lookup data. This could cause a problem for the next delivery
5722 process started by the queue runner, because the external queue runner
5723 process does not know about the closure. So the next delivery process
5724 still has data in the lookup cache. In the case of a file lookup, there was
5725 no problem because closing a file descriptor in a subprocess doesn't affect
5726 the parent. However, if the lookup was caching a connection to a database,
5727 the connection was closed, and the second delivery process was likely to
5728 see errors such as "PGSQL: query failed: server closed the connection
5729 unexpectedly". The problem has been fixed by closing all cached lookups
5730 in a queue runner before running a delivery process.
5732 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5733 seem to have the "const" qualifier which it has on other OS. I've
5736 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5737 *of the same type* (client or server) with the same public name that an
5738 error should be diagnosed.
5740 9. When Exim looked up a host name for an IP address, but failed to find the
5741 original IP address when looking up the host name (a safety check), it
5742 output the message "<ip address> does not match any IP for NULL", which was
5743 confusing, to say the least. The bug was that the host name should have
5744 appeared instead of "NULL".
5746 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5747 user that is built into the binary, and the -C or -D options is used, root
5748 privilege is dropped before the configuration file is read. In addition,
5749 logging is switched to stderr instead of the normal log files. If the
5750 configuration then re-defines the Exim user, the unprivileged environment
5751 is probably not what is expected, so Exim logs a panic warning message (but
5754 However, if deliver_drop_privilege is set, the unprivileged state may well
5755 be exactly what is intended, so the warning has been cut out in that case,
5756 and Exim is allowed to try to write to its normal log files.
5762 1. smtp_return_error_details was not giving details for temporary sender
5763 or receiver verification errors.
5765 2. Diagnose a configuration error if two authenticators have the same public
5768 3. Exim used not to create the message log file for a message until the first
5769 delivery attempt. This could be confusing when incoming messages were held
5770 for policy or load reasons. The message log file is now created at the time
5771 the message is received, and an initial "Received" line is written to it.
5773 4. The automatically generated man page for command line options had a minor
5774 bug that caused no ill effects; however, a more serious problem was that
5775 the procedure for building the man page automatically didn't always
5776 operate. Consequently, release 4.20 contains an out-of-date version. This
5777 shouldn't happen again.
5779 5. When building Exim with embedded Perl support, the script that builds the
5780 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5781 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5783 6. The freeze_tell option was not being used for messages that were frozen on
5784 arrival, either by an ACL or by local_scan().
5786 7. Added the smtp_incomplete_transaction log selector.
5788 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5789 was accepting AUTH without a new EHLO.
5791 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5792 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5795 10. Logging of TCP/IP connections (when configured) now happens in the main
5796 daemon process instead of the child process, so that the TCP/IP connection
5797 count is more accurate (but it can never be perfect).
5799 11. The use of "drop" in a nested ACL was not being handled correctly in the
5800 outer ACL. Now, if condition failure induced by the nested "drop" causes
5801 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5802 or "require"), the connection is dropped.
5804 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5805 that yield "discard" can now be used with an "accept" or a "discard" verb,
5806 but an error is generated for any others (because I can't see a useful way
5807 to define what should happen).
5809 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5810 are now processed in the same way as lines in the Exim configuration file.
5811 In particular, continuation lines are supported.
5813 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5815 15. Added -ti meaning -t -i.
5817 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5818 domains, and warn by logging if others are found.
5820 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5821 returned from GIFCONF when Exim is trying to find the list of interfaces on
5822 a host. The code in os.c has been modified to copy each ifreq to an aligned
5823 structure in all cases.
5825 Also, in some cases, the returned ifreq's were being copied to a 'struct
5826 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5827 means the last couple of bytes of an IPv6 address could be chopped if the
5828 ifreq contained only a normal sockaddr (14 bytes storage).
5830 18. Named domain lists were not supported in the hosts_treat_as_local option.
5831 An entry such as +xxxx was not recognized, and was treated as a literal
5834 19. Ensure that header lines added by a DATA ACL are included in the reject log
5835 if the ACL subsequently rejects the message.
5837 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5838 MD5 (which is deprecated).
5840 21. When testing a filter file using -bf, Exim was writing a message when it
5841 took the sender from a "From " line in the message, but it was not doing so
5842 when it took $return_path from a Return-Path: header line. It now does.
5844 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5845 with a valid header line field name (a series of printing characters
5846 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5848 23. Changed "disc" in the source to "disk" to conform to the documentation and
5849 the book and for uniformity.
5851 24. Ignore Sendmail's -Ooption=value command line item.
5853 25. When execve() failed while trying to run a command in a pipe transport,
5854 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5855 could be confused with a return value of 69 from the command itself. This
5856 has been changed to 127, the value the shell returns if it is asked to run
5857 a non-existent command. The wording for the related log line suggests a
5858 non-existent command as the problem.
5860 26. If received_header_text expands to an empty string, do not add a Received:
5861 header line to the message. (Well, it adds a token one on the spool, but
5862 marks it "old" so that it doesn't get used or transmitted.)
5864 27. Installed eximstats 1.28 (addition of -nt option).
5866 28. There was no check for failure on the call to getsockname() in the daemon
5867 code. This can fail if there is a shortage of resources on the system, with
5868 ENOMEM, for example. A temporary error is now given on failure.
5870 29. Contrary to the C standard, it seems that in some environments, the
5871 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5872 program. Exim now does this explicitly; it affects the formatting of
5873 timestamps using strftime().
5875 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5876 complaints. I've now initialized all the variables, to avoid this.
5878 32. Header lines added by a system filter were not being "seen" during
5879 transport-time rewrites.
5881 33. The info_callback() function passed to OpenSSL is set up with type void
5882 (*)(SSL *, int, int), as described somewhere. However, when calling the
5883 function (actually a macro) that sets it up, the type void(*)() is
5884 expected. I've put in a cast to prevent warnings from picky compilers.
5886 34. If a DNS black list lookup found a CNAME record, but there were no A
5887 records associated with the domain it pointed at, Exim crashed.
5889 35. If a DNS black list lookup returned more than one A record, Exim ignored
5890 all but the first. It now scans all returned addresses if a particular IP
5891 value is being sought. In this situation, the contents of the
5892 $dnslist_value variable are a list of all the addresses, separated by a
5895 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5896 to accept a host name and all its aliases if the forward lookup for any of
5897 them yielded the IP address of the incoming connection. Now it accepts only
5898 those names whose forward lookup yields the correct IP address. Any other
5899 names are discarded. This closes a loophole whereby a rogue DNS
5900 administrator could create reverse DNS records to break through a
5901 wildcarded host restriction in an ACL.
5903 37. If a user filter or a system filter that ran in a subprocess used any of
5904 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5905 the wrong values were passed to the pipe command ($thisaddress had the
5906 value of $0, $0 had the value of $1, etc). This bug was introduced by
5907 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5909 38. Improved the line breaking for long SMTP error messages from ACLs.
5910 Previously, if there was no break point between 40 and 75 characters, Exim
5911 left the rest of the message alone. Two changes have been made: (a) I've
5912 reduced the minimum length to 35 characters; (b) if it can't find a break
5913 point between 35 and 75 characters, it looks ahead and uses the first one
5914 that it finds. This may give the occasional overlong line, but at least the
5915 remaining text gets split now.
5917 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5918 file descriptors might be low, and that setting 1000 would always raise it.
5919 It turns out that in some environments, the limit is already over 1000 and
5920 that lowering it causes trouble. So now Exim takes care not to decrease it.
5922 40. When delivering a message, the value of $return_path is set to $sender_
5923 address at the start of routing (routers may change the value). By an
5924 oversight, this default was not being set up when an address was tested by
5925 -bt or -bv, which affected the outcome if any router or filter referred to
5928 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5929 log only when "message" or "log_message" are set. However, if one of the
5930 conditions was an address verification, or a call to a nested ACL, the
5931 messages generated by the underlying test were being passed through. This
5932 no longer happens. The underlying message is available in $acl_verify_
5933 message for both "message" and "log_message" expansions, so it can be
5934 passed through if needed.
5936 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5937 new expansion $bh_ to give the encoded byte string without charset
5938 translation. Translation happens only if iconv() is available; HAVE_ICONV
5939 indicates this at build time. HEADERS_CHARSET gives the charset to
5940 translate to; headers_charset can change it in the configuration, and
5941 "headers charset" can change it in an individual filter file.
5943 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5944 that creates RFC 2047 encoded "words" labels them as that charset instead
5945 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5946 expansion operator; (ii) when Exim creates a From: line for a local
5947 message; (iii) when a header line is rewritten to include a "phrase" part.
5949 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5950 buggy, causing it to skip the first lines of messages whose message ID
5951 ended in 'D'. This would not have bitten before Exim release 4.14, because
5952 message IDs were unlikely to end in 'D' before then. The effect was to have
5953 incorrect size information for certain domains.
5955 45. #include "config.h" was missing at the start of the crypt16.c module. This
5956 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5959 46. If there was a timeout during a "random" callout check, Exim treated it as
5960 a failure of the random address, and carried on sending RSET and the real
5961 address. If the delay was just some slowness somewhere, the response to the
5962 original RCPT would be taken as a response to RSET and so on, causing
5963 mayhem of various kinds.
5965 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5966 when I implemented it. It didn't allow for the fact that some option values
5967 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5968 the right test for positive values.
5970 48. Domain names in DNS records are case-independent. Exim always looks them up
5971 in lower case. Some resolvers return domain names in exactly the case they
5972 appear in the zone file, that is, they may contain uppercase letters. Not
5973 all resolvers do this - some return always lower case. Exim was treating a
5974 change of case by a resolver as a change of domain, similar to a widening
5975 of a domain abbreviation. This triggered its re-routing code and so it was
5976 trying to route what was effectively the same domain again. This normally
5977 caused routing to fail (because the router wouldn't handle the domain
5978 twice). Now Exim checks for this case specially, and just changes the
5979 casing of the domain that it ultimately uses when it transmits the message
5982 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5985 50. If a filter generated a file delivery with a non-absolute name (possible if
5986 no home directory exists for the router), the forbid_file option was not
5989 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5990 the existing equality matching.
5992 52. Exim was using ints instead of ino_t variables in some places where it was
5993 dealing with inode numbers.
5995 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5996 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5997 called TMPDIR, and if it finds it is different, it changes its value.
5999 54. The smtp_printf() function is now made available to local_scan() so
6000 additional output lines can be written before returning. There is also an
6001 smtp_fflush() function to enable the detection of a dropped connection.
6002 The variables smtp_input and smtp_batched_input are exported to
6005 55. Changed the default runtime configuration: the message "Unknown user"
6006 has been removed from the ACL, and instead placed on the localuser router,
6007 using the cannot_route_message feature. This means that any verification
6008 failures that generate their own messages won't get overridden. Similarly,
6009 the "Unrouteable address" message that was in the ACL for unverifiable
6010 relay addresses has also been removed.
6012 56. Added hosts_avoid_esmtp to the smtp transport.
6014 57. The exicyclog script was not checking for the esoteric option
6015 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6016 will work only if exicyclog is run under the appropriate euid.
6018 58. Following a discussion on the list, the rules by which Exim recognises line
6019 endings on incoming messages have been changed. The -dropcr and drop_cr
6020 options are now no-ops, retained only for backwards compatibility. The
6021 following line terminators are recognized: LF CRLF CR. However, special
6022 processing applies to CR:
6024 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6025 nor a local message in the state where . is a terminator.
6027 (ii) If a bare CR is encountered in a header line, an extra space is added
6028 after the line terminator so as not to end the header. The reasoning
6029 behind this is that bare CRs in header lines are most likely either
6030 to be mistakes, or people trying to play silly games.
6032 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6033 was being incorrectly given as 18 bytes larger than it should have been.
6034 This is a VOB (very old bug).
6036 60. This may never have affected anything current, but just in case it has:
6037 When the local host is found other than at the start of a list of hosts,
6038 the local host, those with the same MX, and any that follow, are discarded.
6039 When the list in question was part of a longer list of hosts, the following
6040 hosts (not currently being processed) were also being discarded. This no
6041 longer happens. I'm not sure if this situation could ever has previously
6044 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6047 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6048 header. This is something that is recommended in a new Internet Draft, and
6049 is something that is documented as being done by Sendmail. There are two
6050 possible values. For messages generated by the autoreply transport, Exim
6053 Auto-Submitted: auto-replied
6055 whereas for all other generated messages (e.g. bounces) it adds
6057 Auto-Submitted: auto-generated
6059 63. The "personal" condition in filters now includes a test for the
6060 Auto-Submitted: header. If it contains the string "auto-" the message it
6061 not considered personal.
6063 64. Added rcpt_include_affixes as a generic transport option.
6065 65. Added queue_only_override (default true).
6067 66. Added the syslog_duplication option.
6069 67. If what should have been the first header line of a message consisted of
6070 a space followed by a colon, Exim was mis-interpreting it as a header line.
6071 It isn't of course - it is syntactically invalid and should therefore be
6072 treated as the start of the message body. The misbehaviour could have
6073 caused a number of strange effects, including loss of data in subsequent
6074 header lines, and spool format errors.
6076 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6077 client host had authenticated. This control can now be exercised by an ACL
6078 for more flexibility.
6080 69. By default, callouts do not happen when testing with -bh. There is now a
6081 variant, -bhc, which does actually run the callout code, including
6082 consulting and updating the callout cache.
6084 70. Added support for saslauthd authentication, courtesy of Alexander
6087 71. If statvfs() failed on the spool or log directories while checking their
6088 size for availability, Exim confusingly gave the error "space shortage".
6089 Furthermore, in debugging mode it crashed with a floating point exception.
6090 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6091 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6092 serious problem, Exim now writes to the main and panic logs when this
6093 happens, with details of the failure. It then refuses to accept the
6094 incoming message, giving the message "spool directory problem" or "log
6095 directory problem" with a 421 code for SMTP messages.
6097 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6098 0, 1, and 2 exist, because some OS complain for execs without them (see
6099 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6100 descriptors. However, the code omitted to check that the open succeeded,
6101 causing mysterious errors if for some reason the permissions on /dev/null
6102 got screwed. Now Exim writes a message to the main and panic logs, and
6103 bombs out if it can't open /dev/null.
6105 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6106 interact so that it is all more flexible. It is supposed to remain
6107 backwards compatible. Also added extra_local_interfaces.
6109 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6110 to bomb out with an assertion failure - to the client this appears as a
6111 connection drop. This problem occurs in the part of the code that was taken
6112 from the Samba project. Fortunately, the assertion is in a very simple
6113 function, so I have fixed this by reproducing the function inline in the
6114 one place where it is called, and arranging for authentication to fail
6115 instead of killing the process with assert().
6117 75. The SPA client code was not working when the server requested OEM rather
6118 than Unicode encoding.
6120 76. Added code to make require_files with a specific uid setting more usable in
6121 the case where statting the file as root fails - usually a non-root-mounted
6122 NFS file system. When this happens and the failure is EACCES, Exim now
6123 forks a subprocess and does the per-uid checking as the relevant uid.
6125 77. Added process_log_path.
6127 78. If log_file_path was not explicitly set, a setting of check_log_space or
6128 check_log_inodes was ignored.
6130 79. If a space check for the spool or log partitions fails, the incident is now
6131 logged. Of course, in the latter case the data may get lost...
6133 80. Added the %p formatting code to string_format() so that it can be used to
6134 print addresses in debug_print(). Adjusted all the address printing in the
6135 debugging in store.c to use %p rather than %d.
6137 81. There was a concern that a line of code in smtp_in.c could overflow a
6138 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6139 initially expressed, the concern was not well-founded, because trailing
6140 spaces are removed early. However, if the trailing spaces were followed by
6141 a NULL, they did not get removed, so the overflow was possible. Two fixes
6144 (a) I re-wrote the offending code in a cleaner fashion.
6145 (b) If an incoming SMTP command contains a NULL character, it is rejected
6148 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6149 runs initgroups(), so that if the Exim user is in any additional groups,
6150 they will be used during message reception.
6156 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob