1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
15 TL/02 The BSD's have an arc4random API. One of the functions to induce
16 adding randomness was arc4random_stir(), but it has been removed in
17 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
18 function when detected.
20 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
21 cause callback expansion.
23 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
24 syntax errors in an expansion can be treated as a string instead of
25 logging or causing an error, due to the internal use of bool_lax
26 instead of bool when processing it.
28 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
29 server certificates when making smtp deliveries.
31 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
33 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
35 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
37 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
38 Merged patch from Sebastian Wiedenroth.
40 JH/05 Fix results-pipe from transport process. Several recipients, combined
41 with certificate use, exposed issues where response data items split
42 over buffer boundaries were not parsed properly. This eventually
43 resulted in duplicates being sent. This issue only became common enough
44 to notice due to the introduction of conection certificate information,
45 the item size being so much larger. Found and fixed by Wolfgang Breyha.
47 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
48 size buffer was used, resulting in syntax errors when an expansion
51 JH/07 Add support for directories of certificates when compiled with a GnuTLS
52 version 3.3.6 or later.
54 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
55 is EXPERIMENTAL_EVENT, the main-configuration and transport options
56 both become "event_action", the variables become $event_name, $event_data
57 and $event_defer_errno. There is a new variable $verify_mode, usable in
58 routers, transports and related events. The tls:cert event is now also
59 raised for inbound connections, if the main configuration event_action
62 TL/06 In test suite, disable OCSP for old versions of openssl which contained
63 early OCSP support, but no stapling (appears to be less than 1.0.0).
65 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
66 server certificate names available under the smtp transport option
67 "tls_verify_cert_hostname" now do not permit multi-component wildcard
70 JH/10 Time-related extraction expansions from certificates now use the main
71 option "timezone" setting for output formatting, and are consistent
72 between OpenSSL and GnuTLS compilations. Bug 1541.
77 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
78 checkers that were complaining about end of non-void function with no
81 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
82 This was a regression intruduced in 4.83 by another bugfix.
84 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
86 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
87 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
93 TF/01 Correctly close the server side of TLS when forking for delivery.
95 When a message was received over SMTP with TLS, Exim failed to clear up
96 the incoming connection properly after forking off the child process to
97 deliver the message. In some situations the subsequent outgoing
98 delivery connection happened to have the same fd number as the incoming
99 connection previously had. Exim would try to use TLS and fail, logging
100 a "Bad file descriptor" error.
102 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
103 utilities have not been installed.
105 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
106 temporary space as the ACL may create new global variables.
108 TL/01 LDAP support uses per connection or global context settings, depending
109 upon the detected version of the libraries at build time.
111 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
112 to extract and use the src ip:port in logging and expansions as if it
113 were a direct connection from the outside internet. PPv2 support was
114 updated based on HAProxy spec change in May 2014.
116 JH/02 Add ${listextract {number}{list}{success}{fail}}.
118 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
119 Properly escape header and check for NULL return.
121 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
124 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
126 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
127 characters in header names, implemented as a verify condition.
128 Contributed by Michael Fischer v. Mollard.
130 TL/05 Rename SPF condition results err_perm and err_temp to standardized
131 results permerror and temperror. Previous values are deprecated but
132 still accepted. In a future release, err_perm and err_temp will be
133 completely removed, which will be a backward incompatibility if the
134 ACL tests for either of these two old results. Patch contributed by
135 user bes-internal on the mailing list.
137 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
139 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
140 selectors, in both main and reject logs.
142 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
145 JH/07 Add malware type "sock" for talking to simple daemon.
147 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
149 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
150 routers/transports under cutthrough routing.
152 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
153 numbers. Touch up "bool" conditional to keep the same definition.
155 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
157 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
159 JH/12 Expand items in router/transport headers_add or headers_remove lists
160 individually rather than the list as a whole. Bug 1452.
162 Required for reasonable handling of multiple headers_ options when
163 they may be empty; requires that headers_remove items with embedded
164 colons must have them doubled (or the list-separator changed).
166 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
167 view the policy declared in the DMARC record. Currently, $dmarc_status
168 is a combined value of both the record presence and the result of the
171 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
173 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
174 dnslookup router and the smtp transport (applying to the forward
177 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
178 of ldap servers used for a specific lookup. Patch provided by Heiko
181 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
182 New variable $lookup_dnssec_authenticated for observability.
184 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
185 Patch submitted by Lars Timman.
187 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
189 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
190 Requires trusted mode and valid format message id, aborts otherwise.
191 Patch contributed by Heiko Schlichting.
193 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
194 certextract with support for various fields. Bug 1358.
196 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
197 is requested by default, modifiable by smtp transport option
200 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
201 operate on certificate variables to give certificate fingerprints
202 Also new ${sha256:cert_variable}.
204 JH/23 The PRDR feature is moved from being Experimental into the mainline.
206 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
207 Christian Aistleitner.
209 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
211 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
212 file. Patch from Wolfgang Breyha.
214 JH/25 Expand the coverage of the delivery $host and $host_address to
215 client authenticators run in verify callout. Bug 1476.
217 JH/26 Port service names are now accepted for tls_on_connect_ports, to
218 align with daemon_smtp_ports. Bug 72.
220 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
221 support and error reporting did not work properly.
223 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
224 and is readable. Patch from Andrew Colin Kissa.
226 TL/14 Enhance documentation of ${run expansion and how it parses the
227 commandline after expansion, particularly in the case when an
228 unquoted variable expansion results in an empty value.
230 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
232 PP/02 Fix internal collision of T_APL on systems which support RFC3123
233 by renaming away from it. Addresses GH issue 15, reported by
236 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
238 TL/15 SECURITY: prevent double expansion in math comparison functions
239 (can expand unsanitized data). Not remotely exploitable.
246 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
248 PP/02 Make -n do something, by making it not do something.
249 When combined with -bP, the name of an option is not output.
251 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
254 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
255 $sender_host_name and config options to manage this, and basic check
258 PP/05 DSCP support for outbound connections and control modifier for inbound.
260 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
261 (Only plugin which currently uses this is kerberos4, which nobody should
262 be using, but we should make it available and other future plugins might
263 conceivably use it, even though it would break NAT; stuff *should* be
264 using channel bindings instead).
266 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
267 name; added for Sendmail compatibility; requires admin caller.
268 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
269 just ignore it); requires trusted caller.
270 Also parse but ignore: -Ac -Am -X<logfile>
273 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
275 TL/02 Add +smtp_confirmation as a default logging option.
277 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
278 Patch by Magnus Holmgren from 2007-02-20.
280 TL/04 Bugzilla 1281 - Spec typo.
281 Bugzilla 1283 - Spec typo.
282 Bugzilla 1290 - Spec grammar fixes.
284 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
286 TL/06 Add Experimental DMARC support using libopendmarc libraries.
288 TL/07 Fix an out of order global option causing a segfault. Reported to dev
289 mailing list by by Dmitry Isaikin.
291 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
293 JH/02 Support "G" suffix to numbers in ${if comparisons.
295 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
297 NM/01 Bugzilla 1197 - Spec typo
298 Bugzilla 1196 - Spec examples corrections
300 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
302 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
303 gnutls_enable_pkcs11, but renamed to more accurately indicate its
306 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
307 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
309 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
310 "acl {{name}{arg}...}", and optional args on acl condition
313 JH/05 Permit multiple router/transport headers_add/remove lines.
315 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
317 JH/07 Avoid using a waiting database for a single-message-only transport.
318 Performance patch from Paul Fisher. Bugzilla 1262.
320 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
323 JH/09 Add $headers_added variable, with content from use of ACL modifier
324 add_header (but not yet added to the message). Bugzilla 199.
326 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
327 Pulled from Bugzilla 817 by Wolfgang Breyha.
329 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
331 (nb: this is the same fix as in Exim 4.80.1)
333 JH/11 Add A= logging on delivery lines, and a client_set_id option on
336 JH/12 Add optional authenticated_sender logging to A= and a log_selector
339 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
341 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
342 advertise SMTP AUTH mechanism to us, instead of a generic
343 protocol violation error. Also, make Exim more robust to bad
344 data from the Dovecot auth socket.
346 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
348 When a queue runner is handling a message, Exim first routes the
349 recipient addresses, during which it prunes them based on the retry
350 hints database. After that it attempts to deliver the message to
351 any remaining recipients. It then updates the hints database using
354 So if a recipient address works intermittently, it can get repeatedly
355 deferred at routing time. The retry hints record remains fresh so the
356 address never reaches the final cutoff time.
358 This is a fairly common occurrence when a user is bumping up against
359 their storage quota. Exim had some logic in its local delivery code
360 to deal with this. However it did not apply to per-recipient defers
361 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
363 This change adds a proper retry rule check during routing so that the
364 final cutoff time is checked against the message's age. We only do
365 this check if there is an address retry record and there is not a
366 domain retry record; this implies that previous attempts to handle
367 the address had the retry_use_local_parts option turned on. We use
368 this as an approximation for the destination being like a local
369 delivery, as in LMTP.
371 I suspect this new check makes the old local delivery cutoff check
372 redundant, but I have not verified this so I left the code in place.
374 TF/02 Correct gecos expansion when From: is a prefix of the username.
376 Test 0254 submits a message to Exim with the header
380 When I ran the test suite under the user fanf2, Exim expanded
381 the header to contain my full name, whereas it should have added
382 a Resent-Sender: header. It erroneously treats any prefix of the
383 username as equal to the username.
385 This change corrects that bug.
387 GF/01 DCC debug and logging tidyup
388 Error conditions log to paniclog rather than rejectlog.
389 Debug lines prefixed by "DCC: " to remove any ambiguity.
391 TF/03 Avoid unnecessary rebuilds of lookup-related code.
393 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
394 Bug spotted by Jeremy Harris; was flawed since initial commit.
395 Would have resulted in OCSP responses post-SNI triggering an Exim
396 NULL dereference and crash.
398 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
400 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
401 Bug detection, analysis and fix by Samuel Thibault.
402 Bugzilla 1331, Debian bug #698092.
404 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
406 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
407 Server implementation by Todd Lyons, client by JH.
408 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
409 config variable "prdr_enable" controls whether the server
410 advertises the facility. If the client requests PRDR a new
411 acl_data_smtp_prdr ACL is called once for each recipient, after
412 the body content is received and before the acl_smtp_data ACL.
413 The client is controlled by bolth of: a hosts_try_prdr option
414 on the smtp transport, and the server advertisement.
415 Default client logging of deliveries and rejections involving
416 PRDR are flagged with the string "PRDR".
418 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
419 fclose(). Diagnosis by Todd Lyons.
421 PP/17 Update configure.default to handle IPv6 localhost better.
422 Patch by Alain Williams (plus minor tweaks).
425 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
426 This is now consistent with GnuTLS, and is now documented: the
427 previous undocumented portable approach to treating the option as
428 unset was to force an expansion failure. That still works, and
429 an empty string is now equivalent.
431 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
432 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
433 not performing validation itself.
435 PP/20 Added force_command boolean option to pipe transport.
436 Patch from Nick Koston, of cPanel Inc.
438 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
441 TF/04 Added udpsend ACL modifer and hexquote expansion operator
443 PP/21 Fix eximon continuous updating with timestamped log-files.
444 Broken in a format-string cleanup in 4.80, missed when I repaired the
445 other false fix of the same issue.
446 Report and fix from Heiko Schlichting.
449 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
450 Report from Prashanth Katuri.
452 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
453 It's SecureTransport, so affects any MacOS clients which use the
454 system-integrated TLS libraries, including email clients.
456 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
457 using a MIME ACL for non-SMTP local injection.
458 Report and assistance in diagnosis by Warren Baker.
460 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
462 JH/16 Fix comparisons for 64b. Bugzilla 1385.
464 TL/09 Add expansion variable $authenticated_fail_id to keep track of
465 last id that failed so it may be referenced in subsequent ACL's.
467 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
470 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
471 ldap library initialization, allowing self-signed CA's to be
472 used. Also properly sets require_cert option later in code by
473 using NULL (global ldap config) instead of ldap handle (per
474 session). Bug diagnosis and testing by alxgomz.
476 TL/12 Enhanced documentation in the ratelimit.pl script provided in
477 the src/util/ subdirectory.
479 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
480 renamed to Transport Post Delivery Action by Jeremy Harris, as
483 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
484 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
485 redis_servers = needs to be configured which will be used by the redis
486 lookup. Patch from Warren Baker, of The Packet Hub.
488 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
490 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
491 hostname or reverse DNS when processing a host list. Used suggestions
492 from multiple comments on this bug.
494 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
496 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
497 Missed a few lines, added it to make the runtest require no keyboard
500 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
501 contains upper case chars. Make router use caseful_local_part.
503 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
504 support when GnuTLS has been built with p11-kit.
510 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
512 This, or similar/improved, will also be change PP/11 of 4.82.
518 PP/01 Handle short writes when writing local log-files.
519 In practice, only affects FreeBSD (8 onwards).
520 Bugzilla 1053, with thanks to Dmitry Isaikin.
522 NM/01 Bugzilla 949 - Documentation tweak
524 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
527 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
529 PP/02 Implemented gsasl authenticator.
531 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
533 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
534 `pkg-config foo` for cflags/libs.
536 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
537 with rest of GSASL and with heimdal_gssapi.
539 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
540 `pkg-config foo` for cflags/libs for the TLS implementation.
542 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
543 properties get this fed in as external SSF. A number of robustness
544 and debugging improvements to the cyrus_sasl authenticator.
546 PP/08 cyrus_sasl server now expands the server_realm option.
548 PP/09 Bugzilla 1214 - Log authentication information in reject log.
549 Patch by Jeremy Harris.
551 PP/10 Added dbmjz lookup type.
553 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
555 PP/12 MAIL args handles TAB as well as SP, for better interop with
556 non-compliant senders.
557 Analysis and variant patch by Todd Lyons.
559 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
560 Bug report from Lars Müller <lars@samba.org> (via SUSE),
561 Patch from Dirk Mueller <dmueller@suse.com>
563 PP/13 tls_peerdn now print-escaped for spool files.
564 Observed some $tls_peerdn in wild which contained \n, which resulted
565 in spool file corruption.
567 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
568 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
569 or write after TLS renegotiation, which otherwise led to messages
572 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
573 as a tracking header (ie: a signed header comes before the signature).
574 Patch from Wolfgang Breyha.
576 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
577 comma-sep list; embedded commas doubled.
579 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
581 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
583 Report and patch from Dmitry Banschikov.
585 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
586 Removed SSL_clear() after SSL_new() which led to protocol negotiation
587 failures. We appear to now support TLS1.1+ with Exim.
589 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
590 lets Exim select keys and certificates based upon TLS SNI from client.
591 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
592 before an outbound SMTP session. New log_selector, +tls_sni.
594 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
595 NULL dereference. Report and patch from Alun Jones.
597 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
598 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
599 Not seeing resolver debug output on NetBSD, but suspect this is a
600 resolver implementation change.
602 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
603 Left warnings. Added "eximon gdb" invocation mode.
605 PP/21 Defaulting "accept_8bitmime" to true, not false.
607 PP/22 Added -bw for inetd wait mode support.
609 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
610 locate the relevant includes and libraries. Made this the default.
612 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
613 Bugzilla 1246, report and most of solution from Tomasz Kusy.
615 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
616 This may cause build issues on older platforms.
618 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
619 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
620 gnutls_require_mac & gnutls_require_protocols (no longer supported).
621 Added SNI support via GnuTLS too.
622 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
624 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
626 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
627 Applied second patch from Janne, implementing suggestion to default
628 multiple-strings-in-record handling to match SPF spec.
630 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
632 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
633 read-only, out of scope).
634 Patch from Wolfgang Breyha, report from Stuart Northfield.
636 PP/29 Fix three issues highlighted by clang analyser static analysis.
637 Only crash-plausible issue would require the Cambridge-specific
638 iplookup router and a misconfiguration.
639 Report from Marcin Mirosław.
641 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
643 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
644 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
645 As part of this, removing so much warning spew let me fix some minor
646 real issues in debug logging.
648 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
649 assignment on my part. Fixed.
651 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
652 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
653 Janne Snabb (who went above and beyond: thank you).
655 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
656 string otherwise requires a connection and a bunch more work and it's
657 relatively easy to get wrong. Should also expose TLS library linkage
660 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
661 64-bit ${eval} (JH/03).
663 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
664 GNU libc to support some of the 64-bit stuff, should not lead to
665 conflicts. Defined before os.h is pulled in, so if a given platform
666 needs to override this, it can.
668 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
669 protection layer was required, which is not implemented.
670 Bugzilla 1254, patch from Wolfgang Breyha.
672 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
673 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
674 tls_dhparam take prime identifiers. Also unbreak combination of
675 OpenSSL+DH_params+TLSSNI.
677 PP/39 Disable SSLv2 by default in OpenSSL support.
683 PP/01 Solaris build fix for Oracle's LDAP libraries.
684 Bugzilla 1109, patch from Stephen Usher.
686 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
688 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
691 TF/02 Fix a couple more cases where we did not log the error message
692 when unlink() failed. See also change 4.74-TF/03.
694 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
695 lock up or crash if it happened to be inside a call to libc when it
696 got a SIGUSR1 from exiwhat.
698 The SIGUSR1 handler appends the current process status to the process
699 log which is later printed by exiwhat. It used to use the general
700 purpose logging code to do this, but several functions it calls are
701 not safe for signals.
703 The new output code in the SIGUSR1 handler is specific to the process
704 log, and simple enough that it's easy to inspect for signal safety.
705 Removing some special cases also simplifies the general logging code.
706 Removing the spurious timestamps from the process log simplifies
709 TF/04 Improved ratelimit ACL condition.
711 The /noupdate option has been deprecated in favour of /readonly which
712 has clearer semantics. The /leaky, /strict, and /readonly update modes
713 are mutually exclusive. The update mode is no longer included in the
714 database key; it just determines when the database is updated. (This
715 means that when you upgrde Exim will forget old rate measurements.)
717 Exim now checks that the per_* options are used with an update mode that
718 makes sense for the current ACL. For example, when Exim is processing a
719 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
720 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
721 must specify per_mail/readonly. If you omit the update mode it defaults to
722 /leaky where that makes sense (as before) or /readonly where required.
724 The /noupdate option is now undocumented but still supported for
725 backwards compatibility. It is equivalent to /readonly except that in
726 ACLs where /readonly is required you may specify /leaky/noupdate or
727 /strict/noupdate which are treated the same as /readonly.
729 A useful new feature is the /count= option. This is a generalization
730 of the per_byte option, so that you can measure the throughput of other
731 aggregate values. For example, the per_byte option is now equivalent
732 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
734 The per_rcpt option has been generalized using the /count= mechanism
735 (though it's more complicated than the per_byte equivalence). When it is
736 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
737 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
738 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
739 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
740 Note that using per_rcpt with a non-readonly update mode in more than
741 one ACL will cause the recipients to be double-counted. (The per_mail
742 and per_byte options don't have this problem.)
744 The handling of very low rates has changed slightly. If the computed rate
745 is less than the event's count (usually one) then this event is the first
746 after a long gap. In this case the rate is set to the same as this event's
747 count, so that the first message of a spam run is counted properly.
749 The major new feature is a mechanism for counting the rate of unique
750 events. The new per_addr option counts the number of different
751 recipients that someone has sent messages to in the last time period. It
752 behaves like per_rcpt if all the recipient addresses are different, but
753 duplicate recipient addresses do not increase the measured rate. Like
754 the /count= option this is a general mechanism, so the per_addr option
755 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
756 example, measure the rate that a client uses different sender addresses
757 with the options per_mail/unique=$sender_address. There are further
758 details in the main documentation.
760 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
762 TF/06 Removed a few PCRE remnants.
764 TF/07 Automatically extract Exim's version number from tags in the git
765 repository when doing development or release builds.
767 PP/02 Raise smtp_cmd_buffer_size to 16kB.
768 Bugzilla 879. Patch from Paul Fisher.
770 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
771 Heavily based on revision 40f9a89a from Simon Arlott's tree.
774 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
776 PP/05 Variable $av_failed, true if the AV scanner deferred.
777 Bugzilla 1078. Patch from John Horne.
779 PP/06 Stop make process more reliably on build failure.
780 Bugzilla 1087. Patch from Heiko Schlittermann.
782 PP/07 Make maildir_use_size_file an _expandable_ boolean.
783 Bugzilla 1089. Patch from Heiko Schlittermann.
785 PP/08 Handle ${run} returning more data than OS pipe buffer size.
786 Bugzilla 1131. Patch from Holger Weiß.
788 PP/09 Handle IPv6 addresses with SPF.
789 Bugzilla 860. Patch from Wolfgang Breyha.
791 PP/10 GnuTLS: support TLS 1.2 & 1.1.
793 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
796 PP/11 match_* no longer expand right-hand-side by default.
797 New compile-time build option, EXPAND_LISTMATCH_RHS.
798 New expansion conditions, "inlist", "inlisti".
800 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
802 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
804 PP/14 fix log_write() format string regression from TF/03.
805 Bugzilla 1152. Patch from Dmitry Isaikin.
811 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
813 PP/02 Harmonised TLS library version reporting; only show if debugging.
814 Layout now matches that introduced for other libraries in 4.74 PP/03.
816 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
818 PP/04 New "dns_use_edns0" global option.
820 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
823 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
824 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
826 TK/01 Updated PolarSSL code to 0.14.2.
827 Bugzilla 1097. Patch from Andreas Metzler.
829 PP/07 Catch divide-by-zero in ${eval:...}.
832 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
835 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
836 format-string attack -- SECURITY: remote arbitrary code execution.
838 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
839 time unintentionally subject to list matching rules, letting the header
840 cause arbitrary Exim lookups (of items which can occur in lists, *not*
841 arbitrary string expansion). This allowed for information disclosure.
843 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
844 INT_MIN/-1 -- value coerced to INT_MAX.
850 NM/01 Workround for PCRE version dependency in version reporting
853 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
854 This fixes portability to compilers other than gcc, notably
855 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
857 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
858 makefiles for portability to HP-UX and POSIX correctness.
860 PP/01 Permit LOOKUP_foo enabling on the make command-line.
861 Also via indented variable definition in the Makefile.
862 (Debugging by Oliver Heesakkers).
864 PP/02 Restore caching of spamd results with expanded spamd_address.
865 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
867 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
868 Improves build reliability. Fix from: Frank Elsner
870 NM/02 Fix wide character breakage in the rfc2047 coding
871 Fixes bug 1064. Patch from Andrey N. Oktyabrski
873 NM/03 Allow underscore in dnslist lookups
874 Fixes bug 1026. Patch from Graeme Fowler
876 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
877 Code patches from Adam Ciarcinski of NetBSD.
879 NM/04 Fixed exiqgrep to cope with mailq missing size issue
882 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
883 is logged, to avoid truncation. Patch from John Horne.
885 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
886 Patch from Jakob Hirsch.
888 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
889 SQL string expansion failure details.
890 Patch from Andrey Oktyabrski.
892 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
893 Patch from Simon Arlott.
895 PP/09 New lookups functionality failed to compile on old gcc which rejects
896 extern declarations in function scope.
897 Patch from Oliver Fleischmann
899 PP/10 Use sig_atomic_t for flags set from signal handlers.
900 Check getgroups() return and improve debugging.
901 Fixed developed for diagnosis in bug 927 (which turned out to be
904 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
905 Patch from Mark Zealey.
907 PP/12 Bugzilla 1056: Improved spamd server selection.
908 Patch from Mark Zealey.
910 PP/13 Bugzilla 1086: Deal with maildir quota file races.
911 Based on patch from Heiko Schlittermann.
913 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
914 Patch from Uwe Doering, sign-off by Michael Haardt.
916 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
917 variable declaration deep within a block. Bug and patch from
920 PP/15 lookups-Makefile IRIX compatibilty coercion.
922 PP/16 Make DISABLE_DKIM build knob functional.
924 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
925 Patch by Simon Arlott
927 TF/03 Fix valgrind.h portability to C89 compilers that do not support
928 variable argument macros. Our copy now differs from upstream.
934 TF/01 Failure to get a lock on a hints database can have serious
935 consequences so log it to the panic log.
937 TF/02 Log LMTP confirmation messages in the same way as SMTP,
938 controlled using the smtp_confirmation log selector.
940 TF/03 Include the error message when we fail to unlink a spool file.
942 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
943 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
944 for maintaining out-of-tree patches for some time.
946 PP/01 Bugzilla 139: Documentation and portability issues.
947 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
948 Handle per-OS dynamic-module compilation flags.
950 PP/02 Let /dev/null have normal permissions.
951 The 4.73 fixes were a little too stringent and complained about the
952 permissions on /dev/null. Exempt it from some checks.
953 Reported by Andreas M. Kirchwitz.
955 PP/03 Report version information for many libraries, including
956 Exim version information for dynamically loaded libraries. Created
957 version.h, now support a version extension string for distributors
958 who patch heavily. Dynamic module ABI change.
960 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
961 privilege escalation vulnerability whereby the Exim run-time user
962 can cause root to append content of the attacker's choosing to
965 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
968 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
969 If dropping privileges for untrusted macros, we disabled normal logging
970 on the basis that it would fail; for the Exim run-time user, this is not
971 the case, and it resulted in successful deliveries going unlogged.
972 Fixed. Reported by Andreas Metzler.
978 PP/01 Date: & Message-Id: revert to normally being appended to a message,
979 only prepend for the Resent-* case. Fixes regression introduced in
980 Exim 4.70 by NM/22 for Bugzilla 607.
982 PP/02 Include check_rfc2047_length in configure.default because we're seeing
983 increasing numbers of administrators be bitten by this.
985 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
987 PP/03 Bugzilla 994: added openssl_options main configuration option.
989 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
991 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
993 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
995 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
996 without a peer certificate, leading to a segfault because of an
997 assumption that peers always have certificates. Be a little more
998 paranoid. Problem reported by Martin Tscholak.
1000 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1001 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1002 NB: ClamAV planning to remove STREAM in "middle of 2010".
1003 CL also introduces -bmalware, various -d+acl logging additions and
1004 more caution in buffer sizes.
1006 PP/09 Implemented reverse_ip expansion operator.
1008 PP/10 Bugzilla 937: provide a "debug" ACL control.
1010 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1012 PP/12 Bugzilla 973: Implement --version.
1014 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1016 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1018 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1020 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1021 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1022 ignore trailing whitespace.
1024 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1026 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1029 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1030 Notification from Dr Andrew Aitchison.
1032 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1033 ExtendedDetectionInfo response format.
1034 Notification from John Horne.
1036 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1039 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1040 XSL and documented dependency on system catalogs, with examples of how
1043 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1046 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1047 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1048 configuration file which is writeable by the Exim user or group.
1050 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1051 of configuration files to cover files specified with the -C option if
1052 they are going to be used with root privileges, not just the default
1055 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1056 option (effectively making it always true).
1058 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1059 files to be used while preserving root privileges.
1061 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1062 that rogue child processes cannot use them.
1064 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1065 run-time user, instead of root.
1067 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1068 Exim run-time user without dropping privileges.
1070 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1071 result string, instead of calling string_vformat() twice with the same
1074 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1075 for other users. Others should always drop root privileges if they use
1076 -C on the command line, even for a whitelisted configure file.
1078 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1080 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1086 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1087 $data_path, and $header_path variables; fixed documentation bugs and
1090 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1091 exipick to access non-standard spools, including the "frozen" queue
1094 NM/01 Bugzilla 965: Support mysql stored procedures.
1095 Patch from Alain Williams
1097 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1099 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1100 Patch from Andreas Metzler
1102 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1103 Patch from Kirill Miazine
1105 NM/05 Bugzilla 671: Added umask to procmail example.
1107 JJ/03 installed exipick 20100323.0, fixing doc bug
1109 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1110 directory. Notification and patch from Dan Rosenberg.
1112 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1114 TK/02 Improve log output when DKIM signing operation fails.
1116 MH/01 Treat the transport option dkim_domain as a colon separated
1117 list, not as a single string, and sign the message with each element,
1118 omitting multiple occurences of the same signer.
1120 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1121 Bugzilla 985, 986. Patch by Simon Arlott
1123 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1124 Patch by Simon Arlott
1126 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1127 MBX locking. Notification from Dan Rosenberg.
1133 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1135 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1137 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1139 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1141 NM/04 Bugzilla 915: Flag broken perl installation during build.
1147 TK/01 Added patch by Johannes Berg that expands the main option
1148 "spamd_address" if it starts with a dollar sign.
1150 TK/02 Write list of recipients to X-Envelope-Sender header when building
1151 the mbox-format spool file for content scanning (suggested by Jakob
1154 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1155 (http://www.dcc-servers.net/) support via dccifd. Activated by
1156 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1158 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1159 by Mark Daniel Reidel <mr@df.eu>.
1161 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1162 When building exim an external PCRE library is now needed -
1163 PCRE is a system library on the majority of modern systems.
1164 See entry on PCRE_LIBS in EDITME file.
1166 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1167 conversation. Added nologin parameter to request.
1168 Patch contributed by Kirill Miazine.
1170 TF/01 Do not log submission mode rewrites if they do not change the address.
1172 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1174 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1175 log files in place. Contributed by Roberto Lima.
1177 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1179 TF/03 Bugzilla 615: When checking the local_parts router precondition
1180 after a local_part_suffix or local_part_prefix option, Exim now
1181 does not use the address's named list lookup cache, since this
1182 contains cached lookups for the whole local part.
1184 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1185 Robert Millan. Documentation is in experimental-spec.txt.
1187 TF/04 Bugzilla 668: Fix parallel build (make -j).
1189 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1191 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1192 Patch provided by Jan Srzednicki.
1194 TF/05 Leading white space used to be stripped from $spam_report which
1195 wrecked the formatting. Now it is preserved.
1197 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1198 that they are available at delivery time.
1200 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1202 TF/08 TLS error reporting now respects the incoming_interface and
1203 incoming_port log selectors.
1205 TF/09 Produce a more useful error message if an SMTP transport's hosts
1206 setting expands to an empty string.
1208 NM/06 Bugzilla 744: EXPN did not work under TLS.
1209 Patch provided by Phil Pennock.
1211 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1212 Patch provided by Richard Godbee.
1214 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1215 acl_smtp_notquit, added index entry.
1217 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1218 Patch provided by Eugene Bujak.
1220 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1221 accept(). Patch provided by Maxim Dounin.
1223 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1224 Patch provided by Phil Pennock.
1226 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1228 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1229 Patch provided by Brad "anomie" Jorsch.
1231 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1232 Patch provided by Dean Brooks.
1234 TK/05 Add native DKIM support (does not depend on external libraries).
1236 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1237 Patch provided by Graeme Fowler.
1239 NM/16 Bugzilla 851: Documentation example syntax fix.
1241 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1243 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1246 NM/19 Bugzilla 745: TLS version reporting.
1247 Patch provided by Phil Pennock.
1249 NM/20 Bugzilla 167: bool: condition support.
1250 Patch provided by Phil Pennock.
1252 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1253 clients. Patch provided by Phil Pennock.
1255 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1256 Patch provided by Brad "anomie" Jorsch.
1258 NM/23 Bugzilla 687: Fix misparses in eximstats.
1259 Patch provided by Heiko Schlittermann.
1261 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1262 Patch provided by Heiko Schlittermann.
1264 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1265 plus update to original patch.
1267 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1269 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1270 Patch provided by David Brownlee.
1272 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1274 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1276 NM/30 Bugzilla 888: TLS documentation bugfixes.
1278 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1280 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1281 Unlike the original bugzilla I have changed all shell scripts in src tree.
1283 NM/33 Bugzilla 898: Transport filter timeout fix.
1284 Patch by Todd Rinaldo.
1286 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1287 Patch by Serge Demonchaux.
1289 NM/35 Bugzilla 39: Base64 decode bug fixes.
1290 Patch by Jakob Hirsch.
1292 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1294 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1296 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1298 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1304 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1305 ALT-N's libdkim that I have put here:
1306 http://duncanthrax.net/exim-experimental/
1308 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1309 were called 'true' and it seems that C99 defines that as a reserved
1310 keyword to be used with 'bool' variable types. That means you could
1311 not include C99-style headers which use bools without triggering
1312 build errors in sieve.c.
1314 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1315 as mailq or other aliases. Changed the --help handling significantly
1316 to do whats expected. exim_usage() emits usage/help information.
1318 SC/01 Added the -bylocaldomain option to eximstats.
1320 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1322 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1324 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1330 PH/01 Another patch from the Sieve maintainer.
1332 PH/02 When an IPv6 address is converted to a string for single-key lookup
1333 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1334 dots are used instead of colons so that keys in lsearch files need not
1335 contain colons. This was done some time before quoting was made available
1336 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1337 (notated using the quote facility) so as to distinguish them from IPv4
1338 keys. This meant that lookups for IP addresses in host lists did not work
1339 for iplsearch lookups.
1341 This has been fixed by arranging for IPv6 addresses to be expressed with
1342 colons if the lookup type is iplsearch. This is not incompatible, because
1343 previously such lookups could never work.
1345 The situation is now rather anomolous, since one *can* have colons in
1346 ordinary lsearch keys. However, making the change in all cases is
1347 incompatible and would probably break a number of configurations.
1349 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1352 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1353 conversion specification without a maximum field width, thereby enabling
1354 a rogue spamd server to cause a buffer overflow. While nobody in their
1355 right mind would setup Exim to query an untrusted spamd server, an
1356 attacker that gains access to a server running spamd could potentially
1357 exploit this vulnerability to run arbitrary code as the Exim user.
1359 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1360 $primary_hostname instead of what libspf2 thinks the hosts name is.
1362 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1363 a directory entry by the name of the lookup key. Previously, if a
1364 symlink pointed to a non-existing file or a file in a directory that
1365 Exim lacked permissions to read, a lookup for a key matching that
1366 symlink would fail. Now it is enough that a matching directory entry
1367 exists, symlink or not. (Bugzilla 503.)
1369 PH/03 The body_linecount and body_zerocount variables are now exported in the
1372 PH/04 Added the $dnslist_matched variable.
1374 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1375 This means they are set thereafter only if the connection becomes
1378 PH/06 Added the client_condition to authenticators so that some can be skipped
1379 by clients under certain conditions.
1381 PH/07 The error message for a badly-placed control=no_multiline_responses left
1382 "_responses" off the end of the name.
1384 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1386 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1387 (without spaces) instead of just copying the configuration text.
1389 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1391 PH/11 Added $max_received_linelength.
1393 PH/12 Added +ignore_defer and +include_defer to host lists.
1395 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1396 way in which PCRE > 7.0 is built.
1398 PH/14 Implemented queue_only_load_latch.
1400 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1401 MAIL command. The effect was to mangle the value on 64-bit systems.
1403 PH/16 Another patch from the Sieve maintainer.
1405 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1407 PH/18 If a system quota error occurred while trying to create the file for
1408 a maildir delivery, the message "Mailbox is full" was not appended to the
1409 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1410 only to a quota excession during the actual writing of the file.
1412 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1413 characters?) which causes problems in log lines. The DN values are now
1414 passed through string_printing() before being added to log lines.
1416 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1417 and InterBase are left for another time.)
1419 PH/21 Added message_body_newlines option.
1421 PH/22 Guard against possible overflow in moan_check_errorcopy().
1423 PH/23 POSIX allows open() to be a macro; guard against that.
1425 PH/24 If the recipient of an error message contained an @ in the local part
1426 (suitably quoted, of course), incorrect values were put in $domain and
1427 $local_part during the evaluation of errors_copy.
1433 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1434 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1437 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1438 issue a MAIL command.
1440 PH/02 In an ACL statement such as
1442 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1444 if a client was not listed at all, or was listed with a value other than
1445 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1446 the condition was not true (as it should be), so access was not denied.
1447 The bug was that the ! inversion was incorrectly passed on to the second
1448 item. This has been fixed.
1450 PH/03 Added additional dnslists conditions == and =& which are different from
1451 = and & when the dns lookup returns more than one IP address.
1453 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1454 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1456 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1457 FSYNC, which compiles an option called disable_fsync that allows for
1458 bypassing fsync(). The documentation is heavily laced with warnings.
1460 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1462 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1463 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1464 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1465 including adding "make clean"; (3) Added -fPIC when compiling the test
1466 dynamically loaded module, to get rid of a warning.
1468 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1469 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1470 = 0s. The bug is otherwise harmless.
1472 PH/07 There was a bug in the dovecot authenticator such that the value of
1473 $auth1 could be overwritten, and so not correctly preserved, after a
1474 successful authentication. This usually meant that the value preserved by
1475 the server_setid option was incorrect.
1477 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1479 PH/09 Installed PCRE release 7.0.
1481 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1482 run for batched SMTP input. It is now run at the start of every message
1483 in the batch. While fixing this I discovered that the process information
1484 (output by running exiwhat) was not always getting set for -bs and -bS
1485 input. This is fixed, and it now also says "batched" for BSMTP.
1487 PH/11 Added control=no_pipelining.
1489 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1490 patch, slightly modified), and move the expansion of helo_data till after
1491 the connection is made in the smtp transport (so it can use these
1494 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1496 PH/14 Added log_selector = +pid.
1498 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1500 PH/16 Add ${if forany and ${if forall.
1502 PH/17 Added dsn_from option to vary the From: line in DSNs.
1504 PH/18 Flush SMTP output before performing a callout, unless control =
1505 no_callout_flush is set.
1507 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1508 was true (the default) a successful delivery failed to delete the retry
1509 item, thus causing premature timeout of the address. The bug is now
1512 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1514 PH/21 Long custom messages for fakedefer and fakereject are now split up
1515 into multiline reponses in the same way that messages for "deny" and
1516 other ACL rejections are.
1518 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1519 with slight modification.
1521 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1522 draft, changing the syntax and factoring some duplicate code".
1524 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1525 for deliveries of the second and subsequent messages over the same SMTP
1528 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1529 ${reduce, with only minor "tidies".
1531 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1533 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1534 expansion side effects.
1536 PH/27 When a message times out after an over-quota error from an Exim-imposed
1537 quota, the bounce message says "mailbox is full". This message was not
1538 being given when it was a system quota that was exceeded. It now should
1541 MH/03 Made $recipients available in local_scan(). local_scan() already has
1542 better access to the recipient list through recipients_list[], but
1543 $recipients can be useful in postmaster-provided expansion strings.
1545 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1546 in the case of a MAIL command with additional options following the
1547 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1548 were accidentally chopped off.
1550 PH/29 SMTP synchronization checks are implemented when a command is read -
1551 there is a check that no more input is waiting when there shouldn't be
1552 any. However, for some commands, a delay in an ACL can mean that it is
1553 some time before the response is written. In this time, more input might
1554 arrive, invalidly. So now there are extra checks after an ACL has run for
1555 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1556 pipelining has not been advertised.
1558 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1560 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1561 returned in the SMTP response when smtp_return_error_details was false.
1562 This has been fixed.
1564 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1565 instead of the C I/O that was originally supplied, because problems were
1566 reported on Solaris.
1568 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1569 Exim which did not show up earlier: it was assuming that a call to
1570 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1571 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1572 was a macro that became an assignment, so it seemed to work. This has
1573 changed to a proper function call with a void return, hence the compile
1574 error. Exim's code has been fixed.
1576 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1579 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1581 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1583 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1584 because it turns out to be too big or there is a timeout) there is no
1585 "Completed" line in the log. When some messages of this type were
1586 selected by exigrep, they were listed as "not completed". Others were
1587 picked up by some special patterns. I have improved the selection
1588 criteria to be more general.
1590 PH/38 The host_find_failed option in the manualroute router can now be set
1591 to "ignore", to completely ignore a host whose IP address cannot be
1592 found. If all hosts are ignored, the behaviour is controlled by the new
1593 host_all_ignored option.
1595 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1596 homing or because of multiple MX records with /mx) generated more than
1597 one IP address, and the following item turned out to be the local host,
1598 all the secondary addresses of the first item were incorrectly removed
1599 from the list, along with the local host and any following hosts (which
1600 is what is supposed to happen).
1602 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1603 whoever called Exim into the -H file. In the case of the daemon it was
1604 behaving confusingly. When first started, it used values for whoever
1605 started the daemon, but after a SIGHUP it used the Exim user (because it
1606 calls itself on a restart). I have changed the code so that it now always
1609 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1610 message are rejected with the same error (e.g. no authentication or bad
1611 sender address), and a DATA command is nevertheless sent (as can happen
1612 with PIPELINING or a stupid MUA), the error message that was given to the
1613 RCPT commands is included in the rejection of the DATA command. This is
1614 intended to be helpful for MUAs that show only the final error to their
1617 PH/42 Another patch from the Sieve maintainer.
1619 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1620 Eximstats - Fixed some broken HTML links and added missing column headers
1622 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1623 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1625 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1627 PH/43 Yet another patch from the Sieve maintainer.
1629 PH/44 I found a way to check for a TCP/IP connection going away before sending
1630 the response to the final '.' that terminates a message, but only in the
1631 case where the client has not sent further data following the '.'
1632 (unfortunately, this is allowed). However, in many cases there won't be
1633 any further data because there won't be any more messages to send. A call
1634 to select() can be used: if it shows that the input is "ready", there is
1635 either input waiting, or the socket has been closed. An attempt to read
1636 the next input character can distinguish the two cases. Previously, Exim
1637 would have sent an OK response which the client would never have see.
1638 This could lead to message repetition. This fix should cure that, at
1639 least in a lot of common cases.
1641 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1642 advertised in response to EHLO.
1648 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1649 fixed by 4.65/MH/01 (is this a record?) are fixed:
1651 (i) An empty string was always treated as zero by the numeric comparison
1652 operators. This behaviour has been restored.
1654 (ii) It is documented that the numeric comparison operators always treat
1655 their arguments as decimal numbers. This was broken in that numbers
1656 starting with 0 were being interpreted as octal.
1658 While fixing these problems I realized that there was another issue that
1659 hadn't been noticed. Values of message_size_limit (both the global option
1660 and the transport option) were treated as octal if they started with 0.
1661 The documentation was vague. These values are now always treated as
1662 decimal, and I will make that clear in the documentation.
1668 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1669 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1672 MH/01 Don't check that the operands of numeric comparison operators are
1673 integers when their expansion is in "skipping" mode (fixes bug
1674 introduced by 4.64-PH/07).
1676 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1677 child addresses, Exim now panics and dies. Previously, because the count
1678 is held in a short int, deliveries were likely to be lost. As such a
1679 large number of recipients for a single message is ridiculous
1680 (performance will be very, very poor), I have chosen to impose a limit
1681 rather than extend the field.
1687 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1688 leftover -K file (the existence of which was triggered by #402).
1689 While we were at it, introduced process PID as part of the -K
1690 filename. This should rule out race conditions when creating
1693 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1694 processing considerably. Previous code took too long for large mails,
1695 triggering a timeout which in turn triggers #401.
1697 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1698 in the DK code in transports.c. sendfile() is not really portable,
1699 hence the _LINUX specificness.
1701 TF/01 In the add_headers option to the mail command in an Exim filter,
1702 there was a bug that Exim would claim a syntax error in any
1703 header after the first one which had an odd number of characters
1706 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1707 callout verification, Exim cached a "reject" for the entire domain. This
1708 is correct for most verifications, but it is not correct for a recipient
1709 verification with use_sender or use_postmaster set, because in that case
1710 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1711 case of MAIL FROM:<> rejection from other early rejections (e.g.
1712 rejection of HELO). When verifying a recipient using a non-null MAIL
1713 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1714 Whatever the result of the callout, the value of the domain cache is
1715 left unchanged (for any other kind of callout, getting as far as trying
1716 RCPT means that the domain itself is ok).
1718 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1721 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1722 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1723 session, the daemon ignores ECONNECT errors and logs others; it now
1724 ignores EPIPE as well.
1726 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1727 (quoted-printable decoding).
1729 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1730 later the small subsequent patch to fix an introduced bug.
1732 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1734 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1736 PH/08 An error is now given if message_size_limit is specified negative.
1738 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1739 to be given (somewhat) arbitrary names.
1741 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1744 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1745 miscellaneous code fixes
1747 PH/10 Added the log_reject_target ACL modifier to specify where to log
1750 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1751 hostname. This is wrong, because it relates to the incoming message (and
1752 probably the interface on which it is arriving) and not to the outgoing
1753 callout (which could be using a different interface). This has been
1754 changed to use the value of the helo_data option from the smtp transport
1755 instead - this is what is used when a message is actually being sent. If
1756 there is no remote transport (possible with a router that sets up host
1757 addresses), $smtp_active_hostname is used.
1759 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1760 tweaks were necessary in order to get it to work (see also 21 below):
1761 (a) The code assumed that strncpy() returns a negative number on buffer
1762 overflow, which isn't the case. Replaced with Exim's string_format()
1764 (b) There were several signed/unsigned issues. I just did the minimum
1765 hacking in of casts. There is scope for a larger refactoring.
1766 (c) The code used strcasecmp() which is not a standard C function.
1767 Replaced with Exim's strcmpic() function.
1768 (d) The code set only $1; it now sets $auth1 as well.
1769 (e) A simple test gave the error "authentication client didn't specify
1770 service in request". It would seem that Dovecot has changed its
1771 interface. Fortunately there's a specification; I followed it and
1772 changed what the client sends and it appears to be working now.
1774 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1777 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1778 address A is aliased to B and C, where B exists and C does not. Without
1779 -v the output is "A verified" because verification stops after a
1780 successful redirection if more than one address is generated. However,
1781 with -v the child addresses are also verified. Exim was outputting "A
1782 failed to verify" and then showing the successful verification for C,
1783 with its parentage. It now outputs "B failed to verify", showing B's
1784 parentage before showing the successful verification of C.
1786 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1787 look up a TXT record in a specific list after matching in a combined
1790 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1791 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1792 they consult the DNS. I had assumed they would set it the way they
1793 wanted; and indeed my experiments on Linux seem to show that in some
1794 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1795 To be on the safe side, however, I have now made the interface to
1796 host_find_byname() similar to host_find_bydns(), with an argument
1797 containing the DNS resolver options. The host_find_byname() function now
1798 sets these options at its start, just as host_find_bydns() does. The smtp
1799 transport options dns_qualify_single and dns_search_parents are passed to
1800 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1801 of host_find_byname() use the default settings of RES_DEFNAMES
1802 (qualify_single) but not RES_DNSRCH (search_parents).
1804 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1805 spool_read_header() do less string testing, by means of a preliminary
1806 switch on the second character of optional "-foo" lines. (This is
1807 overdue, caused by the large number of possibilities that now exist.
1808 Originally there were few.) While I was there, I also converted the
1809 str(n)cmp tests so they don't re-test the leading "-" and the first
1810 character, in the hope this might squeeze out yet more improvement.
1812 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1813 flag allowing group syntax was set by the header_syntax check but not
1814 turned off, possible causing trouble later; (2) The flag was not being
1815 set at all for the header_verify test, causing "group"-style headers to
1816 be rejected. I have now set it in this case, and also caused header_
1817 verify to ignore an empty address taken from a group. While doing this, I
1818 came across some other cases where the code for allowing group syntax
1819 while scanning a header line wasn't quite right (mostly, not resetting
1820 the flag correctly in the right place). These bugs could have caused
1821 trouble for malformed header lines. I hope it is now all correct.
1823 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1824 with the "reply" argument non-NULL. The code, however (which originally
1825 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1826 but it didn't always do it. This confused somebody who was copying the
1827 code for some other use. I have removed all the tests.
1829 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1830 feature that was used to support insecure browsers during the U.S. crypto
1831 embargo. It requires special client support, and Exim is probably the
1832 only MTA that supported it -- and would never use it because real RSA is
1833 always available. This code has been removed, because it had the bad
1834 effect of slowing Exim down by computing (never used) parameters for the
1835 RSA_EXPORT functionality.
1837 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1838 authenticator to fail if there's a tab character in the incoming data
1839 (there should never be unless someone is messing about, as it's supposed
1840 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1841 if the connection is using TLS or if the remote IP is the same as the
1842 local IP, and the "valid-client-cert option" if a client certificate has
1845 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1846 authenticators. This can be used for authorization after authentication
1847 succeeds. (In the case of plaintext, it servers for both authentication
1850 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1851 if any retry times were supplied.
1853 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1854 connection, where there is no client IP address to check. In this
1855 situation, the verify now always succeeds.
1857 PH/25 Applied John Jetmore's -Mset patch.
1859 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1861 PH/27 In a string expansion for a processed (not raw) header when multiple
1862 headers of the same name were present, leading whitespace was being
1863 removed from all of them, but trailing whitespace was being removed only
1864 from the last one. Now trailing whitespace is removed from each header
1865 before concatenation. Completely empty headers in a concatenation (as
1866 before) are ignored.
1868 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1869 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1871 PH/29 [Removed. This was a change that I later backed out, and forgot to
1872 correct the ChangeLog entry (that I had efficiently created) before
1873 committing the later change.]
1875 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1876 address errors (4xx response to RCPT) over the same connection as other
1877 messages routed to the same hosts. Such deliveries are always "forced",
1878 so retry times are not inspected. This resulted in far too many retries
1879 for the affected addresses. The effect occurred only when there were more
1880 hosts than the hosts_max_try setting in the smtp transport when it had
1881 the 4xx errors. Those hosts that it had tried were not added to the list
1882 of hosts for which the message was waiting, so if all were tried, there
1883 was no problem. Two fixes have been applied:
1885 (i) If there are any address or message errors in an SMTP delivery, none
1886 of the hosts (tried or untried) are now added to the list of hosts
1887 for which the message is waiting, so the message should not be a
1888 candidate for sending over the same connection that was used for a
1889 successful delivery of some other message. This seems entirely
1890 reasonable: after all the message is NOT "waiting for some host".
1891 This is so "obvious" that I'm not sure why it wasn't done
1892 previously. Hope I haven't missed anything, but it can't do any
1893 harm, as the worst effect is to miss an optimization.
1895 (ii) If, despite (i), such a delivery is accidentally attempted, the
1896 routing retry time is respected, so at least it doesn't keep
1897 hammering the server.
1899 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1900 in ${readsocket because some servers need this prod.
1902 PH/32 Added some extra debug output when updating a wait-xxx database.
1904 PH/33 The hint "could be header name not terminated by colon", which has been
1905 given for certain expansion errors for a long time, was not being given
1906 for the ${if def:h_colon_omitted{... case.
1908 PH/34 The spec says: "With one important exception, whenever a domain list is
1909 being scanned, $domain contains the subject domain." There was at least
1910 one case where this was not true.
1912 PH/35 The error "getsockname() failed: connection reset by peer" was being
1913 written to the panic log as well as the main log, but it isn't really
1914 panic-worthy as it just means the connection died rather early on. I have
1915 removed the panic log writing for the ECONNRESET error when getsockname()
1918 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1919 runs only) independently of the message's sender address. This meant
1920 that, if the 4xx error was in fact related to the sender, a different
1921 message to the same recipient with a different sender could confuse
1922 things. In particualar, this can happen when sending to a greylisting
1923 server, but other circumstances could also provoke similar problems.
1924 I have changed the default so that the retry time for these errors is now
1925 based a combination of the sender and recipient addresses. This change
1926 can be overridden by setting address_retry_include_sender=false in the
1929 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1930 remote server are returned as part of bounce messages. This was not
1931 happening for LMTP over a pipe (the lmtp transport), but now it is the
1932 same for both kinds of LMTP.
1934 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1935 in header lines that were in fact CNAMEs. This is no longer the case.
1937 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1938 and queue runs started by the daemon processed all messages. This has
1939 been fixed so that -R and -S can now usefully be given with -q<time>.
1941 PH/40 Import PCRE release 6.7 (fixes some bugs).
1943 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1945 PH/42 Give an error if -q is specified more than once.
1947 PH/43 Renamed the variables $interface_address and $interface_port as
1948 $received_ip_address and $received_port, to make it clear that these
1949 values apply to message reception, and not to the outgoing interface when
1950 a message is delivered. (The old names remain recognized, of course.)
1952 PH/44 There was no timeout on the connect() call when using a Unix domain
1953 socket in the ${readsocket expansion. There now is.
1955 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1956 be meaningful with "accept".
1958 SC/01 Eximstats V1.43
1959 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1961 SC/02 Eximstats V1.44
1962 Use a glob alias rather than an array ref in the generated
1963 parser. This improves both readability and performance.
1965 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1966 Collect SpamAssassin and rejection statistics.
1967 Don't display local sender or destination tables unless
1968 there is data to show.
1969 Added average volumes into the top table text output.
1971 SC/04 Eximstats V1.46
1972 Collect data on the number of addresses (recipients)
1973 as well as the number of messages.
1975 SC/05 Eximstats V1.47
1976 Added 'Message too big' to the list of mail rejection
1977 reasons (thanks to Marco Gaiarin).
1979 SC/06 Eximstats V1.48
1980 Mainlog lines which have GMT offsets and are too short to
1981 have a flag are now skipped.
1983 SC/07 Eximstats V1.49 (Alain Williams)
1984 Added the -emptyok flag.
1986 SC/08 Eximstats V1.50
1987 Fixes for obtaining the IP address from reject messages.
1989 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1990 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1991 whitesspace changes from 4.64-PH/27
1993 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1996 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1997 are found, allow negative numbers in numeric criteria)
1999 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2001 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2002 to match changes made in 4.64-PH/43
2004 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2006 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2007 "domain" is actually an IP address, and force a failure. This locks out
2008 those revolvers/nameservers that support "A-for-A" lookups, in
2009 contravention of the specifications.
2011 PH/48 When a host name was looked up from an IP address, and the subsequent
2012 forward lookup of the name timed out, the host name was left in
2013 $sender_host_name, contrary to the specification.
2015 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2016 restricted to single-key lookups, Exim was not diagnosing an error if
2017 * or *@ was used with a query-style lookup.
2019 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2021 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2022 long ago, but noone interested enough thought of it. Let's just say that
2023 the "1.1" means that there are some new functions that weren't there at
2024 some point in the past.
2026 PH/51 Error processing for expansion failure of helo_data from an smtp
2027 transport during callout processing was broken.
2029 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2030 tested/used via the -bh/-bhc/-bs options.
2032 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2033 bug, fixed in subsequent PCRE releases).
2035 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2036 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2038 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2044 SC/01 Use a glob alias rather than an array ref in eximstats generated
2045 parser. This improves both readability and performance.
2047 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2048 Don't display local sender or destination tables in eximstats unless
2049 there is data to show.
2050 Added average volumes into the eximstats top table text output.
2052 SC/03 Collect data on the number of addresses (recipients) as well
2053 as the number of messages in eximstats.
2055 TF/01 Correct an error in the documentation for the redirect router. Exim
2056 does (usually) call initgroups() when daemonizing.
2058 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2059 with consistent privilege compared to when running as a daemon.
2061 TF/03 Note in the spec that $authenticated_id is not set for local
2062 submissions from trusted users.
2064 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2065 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2067 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2068 by adding some example configuration directives to the default
2069 configuration file. A little bit of work is required to uncomment the
2070 directives and define how usernames and passwords are checked, but
2071 there is now a framework to start from.
2073 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2074 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2075 without this. I don't know how relevant this is to other LDAP libraries.
2077 PH/02 Add the verb name to the "unknown ACL verb" error.
2079 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2081 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2083 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2084 directory not expanded when it should be if an expanded home directory
2085 was set for the address (which is overridden by the transport).
2087 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2090 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2091 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2092 because it is too late at that time, and has no effect.
2094 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2095 security issue with \' (bugzilla #107). I could not use the
2096 PQescapeStringConn() function, because it needs a PGconn value as one of
2099 PH/08 When testing addresses using -bt, indicate those final addresses that
2100 are duplicates that would not cause an additional delivery. At least one
2101 person was confused, thinking that -bt output corresponded to deliveries.
2102 (Suppressing duplicates isn't a good idea as you lose the information
2103 about possibly different redirections that led to the duplicates.)
2105 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2106 systems where poll() doesn't work, in particular OS X.
2108 PH/10 Added more information to debugging output for retry time not reached.
2110 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2111 operations in malware.c.
2113 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2116 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2117 syslog_duplication set false, log lines that would normally be written
2118 both the the main log and to the reject log were not written to syslog at
2121 PH/14 In the default configuration, change the use of "message" in ACL warn
2122 statements to "add_header".
2124 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2125 not followed by a command (e.g. "seen endif").
2127 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2128 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2131 PH/17 Added extra conditions to the default value of delay_warning_condition
2135 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2136 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2137 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2140 The Auto-Submitted: and various List- headers are standardised, whereas I
2141 don't think Precedence: ever was.
2143 PH/18 Refactored debugging code in route_finduser() to show more information,
2144 in particular, the error code if getpwnam() issues one.
2146 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2147 This is apparently needed in addition to the PH/07 change above to avoid
2148 any possible encoding problems.
2150 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2151 but not after initializing Perl.
2153 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2154 output them only if debugging. By default they are written stderr,
2155 apparently, which is not desirable.
2157 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2160 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2163 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2165 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2166 authenticated or an ident call has been made. Suppress the default
2167 values for $authenticated_id and $authenticated_sender (but permit -oMai
2168 and -oMas) when testing with -bh.
2170 PH/24 Re-jigged the order of the tests in the default configuration so that the
2171 tests for valid domains and recipients precede the DNS black list and CSA
2172 tests, on the grounds that those ones are more expensive.
2174 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2175 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2176 This bug exists in every version of Exim that I still have, right back to
2179 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2180 However, an attempt to turn on case-sensitivity in a regex key by
2181 including (?-i) didn't work because the subject string was already
2182 lowercased, and the effects were non-intuitive. It turns out that a
2183 one-line patch can be used to allow (?-i) to work as expected.
2189 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2190 other effects) broke the use of negated acl sub-conditions.
2192 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2195 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2196 "Deny" causes Exim to reject the incoming connection with a 554 error.
2197 Unfortunately, if there is a major crisis, such as a disk failure,
2198 tcp-wrappers gives "deny", whereas what one would like would be some
2199 kind of temporary error. A kludge has been added to help with this.
2200 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2201 554 error is used if errno is still zero or contains ENOENT (which occurs
2202 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2205 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2207 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2208 errors. Otherwise a message that provokes a temporary error (when other
2209 messages do not) can cause a whole host to time out.
2211 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2212 addresses were routed directly to files or pipes from a redirect router.
2213 File deliveries just didn't batch; pipe deliveries might have suffered
2216 PH/06 A failure to get a lock for a hints database would erroneously always say
2217 "Failed to get write lock", even when it was really a read lock.
2219 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2220 of 0600. This has been changed to use the value of the lockfile_mode
2221 option (which defaults to 0600).
2223 PH/08 Applied small patch from the Sieve maintainer.
2225 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2226 folder from quota calculations, a direct delivery into this folder messed
2227 up the contents of the maildirsize file. This was because the regex was
2228 used only to exclude .Trash (or whatever) when the size of the mailbox
2229 was calculated. There was no check that a delivery was happening into an
2230 excluded directory. This bug has been fixed by ignoring all quota
2231 processing for deliveries into excluded directories.
2233 PH/10 Added the maildirfolder_create_regex option to appendfile.
2239 PH/01 The code for finding all the local interface addresses on a FreeBSD
2240 system running IPv6 was broken. This may well have applied to all BSD
2241 systems, as well as to others that have similar system calls. The broken
2242 code found IPv4 interfaces correctly, but gave incorrect values for the
2243 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2244 that it would not match correctly against @[] and not recognize the IPv6
2247 PH/02 The ipliteral router was not recognizing addresses of the form user@
2248 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2250 PH/03 Added disable_ipv6.
2252 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2253 decoded form, because it is most often used to construct To: headers
2254 lines in autoreplies, and the decoded form may well be syntactically
2255 invalid. However, $reply_address has leading white space removed, and all
2256 newlines turned into spaces so that the autoreply transport does not
2259 PH/05 If group was specified without a user on a router, and no group or user
2260 was specified on a transport, the group from the router was ignored.
2262 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2263 for visible compile-time settings that can be used to change these
2264 numbers, for those that want even more. Backwards compatibility with old
2265 spool files has been maintained. However, going back to a previous Exim
2266 release will lost any variables that are in spool files.
2268 PH/07 Two small changes when running in the test harness: increase delay when
2269 passing a TCP/IP connection to a new process, in case the original
2270 process has to generate a bounce, and remove special handling of
2271 127.0.0.2 (sic), which is no longer necessary.
2273 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2274 be the same on different OS.
2276 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2279 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2280 whether --show-vars was specified or not
2282 JJ/02 exipick: Added support for new ACL variable spool format introduced
2285 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2286 syntactically invalid From: or Reply-to: line, and a filter used this to
2287 generate an autoreply, and therefore failed to obtain an address for the
2288 autoreply, Exim could try to deliver to a non-existent relative file
2289 name, causing unrelated and misleading errors. What now happens is that
2290 it logs this as a hard delivery error, but does not attempt to create a
2293 PH/11 The exinext utility has a -C option for testing purposes, but although
2294 the given file was scanned by exinext itself; it wasn't being passed on
2295 when Exim was called.
2297 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2298 an end-of-file indication when reading a command response.
2300 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2301 compiled. In many other places in Exim, IPv6 addresses are always
2302 recognized, so I have changed this. It also means that IPv4 domain
2303 literals of the form [IPV4:n.n.n.n] are now always recognized.
2305 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2306 used if the router is not running as root, for example, when verifying at
2307 ACL time, or when using -bh. The debugging output from this situation was
2308 non-existent - all you got was a failure to exec. I have made two
2311 (a) Failures to set uid/gid, the current directory, or a process leader
2312 in a subprocess such as that created by queryprogram now generate
2313 suitable debugging ouput when -d is set.
2315 (b) The queryprogram router detects when it is not running as root,
2316 outputs suitable debugging information if -d is set, and then runs
2317 the subprocess without attempting to change uid/gid.
2319 PH/15 Minor change to Makefile for building test_host (undocumented testing
2322 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2323 additional section of a DNS packet that returns MX or SRV records.
2324 Instead, it always explicitly searches for A/AAAA records. This avoids
2325 major problems that occur when a DNS server includes only records of one
2326 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2327 fixed another bug: if SRV records were looked up and the corresponding
2328 address records were *not* found in the additional section, the port
2329 values from the SRV records were lost.
2331 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2332 using the correct key (the original address) when searching the retry
2333 rules in order to find which one to use for generating the retry hint.
2335 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2336 adding the default one. Similarly, if it contains a Reply-To: header, the
2337 errors_reply_to option, if set, is not used.
2339 PH/19 When calculating a retry time, Exim used to measure the "time since
2340 failure" by looking at the "first failed" field in the retry record. Now
2341 it does not use this if it is later than than the arrival time of the
2342 message. Instead it uses the arrival time. This makes for better
2343 behaviour in cases where some deliveries succeed, thus re-setting the
2344 "first failed" field. An example is a quota failure for a huge message
2345 when small messages continue to be delivered. Without this change, the
2346 "time since failure" will always be short, possible causing more frequent
2347 delivery attempts for the huge message than are intended.
2348 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2350 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2351 $1, $2, $3) because the numerical variables can be reset during some
2352 expansion items (e.g. "match"), thereby losing the authentication data.
2354 PH/21 Make -bV show the size of off_t variables so that the test suite can
2355 decide whether to run tests for quotas > 2G.
2357 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2358 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2359 filecount value is greater than 2G or if a quota value is greater than 2G
2360 on a system where the size of off_t is not greater than 4, a panic error
2363 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2364 never match. The debug and -bh output now contains an explicit error
2365 message indicating a malformed IPv4 address or mask.
2367 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2368 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2369 PH/23 above applies.
2371 PH/25 Do not write to syslog when running in the test harness. The only
2372 occasion when this arises is a failure to open the main or panic logs
2373 (for which there is an explicit test).
2375 PH/26 Added the /no_tell option to "control=freeze".
2377 PH/27 If a host name lookup failed very early in a connection, for example, if
2378 the IP address matched host_lookup and the reverse lookup yielded a name
2379 that did not have a forward lookup, an error message of the form "no IP
2380 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2381 could be logged. Now it outputs the IP address instead of "NULL".
2383 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2384 allows the sender and the authenticated sender to be set when
2385 submitting a message from within Exim. Since child_open_exim() is
2386 documented for local_scan(), the new function should be too.
2388 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2389 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2390 results in an empty string is now treated as unset.
2392 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2394 PH/31 Added sender_verify_fail logging option.
2396 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2397 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2398 tidied the source and removed it altogether.
2400 PH/33 When a queue run was abandoned because the load average was too high, a
2401 log line was always written; now it is written only if the queue_run log
2402 selector is set. In addition, the log line for abandonment now contains
2403 information about the queue run such as the pid. This is always present
2404 in "start" and "stop" lines but was omitted from the "abandon" line.
2406 PH/34 Omit spaces between a header name and the colon in the error message that
2407 is given when verify = headers_syntax fails (if there are lots of them,
2408 the message gets confusing).
2410 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2411 names, as there are now some PTR records that contain slashes. This check
2412 is only to protect against broken name servers that fall over on strange
2413 characters, so the fact that it applies to all lookups doesn't matter.
2415 PH/36 Now that the new test suite is complete, we can remove some of the
2416 special code in Exim that was needed for the old test suite. For example,
2417 sorting DNS records because real resolvers return them in an arbitrary
2418 order. The new test suite's fake resolver always returns records in the
2421 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2422 bounces) except when queue_only is set, to avoid logging races between
2423 the different processes.
2425 PH/38 Panic-die if .include specifies a non-absolute path.
2427 PH/39 A tweak to the "H" retry rule from its user.
2429 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2430 a label. They prevented compilation on older perls.
2432 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2433 a warning to be raised on newish perls.
2435 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2436 on queue. Changes to match documented behaviour of showing count of
2437 messages matching specified criteria.
2439 PH/40 Changed the default ident timeout from 30s to 5s.
2441 PH/41 Added support for the use of login_cap features, on those BSD systems
2442 that have them, for controlling the resources used by pipe deliveries.
2444 PH/42 The content-scanning code uses fopen() to create files in which to put
2445 message data. Previously it was not paying any attention to the mode of
2446 the files. Exim runs with umask(0) because the rest of the code creates
2447 files with open(), and sets the required mode explicitly. Thus, these
2448 files were ending up world-writeable. This was not a big issue, because,
2449 being within the spool directory, they were not world-accessible. I have
2450 created a function called modefopen, which takes an additional mode
2451 argument. It sets umask(777), creates the file, chmods it to the required
2452 mode, then resets the umask. All the relevant calls to fopen() in the
2453 content scanning code have been changed to use this function.
2455 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2456 to 24 hours. This avoids potential overflow problems when processing G
2457 and H retry rules. I suspect nobody ever tinkers with this value.
2459 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2461 PH/45 When the plaintext authenticator is running as a client, the server's
2462 challenges are checked to ensure they are valid base64 strings. By
2463 default, the authentication attempt is cancelled if an invalid string is
2464 received. Setting client_ignore_invalid_base64 true ignores these errors.
2465 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2466 they are received. Thus, the responses can be made to depend on the
2467 challenges. If an invalid string is ignored, an empty string is placed in
2470 PH/46 Messages that are created by the autoreply transport now contains a
2471 References: header, in accordance with RFCs 2822 and 3834.
2473 PH/47 Added authenticated_sender_force to the smtp transport.
2475 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2477 PH/49 Installed latest patch from the Sieve maintainer.
2479 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2480 was also set, the appendfile transport was unnecessarily scanning a
2481 directory of message files (e.g. for maildir delivery) to find the count
2482 of files (along with the size), even though it did not need this
2483 information. It now does the scan only if it needs to find either the
2484 size of the count of files.
2486 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2488 PH/52 Two bugs concerned with error handling when the smtp transport is
2491 (i) Exim was not creating retry information for temporary errors given
2492 for individual recipients after the DATA command when the smtp transport
2493 was used in LMTP mode. This meant that they could be retried too
2494 frequently, and not timed out correctly.
2496 (ii) Exim was setting the flag that allows error details to be returned
2497 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2498 recipients that were returned after the DATA command.
2500 PH/53 This is related to PH/52, but is more general: for any failing address,
2501 when detailed error information was permitted to be returned to the
2502 sender, but the error was temporary, then after the final timeout, only
2503 "retry timeout exceeded" was returned. Now it returns the full error as
2504 well as "retry timeout exceeded".
2506 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2507 do this, and (what is worse) MTAs that accept it.
2509 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2510 will now be deprecated.
2512 PH/56 New os.c-cygwin from the Cygwin maintainer.
2514 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2515 formats (previously only available in exim formats via -bpr, -bpru,
2516 and -bpra. Now also available in native and exiqgrep formats)
2518 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2519 with very large, slow to parse queues
2521 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2523 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2525 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2526 responses to authentication challenges, though it was showing the
2527 challenges; (ii) I've removed the CR characters from the debug output for
2530 PH/58 Allow for the insertion of a newline as well as a space when a string
2531 is turned into more than one encoded-word during RFC 2047 encoding. The
2532 Sieve code now uses this.
2534 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2535 data_4xx, lost_connection, tls_required.
2537 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2538 message was being sent as an SMTP response.
2540 PH/61 Add -l and -k options to exicyclog.
2542 PH/62 When verifying, if an address was redirected to one new address, so that
2543 verification continued, and the new address failed or deferred after
2544 having set something in $address_data, the value of $address_data was not
2545 passed back to the ACL. This was different to the case when no
2546 redirection occurred. The value is now passed back in both cases.
2548 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2549 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2550 use login_cap.h, so on its own it isn't the distinguishing feature. The
2551 new name refers directly to the setclassresources() function.
2553 PH/65 Added configuration files for NetBSD3.
2555 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2557 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2558 is preferred over IPv4.
2560 PH/68 The bounce_return_message and bounce_return_body options were not being
2561 honoured for bounces generated during the reception of non-SMTP messages.
2562 In particular, this applied to messages rejected by the ACL. This bug has
2563 been fixed. However, if bounce_return_message is true and bounce_return_
2564 body is false, the headers that are returned for a non-SMTP message
2565 include only those that have been read before the error was detected.
2566 (In the case of an ACL rejection, they have all been read.)
2568 PH/69 The HTML version of the specification is now built in a directory called
2569 spec_html instead of spec.html, because the latter looks like a path with
2570 a MIME-type, and this confuses some software.
2572 PH/70 Catch two compiler warnings in sieve.c.
2574 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2575 function verify_get_ident() calls ip_connect() to connect a socket, but
2576 if the "connect()" function timed out, ip_connect() used to close the
2577 socket. However, verify_get_ident() also closes the socket later, and in
2578 between Exim writes to the log, which may get opened at this point. When
2579 the socket was closed in ip_connect(), the log could get the same file
2580 descriptor number as the socket. This naturally causes chaos. The fix is
2581 not to close the socket in ip_connect(); the socket should be closed by
2582 the function that creates it. There was only one place in the code where
2583 this was missing, in the iplookup router, which I don't think anybody now
2584 uses, but I've fixed it anyway.
2586 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2587 well as to direct DNS lookups. Otherwise the handling of names in host
2588 lists is inconsistent and therefore confusing.
2594 PH/01 Two changes to the default runtime configuration:
2596 (1) Move the checks for relay_from_hosts and authenticated clients from
2597 after to before the (commented out) DNS black list checks.
2599 (2) Add control=submission to the relay_from_hosts and authenticated
2600 clients checks, on the grounds that messages accepted by these
2601 statements are most likely to be submissions.
2603 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2605 (1) Generate an error if the third argument for the ${prvs expansion is
2608 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2611 (3) Reset the variables that are obtained from the first argument of
2612 ${prvscheck and used in the second argument before leaving the code,
2613 because their memory is reclaimed, so using them afterwards may do
2616 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2617 one (it's much easier than Tom thought :-).
2619 (5) Because of (4), we can now allow for the use of $prvscheck_result
2620 inside the third argument.
2622 PH/03 For some reason, the default setting of PATH when running a command from
2623 a pipe transport was just "/usr/bin". I have changed it to
2626 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2627 anything to be listed in the output from -bV.
2629 PH/05 When a filter generated an autoreply, the entire To: header line was
2630 quoted in the delivery log line, like this:
2632 => >A.N.Other <ano@some.domain> <original@ddress> ...
2634 This has been changed so that it extracts the operative address. There
2635 may be more than one such address. If so, they are comma-separated, like
2638 => >ano@some.domain,ona@other.domain <original@ddress> ...
2640 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2641 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2642 not being looked up in the rDNS to get a host name, Exim was showing the
2643 IP address twice in Received: lines, even though the IP addresses were
2644 identical. For example:
2646 Received: from [1.2.3.4] (helo=[1.2.3.4])
2648 However, if the real host name was known, it was omitting the HELO data
2649 if it matched the actual IP address. This has been tidied up so that it
2650 doesn't show the same IP address twice.
2652 PH/07 When both +timestamp and +memory debugging was on, the value given by
2653 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2654 called by the debug printing, thereby overwriting the timestamp buffer.
2655 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2657 PH/08 When the original message was included in an autoreply transport, it
2658 always said "this is a copy of the message, including all the headers",
2659 even if body_only or headers_only was set. It now gives an appropriate
2662 PH/09 Applied a patch from the Sieve maintainer which:
2664 o fixes some comments
2665 o adds the (disabled) notify extension core
2666 o adds some debug output for the result of if/elsif tests
2667 o points to the current vacation draft in the documentation
2668 and documents the missing references header update
2672 o fixes a bug in processing the envelope test (when testing
2673 multiple envelope elements, the last element determinted the
2676 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2677 Electronic Mail") by including:
2679 Auto-submitted: auto-generated
2681 in the messages that it generates (bounce messages and others, such as
2682 warnings). In the case of bounce messages for non-SMTP mesages, there was
2683 also a typo: it was using "Auto_submitted" (underscore instead of
2684 hyphen). Since every message generated by Exim is necessarily in response
2685 to another message, thes have all been changed to:
2687 Auto-Submitted: auto-replied
2689 in accordance with these statements in the RFC:
2691 The auto-replied keyword:
2693 - SHOULD be used on messages sent in direct response to another
2694 message by an automatic process,
2696 - MUST NOT be used on manually-generated messages,
2698 - MAY be used on Delivery Status Notifications (DSNs) and Message
2699 Disposition Notifications (MDNs),
2701 - MUST NOT be used on messages generated by automatic or periodic
2702 processes, except for messages which are automatic responses to
2705 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2706 to the default Received: header definition.
2708 PH/12 Added log selector acl_warn_skipped (default on).
2710 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2711 variables because (a) they are in the wrong storage pool and (b) even if
2712 they were copied, it wouldn't work properly because of the caching.
2714 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2715 checking when decoding. Apparently there are clients that generate
2716 overlong encoded strings. Why am I not surprised?
2718 PH/15 If the first argument of "${if match_address" was not empty, but did not
2719 contain an "@" character, Exim crashed. Now it writes a panic log message
2720 and treats the condition as false.
2722 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2724 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2725 extension "envelope-auth". The code is finished and in agreement with
2726 other implementations, but there is no documentation so far and in fact,
2727 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2728 not changing the active code.
2730 Print executed "if" and "elsif" statements when debugging is used. This
2731 helps a great deal to understand what a filter does.
2733 Document more things not specified clearly in RFC3028. I had all this
2734 sorted out, when out of a sudden new issues came to my mind. Oops."
2736 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2739 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2740 canonical form (as documented). However, after a host name lookup from
2741 the IP address, check_host() was doing a simple string comparison with
2742 addresses acquired from the DNS when checking that the found name did
2743 have the original IP as one of its addresses. Since any found IPv6
2744 addresses are likely to be in abbreviated form, the comparison could
2745 fail. Luckily, there already exists a function for doing the comparison
2746 by converting both addresses to binary, so now that is used instead of
2747 the text comparison.
2749 PH/20 There was another similar case to PH/19, when a complete host name was
2750 given in a host list; looking up its IP address could give an abbreviated
2751 form, whereas the current host's name might or might not be abbreviated.
2752 The same fix has been applied.
2758 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2759 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2762 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2763 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2765 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2767 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2768 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2769 Also, for "Auto-Submitted", treat anything other than "no" as
2770 non-personal, in accordance with RFC 3834. (Previously it treated
2771 anything starting "auto-" as non-personal.)
2773 TF/01 The control=submission/name=... option had a problem with syntax
2774 errors if the name included a slash character. The /name= option
2775 now slurps the rest of the string, so it can include any characters
2776 but it must come last in the list of options (after /sender_retain
2779 PH/05 Some modifications to the interface to the fake nameserver for the new
2787 TK/01 Added the "success_on_redirect" address verification option. See
2788 NewStuff for rationale and an example.
2790 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2792 PH/02 Patch to exigrep to allow it to work on syslog lines.
2794 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2795 fread() to skip over the body file's header line, because in Cygwin the
2796 header line is locked and is inaccessible.
2798 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2799 co-exist for some time) to make it clear that it is the Exim ID that is
2800 referenced, not the Message-ID: header line.
2802 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2803 string_format() function, because snprintf() does not exist on all
2806 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2807 ${stat: expansion item.
2809 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2810 protocol synchronization error", to keep the pedants happy.
2812 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2813 well as for IRIX systems, when gcc is being used. See the host.c source
2816 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2818 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2821 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2822 transport and to the smtp transport in LMTP mode.
2824 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2826 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2827 run a filter in a subprocess. This could lead to confusion in subsequent
2828 lookups in the parent process. There should also be a search_tidyup() at
2829 the end of the subprocess.
2831 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2832 only if the host matched helo_try_verify_hosts, which caused the
2833 verification to occur when the EHLO/HELO command was issued. The ACL just
2834 tested the remembered result. Now, if a previous verification attempt has
2835 not happened, "verify = helo" does it there and then.
2837 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2839 TK/03 Fix log output including CR from clamd.
2841 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2842 exist provoked a memory error which could cause a segfault.
2844 PH/15 Installed PCRE 6.2
2846 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2848 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2849 of the problem. Specifically, suggested +O2 rather than +O1 for the
2852 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2854 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2855 record, the port was not correctly logged when the outgoing_port log
2856 selector was set (it logged the transort's default port).
2858 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2859 fallback_hosts, and "hosts" in the smtp transport.
2861 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2862 host errors such as "Connection refused".
2864 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2865 authentication with radiusclient 0.4.9:
2867 - Error returned from rc_read_config was caught wrongly
2868 - Username/password not passed on to radius server due to wrong length.
2870 The presumption is that some radiusclient API changes for 4.51/PH/17
2871 were not taken care of correctly. The code is still untested by me (my
2872 Linux distribution still has 0.3.2 of radiusclient), but it was
2873 contributed by a Radius user.
2875 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2876 expanding the "port" option of the smtp transport.
2878 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2879 while reading a MIME header. Thanks to Tom Hughes for a patch.
2881 PH/24 Include config.h inside local_scan.h so that configuration settings are
2884 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2885 that in an ACL for RCPT (for example), you can examine exactly what was
2888 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2889 commands, but it was not correctly comparing the address with the actual
2890 client host address. Thus, it would show the EHLO address in Received:
2891 header lines when this was not necessary.
2893 PH/27 Added the % operator to ${eval:}.
2895 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2896 it should be ignoring failures (because with -C, for example, it has lost
2897 privilege). It wasn't ignoring creation failures other than "already
2900 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2903 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2904 because an input line was too long, either on its own, or by virtue of
2905 too many continuations, the temporary file was not being removed, and the
2906 return code was incorrect.
2908 PH/31 Missing "BOOL" in function definition in filtertest.c.
2910 PH/32 Applied Sieve patches from the maintainer.
2912 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2914 PH/33 Added "verify = not_blind".
2916 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2917 Local/Makefile (with some defaults set). These are used in built scripts
2918 such as exicyclog, but they have never been used in the exim_install
2919 script (though there are many overriding facilities there). I have
2920 arranged that the exim_install script now takes note of these two
2923 PH/35 Installed configuration files for Dragonfly.
2925 PH/36 When a locally submitted message by a trusted user did not contain a
2926 From: header, and the sender address was obtained from -f or from an SMTP
2927 MAIL command, and the trusted user did not use -F to supply a sender
2928 name, $originator_name was incorrectly used when constructing a From:
2929 header. Furthermore, $originator_name was used for submission mode
2930 messages from external hosts without From: headers in a similar way,
2931 which is clearly wrong.
2933 PH/37 Added control=suppress_local_fixups.
2935 PH/38 When log_selector = +received_sender was set, and the addition of the
2936 sender made the log line's construction buffer exactly full, or one byte
2937 less than full, an overflow happened when the terminating "\n" was
2940 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2941 when the result of a list match is failure because a DNS lookup failed.
2943 PH/40 RM_COMMAND is now used in the building process.
2945 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2946 the "build-* directories that it finds.
2948 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2949 address in a domain literal was a prefix of an interface address.
2951 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2952 when verifying a sender address, unless rewrite_headers is false.
2954 PH/44 Wrote a long comment about why errors_to addresses are verified as
2955 recipients, not senders.
2957 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2958 the ratelimit ACL was added.
2960 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2962 PH/46 Added extra information about PostgreSQL errors to the error string.
2964 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2965 avoiding the need to install special zones in a real server. This is
2966 backwards compatible; if it can't find the fake resolver, it drops back.
2967 Thus, both old and new test suites can be run.
2969 TF/02 Added util/ratelimit.pl
2971 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2972 clock is set back in time.
2974 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2975 Candler <B.Candler@pobox.com>.
2977 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2978 OK for addresses that are the result of redirections.
2980 PH/48 A number of further additions for the benefit of the new test suite,
2981 including a fake gethostbyname() that interfaces to the fake DNS resolver
2984 TF/06 The fix for widen_domains has also been applied to qualify_single and
2985 search_parents which are the other dnslookup options that can cause
2988 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2991 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2993 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2994 service" if it encountered an MX record with an empty target hostname.
2995 The message is now "an MX or SRV record indicated no SMTP service".
2997 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2998 if the DNS of the sending site is misconfigured. This is quite a
2999 common situation. This change restores the behaviour of treating a
3000 helo verification defer as a failure.
3002 PH/51 If self=fail was set on a router, the bounce message did not include the
3003 actual error message.
3009 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3011 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3012 command itself ended in error, the underlying message about the transport
3013 filter timeout was being overwritten with the pipe command error. Now the
3014 underlying error message should be appended to the second error message.
3016 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3018 PH/02 Reduce the amount of output that "make" produces by default. Full output
3019 can still be requested.
3021 PH/03 The warning log line about a condition test deferring for a "warn" verb
3022 was being output only once per connection, rather than after each
3023 occurrence (because it was using the same function as for successful
3024 "warn" verbs). This seems wrong, so I have changed it.
3026 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3027 it should not have, which might have caused a crash in the right
3028 circumstances, but probably never did.
3030 PH/04 Installed a modified version of Tony Finch's patch to make submission
3031 mode fix the return path as well as the Sender: header line, and to
3032 add a /name= option so that you can make the user's friendly name appear
3035 TF/03 Added the control = fakedefer ACL modifier.
3037 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3038 Mark Lowes for thorough testing.
3040 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3042 TK/03 Merged latest SRS patch from Miles Wilton.
3044 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3045 with the definition in sysexits.h (which is #included earlier).
3046 Fortunately, Exim does not actually use EX_OK. The code used to try to
3047 preserve the sysexits.h value, by assumimg that macro definitions were
3048 scanned for macro replacements. I have been disabused of this notion,
3049 so now the code just undefines EX_OK before #including unistd.h.
3051 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3052 in the smtp transport. When a block could not be written in a single
3053 write() function, the timeout was being re-applied to each part-write.
3054 This seems wrong - if the receiver was accepting one byte at a time it
3055 would take for ever. The timeout is now adjusted when this happens. It
3056 doesn't have to be particularly precise.
3058 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3059 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3061 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3062 without a domain if the check to <postmaster@domain> fails.
3064 SC/01 Eximstats: added -xls and the ability to specify output files
3065 (patch written by Frank Heydlauf).
3067 SC/02 Eximstats: use FileHandles for outputing results.
3069 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3071 SC/04 Eximstats: fixed display of large numbers with -nvr option
3073 SC/05 Eximstats: fixed merging of reports with empty tables.
3075 SC/06 Eximstats: added the -include_original_destination flag
3077 SC/07 Eximstats: removed tabs and trailing whitespace.
3079 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3081 TK/06 MBOX spool code: Add real "From " MBOX separator line
3082 so the .eml file is really in mbox format (even though
3083 most programs do not really care). Patch from Alex Miller.
3085 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3086 The latter is generated from $received_to and is only set if the
3087 message has one envelope recipient. SA can use these headers,
3088 obviously out-of-the-box. Patch from Alex Miller.
3090 PH/08 The ${def test on a variable was returning false if the variable's
3091 value was "0", contrary to what the specification has always said!
3092 The result should be true unless the variable is empty.
3094 PH/09 The syntax error of a character other than { following "${if
3095 def:variable_name" (after optional whitespace) was not being diagnosed.
3096 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3097 accidental colon was present, for example, could give incorrect results.
3099 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3100 result is used (not including appendfile, where other changes are about
3103 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3104 This involved changing a lot of size variables from int to off_t. It
3105 should work with maildirs and everything.
3107 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3108 spamd dying while we are connected to it.
3110 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3113 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3114 function for BDB 4.3.
3116 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3118 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3119 the sources for PCRE have been re-arranged and more files are now
3122 PH/15 The code I had for printing potentially long long variables in PH/11
3123 above was not the best (it lost precision). The length of off_t variables
3124 is now inspected at build time, and an appropriate printing format (%ld
3125 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3126 to be "long long int" or "long int". This is needed for the internal
3127 formatting function string_vformat().
3129 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3130 the configuration file to be ":syslog", then the script "guesses" where
3131 the logs files are, rather than using the compiled in default. In our
3132 case the guess is not the same as the compiled default, so the script
3133 suddenly stopped working when I started to use syslog. The patch checks
3134 to see if log_file_path is "". If so, it attempts to read it from exim
3135 with no configuration file to get the compiled in version, before it
3136 falls back to the previous guessing code."
3138 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3139 implementing BATV in an Exim configuration. See NewStuff for the gory
3142 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3143 Makefile that are specific to HP-UX.
3145 PH/18 If the "use_postmaster" option was set for a recipient callout together
3146 with the "random" option, the postmaster address was used as the MAIL
3147 FROM address for the random test, but not for the subsequent recipient
3148 test. It is now used for both.
3150 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3151 patch removes a few documentation additions to RFC 3028, because the
3152 latest draft now contains them. It adds the new en;ascii-case comparator
3153 and a new error check for 8bit text in MIME parts. Comparator and
3154 require names are now matched exactly. I enabled the subaddress
3155 extension, but it is not well tested yet (read: it works for me)."
3157 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3158 rework some of the code of TK/09 above to avoid the hardwired use of
3159 "%lld" and "long long". Replaced the call to snprintf() with a call to
3162 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3163 records point to non-existent hosts", "retry timeout exceeded", and
3164 "retry time not reached for any host after a long failure period".
3166 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3167 experimental DomainKeys support:
3169 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3170 (2) On an error such as an illegally used "control", the wrong name for
3171 the control was given.
3173 These problems did NOT occur unless DomainKeys support was compiled.
3175 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3177 PH/24 Added ${if match_ip condition.
3179 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3180 databases so that it will be absolutely obvious if a crash occurs in the
3181 DB library. This is a regular occurrence (often caused by mis-matched
3184 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3185 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3186 picked up on a user's system that detects such things. There doesn't seem
3187 to be a gcc warning option for this - only an attribute that has to be
3188 put on the function's prototype. It seems that in Fedora Core 4 they have
3189 set this on a number of new functions. No doubt there will be more in due
3192 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3193 specify a transport. However, if an address that was verified by such a
3194 router was the subject of a callout, Exim crashed because it tried to
3195 read the rcpt_include_affixes from the non-existent transport. Now it
3196 just assumes that the setting of that option is false. This bug was
3197 introduced by 4.51/PH/31.
3199 PH/28 Changed -d+all to exclude +memory, because that information is very
3200 rarely of interest, but it makes the output a lot bigger. People tend to
3201 do -d+all out of habit.
3203 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3204 code in os-type was giving problems when libc.so lives in lib64, like on
3207 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3208 aren't the modern standard, and it seems that some systems' include files
3209 don't always have them. Exim was already checking for some of the newer
3210 ones like T_AAAA, and defining it itself. I've added checks for all the
3211 record types that Exim uses.
3213 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3214 not automatically generating a new one, as it is supposed to. This
3215 prevented TLS from working. If the file did exist, but contained invalid
3216 data, a new version was generated, as expected. It was only the case of a
3217 non-existent file that was broken.
3219 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3220 with a change in libdomainkeys > 0.64.
3222 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3223 from DNS. If the selector record carries the flag, it now has
3224 precedence over the domain-wide flag.
3226 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3228 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3229 the use of a port name that isn't defined in /etc/services) occurred, the
3230 message was deferred as in a normal delivery, and thus remained on the
3231 spool, instead of being failed because of the mua_wrapper setting. This
3232 is now fixed, and I tidied up some of the mua_wrapper messages at the
3235 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3236 the messages in a hash of arrays rather than using individual hashes.
3237 This is a bit cleaner and results in dramatic memory savings, albeit
3238 at a slight CPU cost.
3240 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3241 as requested by Marc Sherman.
3243 SC/10 Eximstats: added histograms for user specified patterns as requested
3246 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3248 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3249 fopen() in the content-scanning modules that did not already have it.
3255 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3256 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3258 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3260 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3262 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3263 to test Sieve filters that use "vacation".
3265 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3266 that changes the way the GnuTLS parameters are stored in the cache file.
3267 The new format can be generated externally. For backward compatibility,
3268 if the data in the cache doesn't make sense, Exim assumes it has read an
3269 old-format file, and it generates new data and writes a new file. This
3270 means that you can't go back to an older release without removing the
3273 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3274 work if there are any delivery delays because "one_time" forces the
3275 parent to be marked "delivered", so its unseen clone is never tried
3276 again. For this reason, Exim now forbids the simultaneous setting of
3279 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3280 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3281 entry if you want to know the details. The fix, however, neglected to
3282 consider the case when local delivery batching is involved. The test for
3283 "previously delivered" was not happening when checking to see if an
3284 address could be batched with a previous (undelivered) one; under
3285 certain circumstances this could lead to multiple deliveries to the same
3288 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3289 in its include files, and this causes problems building Exim.
3291 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3292 header_syntax) but Exim was just ignoring anything given after a slash.
3293 In particular, this caused confusion with an attempt to use "verify =
3294 reverse_host_lookup/defer_ok". An error is now given when options are
3295 supplied for verify items that do not have them. (Maybe reverse_host_
3296 lookup should have a defer_ok option, but that's a different point.)
3298 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3299 defined by RFC 821) to 2048, because there were problems with some AUTH
3300 commands, and RFC 1869 says the size should be increased for extended
3301 SMTP commands that take arguments.
3303 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3306 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3307 "unknown" error; now it says that the functionality isn't in the binary.
3309 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3310 an address' error message when a string expansion fails (syntax or
3311 whatever). Otherwise the password may appear in the log. Following change
3312 PH/42 below, there is no longer a chance of it appearing in a bounce
3315 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3317 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3318 of its addresses was ever tried. (Bugzilla bug #2.)
3320 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3321 the result incorrectly in the debug output. (It correctly added a newline
3322 to what was transported.)
3324 TF/01 Added $received_time.
3326 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3327 commented out examples of how to interface to a virus scanner and to
3328 SpamAssassin. Also added commented examples of av_scanner and
3329 spamd_address settings.
3331 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3332 and controls are allowed in which ACLs. There were a couple of minor
3333 errors. Some of the entries in the conditions table (which is a table of
3334 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3335 negation of where the condition IS allowed.
3337 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3339 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3340 header file does not have a version number, so I've had to invent a new
3341 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3342 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3343 radiusclient), but it was contributed by a Radius user.
3345 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3346 files or directories, for OpenSSL.
3348 PH/19 When an Exim process that is running as root has to create an Exim log
3349 file, it does so in a subprocess that runs as exim:exim so as to get the
3350 ownership right at creation (otherwise, other Exim processes might see
3351 the file with the wrong ownership). There was no test for failure of this
3352 fork() call, which would lead to the process getting stuck as it waited
3353 for a non-existent subprocess. Forks do occasionally fail when resources
3354 run out. I reviewed all the other calls to fork(); they all seem to check
3357 PH/20 When checking for unexpected SMTP input at connect time (before writing
3358 the banner), Exim was not dealing correctly with a non-positive return
3359 from the read() function. If the client had disconnected by this time,
3360 the result was a log entry for a synchronization error with an empty
3361 string after "input=" when read() returned zero. If read() returned -1
3362 (an event I could not check), uninitialized data bytes were printed.
3363 There were reports of junk text (parts of files, etc) appearing after
3366 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3368 PH/22 Added support for macro redefinition, and (re)definition in between
3369 driver and ACL definitions.
3371 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3372 forgetting to use the resulting value; it was using the unexpanded value.
3374 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3375 hadn't been configured. The fix is from Juergen Kreileder, who
3376 understands it better than I do:
3378 "Here's what I see happening with three configured cyrus_sasl
3379 authenticators configured (plain, login, cram-md5):
3381 On startup auth_cyrus_sasl_init() gets called for each of these.
3382 This means three calls to sasl_listmech() without a specified mech_list.
3383 => SASL tests which mechs of all available mechs actually work
3384 => three warnings about OTP not working
3385 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3387 With the patch, sasl_listmech() also gets called three times. But now
3388 SASL's mech_list option is set to the server_mech specified in the the
3389 authenticator. Or in other words, the answer from sasl_listmech()
3390 gets limited to just the mech you're testing for (which is different
3392 => the return list contains just 'plain' or 'login', 'cram-md5' or
3393 nothing depending on the value of ob->server_mech.
3395 I've just tested the patch: Authentication still works fine,
3396 unavailable mechs specified in the exim configuration are still
3397 caught, and the auth.log warnings about OTP are gone."
3399 PH/25 When debugging is enabled, the contents of the command line are added
3400 to the debugging output, even when log_selector=+arguments is not
3403 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3404 answer is "GNU", and only if the return is "GNU/something" is the answer
3407 PH/27 $acl_verify_message is now set immediately after the failure of a
3408 verification in an ACL, and so is available in subsequent modifiers. In
3409 particular, the message can be preserved by coding like this:
3411 warn !verify = sender
3412 set acl_m0 = $acl_verify_message
3414 Previously, $acl_verify_message was set only while expanding "message"
3415 and "log_message" when a very denied access.
3417 PH/28 Modified OS/os.c-Linux with
3419 -#ifndef OS_LOAD_AVERAGE
3420 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3422 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3423 nomenclature these days.)
3425 PH/29 Installed patch from the Sieve maintainer that adds the options
3426 sieve_useraddress and sieve_subaddress to the redirect router.
3428 PH/30 In these circumstances:
3429 . Two addresses routed to the same list of hosts;
3430 . First host does not offer TLS;
3431 . First host accepts first address;
3432 . First host gives temporary error to second address;
3433 . Second host offers TLS and a TLS session is established;
3434 . Second host accepts second address.
3435 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3436 and peerdn, if requested) that were in fact used only for the second
3439 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3440 attention to any local part prefix or suffix that was matched by the
3441 router that accepted the address. It now behaves in the same way as it
3442 does for delivery: the affixes are removed from the local part unless
3443 rcpt_include_affixes is set on the transport.
3445 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3446 timeout during the DATA phase of an incoming message.
3448 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3449 applied a patch sanctioned by the Sieve maintainer.
3451 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3452 the uid or gid is negative. A case of a negative gid caused this to be
3453 noticed. The fix allows for either to be negative.
3455 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3456 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3459 PH/36 The change PH/12 above was broken. Fixed it.
3461 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3462 the grounds that routing the same address twice would always produce the
3463 same answer. This might have been true once, but it is certainly no
3464 longer true now. Routing a child address may depend on the previous
3465 routing that produced that child. Some complicated redirection strategies
3466 went wrong when messages had multiple recipients, and made Exim's
3467 behaviour dependent on the order in which the addresses were given.
3469 I have moved the duplicate checking until after the routing is complete.
3470 Exim scans the addresses that are assigned to local and remote
3471 transports, and removes any duplicates. This means that more work will be
3472 done, as duplicates will always all be routed, but duplicates are
3473 presumably rare, so I don't expect this is of any significance.
3475 For deliveries to pipes, files, and autoreplies, the duplicate checking
3476 still happens during the routing process, since they are not going to be
3479 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3480 It corrects a timeout issue with spamd. This is Ian's comment: "The
3481 background is that sometimes spamd either never reads data from a
3482 connection it has accepted, or it never writes response data. The exiscan
3483 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3484 blindly assumes that writes won't block so it may never time out."
3486 PH/39 Allow G after quota size as well as K and M.
3488 PH/40 The value set for $authenticated_id in an authenticator may not contain
3489 binary zeroes or newlines because the value is written to log lines and
3490 to spool files. There was no check on this. Now the value is run through
3491 the string_printing() function so that such characters are converted to
3492 printable escape sequences.
3494 PH/41 $message_linecount is a new variable that contains the total number of
3495 lines in the message. Compare $body_linecount, which is the count for the
3498 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3499 bounce and delay warning messages, except in certain special cases, which
3502 (a) An SMTP error message from a remote host;
3503 (b) A message specified in a :fail: redirection;
3504 (c) A message specified in a "fail" command in a system filter;
3505 (d) A message specified in a FAIL return from the queryprogram router;
3506 (e) A message specified by the cannot_route_message router option.
3508 In these cases only, Exim does include the error details in bounce and
3509 warning messages. There are also a few cases where bland messages such
3510 as "unrouteable address" or "local delivery error" are given.
3512 PH/43 $value is now also set for the "else" part of a ${run expansion.
3514 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3515 being worked on, but at least Exim now implements the latest version to
3518 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3519 process to complete was treated as a delivery failure, a timeout while
3520 writing the message to the pipe was logged, but erroneously treated as a
3521 successful delivery. Such timeouts include transport filter timeouts. For
3522 consistency with the overall process timeout, these timeouts are now
3523 treated as errors, giving rise to delivery failures by default. However,
3524 there is now a new Boolean option for the pipe transport called
3525 timeout_defer, which, if set TRUE, converts the failures into defers for
3526 both kinds of timeout. A transport filter timeout is now identified in
3529 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3530 systems where "make" and "gmake" are different, calling "gmake" at top
3531 level broke things. I've arranged for the value of $(MAKE) to be passed
3532 from the Makefile to this script so that it can call the same version of
3536 A note about Exim versions 4.44 and 4.50
3537 ----------------------------------------
3539 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3540 changes of various kinds. As a consequence, a big documentation update was
3541 needed. This delayed the release for rather longer than seemed good, especially
3542 in the light of a couple of (minor) security issues. Therefore, the changes
3543 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3544 release. So 4.44 and 4.50 are in effect two different branches that both start
3547 I have left the 4.50 change log unchanged; it contains all the changes since
3548 4.43. The change log for 4.44 is below; many of its items are identical to
3549 those for 4.50. This seems to be the most sensible way to preserve the
3550 historical information.
3556 1. Minor wording change to the doc/README.SIEVE file.
3558 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3559 computation of the current number of files was incorrect.
3561 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3562 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3565 4. Give more explanation in the error message when the command for a transport
3566 filter fails to execute.
3568 5. There are several places where Exim runs a non-Exim command in a
3569 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3570 was being done only for the command run by the queryprogram router. It is
3571 now done for all such subprocesses. The other cases are: ${run, transport
3572 filters, and the commands run by the lmtp and pipe transports.
3574 6. Added CONFIGURE_GROUP build-time option.
3576 7. Some older OS have a limit of 256 on the maximum number of file
3577 descriptors. Exim was using setrlimit() to set 1000 as a large value
3578 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3579 systems. I've change it so that if it can't get 1000, it tries for 256.
3581 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3582 was an oversight, and furthermore, ever since the addition of extra
3583 controls (e.g. 4.43/32), the checks on when to allow different forms of
3584 "control" were broken. There should now be diagnostics for all cases when a
3585 control that does not make sense is encountered.
3587 9. Added the /retain_sender option to "control=submission".
3589 10. $recipients is now available in the predata ACL (oversight).
3591 11. Tidy the search cache before the fork to do a delivery from a message
3592 received from the command line. Otherwise the child will trigger a lookup
3593 failure and thereby defer the delivery if it tries to use (for example) a
3594 cached ldap connection that the parent has called unbind on.
3596 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3597 of $address_data from the recipient verification was clobbered by the
3598 sender verification.
3600 13. The value of address_data from a sender verification is now available in
3601 $sender_address_data in subsequent conditions in the ACL statement.
3603 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3605 15. Added a new option "connect=<time>" to callout options, to set a different
3608 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3609 was its contents. (It was OK if the option was not defined at all.)
3611 17. A "Completed" log line is now written for messages that are removed from
3612 the spool by the -Mrm option.
3614 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3615 information about exactly what failed.
3617 19. Added -dd to debug only the daemon process.
3619 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3620 handles timeouts, both on the server side and network timeouts. Renamed the
3621 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3623 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3624 It is now set to "smtps".
3626 22. $host_address is now set to the target address during the checking of
3627 ignore_target_hosts.
3629 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3630 being passed; this would have caused $sender_host_name to have been used if
3631 matching the list had actually called for a host name (not very likely,
3632 since this list is usually IP addresses). A host name is now passed as
3635 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3636 code that specifies a non-restarting handler (typically sigaction() in
3637 modern systems) in an attempt to fix a rare and obscure crash bug.
3639 25. Narrowed the window for a race in the daemon that could cause it to ignore
3640 SIGCHLD signals. This is not a major problem, because they are used only to
3641 wake it up if nothing else does.
3643 26. A malformed maildirsize file could cause Exim to calculate negative values
3644 for the mailbox size or file count. Odd effects could occur as a result.
3645 The maildirsize information is now recalculated if the size or filecount
3648 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3649 support for a long time. Removed HAVE_SYS_VFS_H.
3651 28. Installed the latest version of exipick from John Jetmore.
3653 29. In an address list, if the pattern was not a regular expression, an empty
3654 subject address (from a bounce message) matched only if the pattern was an
3655 empty string. Non-empty patterns were not even tested. This was the wrong
3656 because it is perfectly reasonable to use an empty address as part of a
3657 database query. An empty address is now tested by patterns that are
3658 lookups. However, all the other forms of pattern expect the subject to
3659 contain a local part and a domain, and therefore, for them, an empty
3660 address still always fails if the pattern is not itself empty.
3662 30. Exim went into a mad DNS loop when attempting to do a callout where the
3663 host was specified on an smtp transport, and looking it up yielded more
3664 than one IP address.
3666 31. Re-factored the code for checking spool and log partition space into a
3667 function that finds that data and another that does the check. The former
3668 is then used to implement four new variables: $spool_space, $log_space,
3669 $spool_inodes, and $log_inodes.
3671 32. The RFC2047 encoding function was originally intended for short strings
3672 such as real names; it was not keeping to the 75-character limit for
3673 encoded words that the RFC imposes. It now respects the limit, and
3674 generates multiple encoded words if necessary. To be on the safe side, I
3675 have increased the buffer size for the ${rfc2047: expansion operator from
3678 33. It is now permitted to omit both strings after an "if" condition; if the
3679 condition is true, the result is "true". As before, when the second string
3680 is omitted, a false condition yields an empty string. This makes it less
3681 cumbersome to write custom ACL and router conditions.
3683 34. Failure to deliver a bounce message always caused it to be frozen, even if
3684 there was an errors_to setting on the router. The errors_to setting is now
3687 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3688 canonical form (fully expanded) before being placed in
3689 $sender_host_address.
3691 36. The table in the code that translates DNS record types into text (T_A to
3692 "A" for instance) was missing entries for NS and CNAME. It is just possible
3693 that this could have caused confusion if both these types were looked up
3694 for the same domain, because the text type is used as part of Exim's
3695 per-process caching. But the chance of anyone hitting this buglet seems
3698 37. The dnsdb lookup has been extended in a number of ways.
3700 (1) There is a new type, "zns", which walks up the domain tree until it
3701 finds some nameserver records. It should be used with care.
3703 (2) There is a new type, "mxh", which is like "mx" except that it returns
3704 just the host names, not the priorities.
3706 (3) It is now possible to give a list of domains (or IP addresses) to be
3707 looked up. The behaviour when one of the lookups defers can be
3708 controlled by a keyword.
3710 (4) It is now possible to specify the separator character for use when
3711 multiple records are returned.
3713 38. The dnslists ACL condition has been extended: it is now possible to supply
3714 a list of IP addresses and/or domains to be looked up in a particular DNS
3717 39. Added log_selector=+queue_time_overall.
3719 40. When running the queue in the test harness, wait just a tad after forking a
3720 delivery process, to get repeatability of debugging output.
3722 41. Include certificate and key file names in error message when GnuTLS fails
3723 to set them up, because the GnuTLS error message doesn't include the name
3724 of the failing file when there is a problem reading it.
3726 42. Allow both -bf and -bF in the same test run.
3728 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3730 44. The "Exiscan patch" is now merged into the mainline Exim source.
3732 45. Sometimes the final signoff response after QUIT could fail to get
3733 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3734 < 0 before doing a fflush(). This bug looks as though it goes back to the
3735 introduction of TLS in release 3.20, but "sometimes" must have been rare
3736 because the tests only now provoked it.
3738 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3739 (this can affect the format of dates).
3741 47. exim_tidydb, when checking for the continued existence of a message for
3742 which it has found a message-specific retry record, was not finding
3743 messages that were in split spool directories. Consequently, it was
3744 deleting retry records that should have stayed in existence.
3746 48. Steve fixed some bugs in eximstats.
3748 49. The SPA authentication driver was not abandoning authentication and moving
3749 on to the next authenticator when an expansion was forced to fail,
3750 contradicting the general specification for all authenticators. Instead it
3751 was generating a temporary error. It now behaves as specified.
3753 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3754 (the order specifies the preference for clients). The order is now AES256,
3755 AES128, 3DES, ARCFOUR128.
3757 51. Small patch to Sieve code - explicitly set From: when generating an
3760 52. Exim crashed if a remote delivery caused a very long error message to be
3761 recorded - for instance if somebody sent an entire SpamAssassin report back
3762 as a large number of 550 error lines. This bug was coincidentally fixed by
3763 increasing the size of one of Exim's internal buffers (big_buffer) that
3764 happened as part of the Exiscan merge. However, to be on the safe side, I
3765 have made the code more robust (and fixed the comments that describe what
3768 53. Now that there can be additional text after "Completed" in log lines (if
3769 the queue_time_overall log selector is set), a one-byte patch to exigrep
3770 was needed to allow it to recognize "Completed" as not the last thing in
3773 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3774 patch that reportedly fixes this has been added. I am not expert enough to
3775 create a test for it. This is what the patch creator wrote:
3777 "I found a little strange behaviour of ldap code when working with
3778 Windows 2003 AD Domain, where users was placed in more than one
3779 Organization Units. When I tried to give exim partial DN, the exit code
3780 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3781 But simultaneously result of request was absolutely normal ldap result,
3782 so I produce this patch..."
3784 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3785 REFERENCE, so I have modified the code to exclude the patch when that macro
3788 55. Some experimental protocols are using DNS PTR records for new purposes. The
3789 keys for these records are domain names, not reversed IP addresses. The
3790 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3791 leaves it alone. Component reversal etc. now happens only for IP addresses.
3794 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3796 57. Double the size of the debug message buffer (to 2048) so that more of very
3797 long debug lines gets shown.
3799 58. The exicyclog utility now does better if the number of log files to keep
3800 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3802 59. Two changes related to the smtp_active_hostname option:
3804 (1) $smtp_active_hostname is now available as a variable.
3805 (2) The default for smtp_banner uses $smtp_active_hostname instead
3806 of $primary_hostname.
3808 60. The host_aton() function is supposed to be passed a string that is known
3809 to be a valid IP address. However, in the case of IPv6 addresses, it was
3810 not checking this. This is a hostage to fortune. Exim now panics and dies
3811 if the condition is not met. A case was found where this could be provoked
3812 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3813 components; fortuitously, this particular loophole had already been fixed
3814 by change 4.50/55 above.
3816 If there are any other similar loopholes, the new check in host_aton()
3817 itself should stop them being exploited. The report I received stated that
3818 data on the command line could provoke the exploit when Exim was running as
3819 exim, but did not say which command line option was involved. All I could
3820 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3821 running as the user.
3824 61. There was a buffer overflow vulnerability in the SPA authentication code
3825 (which came originally from the Samba project). I have added a test to the
3826 spa_base64_to_bits() function which I hope fixes it.
3829 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3830 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3832 63. The daemon start-up calls getloadavg() while still root for those OS that
3833 need the first call to be done as root, but it missed one case: when
3834 deliver_queue_load_max is set with deliver_drop_privilege. This is
3835 necessary for the benefit of the queue runner, because there is no re-exec
3836 when deliver_drop_privilege is set.
3838 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3839 This has been fixed.
3841 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3842 was in use, was not putting the data itself into the right store pool;
3843 consequently, it could be overwritten for a subsequent message in the same
3844 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3847 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3849 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3850 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3851 the function were treating the return as a boolean value, which happened to
3852 work because 0=false and not-0=true, but is not correct code.
3854 68. The host_aton() function was not handling scoped IPv6 addresses (those
3855 with, for example, "%eth0" on the end) correctly.
3857 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3858 negated items (that is, ~something) in unsigned ints. Some compilers
3859 apparently mutter when there is no cast.
3861 70. If an address verification called from an ACL failed, and did not produce a
3862 user-specific message (i.e. there was only a "system" message), nothing was
3863 put in $acl_verify_message. In this situation, it now puts the system
3866 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3867 session; change 4.31/43 added the unwanted input to the log line - except
3868 that it did not do this in the start of session case. It now does.
3870 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3871 This is wrong and can cause the other end to generate a synchronization
3872 error if it is another Exim or anything else that does the synchronization
3873 check. A QUIT command is no longer sent after a timeout.
3875 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3876 during host lookups.
3878 74. The defer_ok option of callout verification was not working if it was used
3879 when verifying addresses in header lines, that is, for this case:
3881 verify = header_sender/callout=defer_ok
3883 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3884 those file descriptors could be used for SMTP connections. If anything
3885 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3886 could be sent to the SMTP client, causing chaos. The daemon now opens
3887 stdin, stdout, and stderr to /dev/null when it puts itself into the
3890 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3891 log by default. The user can override this with suitable Perl magic.
3893 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3894 the log message when discard triggers, was not working for the DATA ACL or
3895 for the non-SMTP ACL.
3897 78. Error message wording change in sieve.c.
3899 79. If smtp_accept_max_per_host was set, the number of connections could be
3900 restricted to fewer than expected, because the daemon was trying to set up
3901 a new connection before checking whether the processes handling previous
3902 connections had finished. The check for completed processes is now done
3903 earlier. On busy systems, this bug wouldn't be noticed because something
3904 else would have woken the daemon, and it would have reaped the completed
3907 80. If a message was submitted locally by a user whose login name contained one
3908 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3909 It caused a spool format error. I have fixed the spool reading code. A
3910 related problem was that the "from" clause in the Received: line became
3911 illegal because of the space(s). It is now covered by ${quote_local_part.
3913 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3916 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3918 83. Patch from Sieve maintainer to fix unterminated string problem in
3919 "vacation" handling.
3921 84. Some minor changes to the Linux configuration files to help with other
3922 OS variants using glibc.
3924 85. One more patch for Sieve to update vacation handling to latest spec.
3927 ----------------------------------------------------
3928 See the note above about the 4.44 and 4.50 releases.
3929 ----------------------------------------------------
3935 1. Change 4.43/35 introduced a bug that caused file counts to be
3936 incorrectly computed when quota_filecount was set in an appendfile
3939 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3940 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3943 3. Give more explanation in the error message when the command for a transport
3944 filter fails to execute.
3946 4. There are several places where Exim runs a non-Exim command in a
3947 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3948 was being done only for the command run by the queryprogram router. It is
3949 now done for all such subprocesses. The other cases are: ${run, transport
3950 filters, and the commands run by the lmtp and pipe transports.
3952 5. Some older OS have a limit of 256 on the maximum number of file
3953 descriptors. Exim was using setrlimit() to set 1000 as a large value
3954 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3955 systems. I've change it so that if it can't get 1000, it tries for 256.
3957 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3958 was an oversight, and furthermore, ever since the addition of extra
3959 controls (e.g. 4.43/32), the checks on when to allow different forms of
3960 "control" were broken. There should now be diagnostics for all cases when a
3961 control that does not make sense is encountered.
3963 7. $recipients is now available in the predata ACL (oversight).
3965 8. Tidy the search cache before the fork to do a delivery from a message
3966 received from the command line. Otherwise the child will trigger a lookup
3967 failure and thereby defer the delivery if it tries to use (for example) a
3968 cached ldap connection that the parent has called unbind on.
3970 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3971 of $address_data from the recipient verification was clobbered by the
3972 sender verification.
3974 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3975 was its contents. (It was OK if the option was not defined at all.)
3977 11. A "Completed" log line is now written for messages that are removed from
3978 the spool by the -Mrm option.
3980 12. $host_address is now set to the target address during the checking of
3981 ignore_target_hosts.
3983 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3984 being passed; this would have caused $sender_host_name to have been used if
3985 matching the list had actually called for a host name (not very likely,
3986 since this list is usually IP addresses). A host name is now passed as
3989 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3990 code that specifies a non-restarting handler (typically sigaction() in
3991 modern systems) in an attempt to fix a rare and obscure crash bug.
3993 15. Narrowed the window for a race in the daemon that could cause it to ignore
3994 SIGCHLD signals. This is not a major problem, because they are used only to
3995 wake it up if nothing else does.
3997 16. A malformed maildirsize file could cause Exim to calculate negative values
3998 for the mailbox size or file count. Odd effects could occur as a result.
3999 The maildirsize information is now recalculated if the size or filecount
4002 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4003 support for a long time. Removed HAVE_SYS_VFS_H.
4005 18. Updated exipick to current release from John Jetmore.
4007 19. Allow an empty sender to be matched against a lookup in an address list.
4008 Previously the only cases considered were a regular expression, or an
4011 20. Exim went into a mad DNS lookup loop when doing a callout where the
4012 host was specified on the transport, if the DNS lookup yielded more than
4015 21. The RFC2047 encoding function was originally intended for short strings
4016 such as real names; it was not keeping to the 75-character limit for
4017 encoded words that the RFC imposes. It now respects the limit, and
4018 generates multiple encoded words if necessary. To be on the safe side, I
4019 have increased the buffer size for the ${rfc2047: expansion operator from
4022 22. Failure to deliver a bounce message always caused it to be frozen, even if
4023 there was an errors_to setting on the router. The errors_to setting is now
4026 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4027 canonical form (fully expanded) before being placed in
4028 $sender_host_address.
4030 24. Updated eximstats to version 1.33
4032 25. Include certificate and key file names in error message when GnuTLS fails
4033 to set them up, because the GnuTLS error message doesn't include the name
4034 of the failing file when there is a problem reading it.
4036 26. Expand error message when OpenSSL has problems setting up cert/key files.
4039 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4040 (this can affect the format of dates).
4042 28. exim_tidydb, when checking for the continued existence of a message for
4043 which it has found a message-specific retry record, was not finding
4044 messages that were in split spool directories. Consequently, it was
4045 deleting retry records that should have stayed in existence.
4047 29. eximstats updated to version 1.35
4048 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4049 1.35 - bugfix such that pie charts by volume are generated correctly
4051 30. The SPA authentication driver was not abandoning authentication and moving
4052 on to the next authenticator when an expansion was forced to fail,
4053 contradicting the general specification for all authenticators. Instead it
4054 was generating a temporary error. It now behaves as specified.
4056 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4057 (the order specifies the preference for clients). The order is now AES256,
4058 AES128, 3DES, ARCFOUR128.
4060 31. Small patch to Sieve code - explicitly set From: when generating an
4063 32. Exim crashed if a remote delivery caused a very long error message to be
4064 recorded - for instance if somebody sent an entire SpamAssassin report back
4065 as a large number of 550 error lines. This bug was coincidentally fixed by
4066 increasing the size of one of Exim's internal buffers (big_buffer) that
4067 happened as part of the Exiscan merge. However, to be on the safe side, I
4068 have made the code more robust (and fixed the comments that describe what
4071 33. Some experimental protocols are using DNS PTR records for new purposes. The
4072 keys for these records are domain names, not reversed IP addresses. The
4073 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4074 leaves it alone. Component reversal etc. now happens only for IP addresses.
4077 34. The host_aton() function is supposed to be passed a string that is known
4078 to be a valid IP address. However, in the case of IPv6 addresses, it was
4079 not checking this. This is a hostage to fortune. Exim now panics and dies
4080 if the condition is not met. A case was found where this could be provoked
4081 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4082 components; fortuitously, this particular loophole had already been fixed
4083 by change 4.50/55 or 4.44/33 above.
4085 If there are any other similar loopholes, the new check in host_aton()
4086 itself should stop them being exploited. The report I received stated that
4087 data on the command line could provoke the exploit when Exim was running as
4088 exim, but did not say which command line option was involved. All I could
4089 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4090 running as the user.
4093 35. There was a buffer overflow vulnerability in the SPA authentication code
4094 (which came originally from the Samba project). I have added a test to the
4095 spa_base64_to_bits() function which I hope fixes it.
4098 36. The daemon start-up calls getloadavg() while still root for those OS that
4099 need the first call to be done as root, but it missed one case: when
4100 deliver_queue_load_max is set with deliver_drop_privilege. This is
4101 necessary for the benefit of the queue runner, because there is no re-exec
4102 when deliver_drop_privilege is set.
4104 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4105 was in use, was not putting the data itself into the right store pool;
4106 consequently, it could be overwritten for a subsequent message in the same
4107 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4110 38. Sometimes the final signoff response after QUIT could fail to get
4111 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4112 < 0 before doing a fflush(). This bug looks as though it goes back to the
4113 introduction of TLS in release 3.20, but "sometimes" must have been rare
4114 because the tests only now provoked it.
4120 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4121 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4122 Now it returns an integer. A number of calls were still expecting a T/F
4123 return. Fortuitously, in all cases, the tests worked in OK situations,
4124 which is the norm. However, things would have gone wrong on any write
4125 failures on the smtp file descriptor. This function is used when sending
4126 messages over SMTP and also when doing verify callouts.
4128 2. When Exim is called to do synchronous delivery of a locally submitted
4129 message (the -odf or -odi options), it no longer closes stderr before doing
4132 3. Implemented the mua_wrapper option.
4134 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4136 5. Implemented the functions header_remove(), header_testname(),
4137 header_add_at_position(), and receive_remove_recipient(), and exported them
4140 6. If an ACL "warn" statement specified the addition of headers, Exim already
4141 inserted X-ACL-Warn: at the start if there was no header name. However, it
4142 was not making this test for the second and subsequent header lines if
4143 there were newlines in the string. This meant that an invalid header could
4144 be inserted if Exim was badly configured.
4146 7. Allow an ACL "warn" statement to add header lines at the start or after all
4147 the Received: headers, as well as at the end.
4149 8. Added the rcpt_4xx retry error code.
4151 9. Added postmaster_mailfrom=xxx to callout verification option.
4153 10. Added mailfrom=xxxx to the callout verification option, for verify=
4156 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4157 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4159 12. Inserted some casts to stop certain compilers warning when using pointer
4160 differences as field lengths or precisions in printf-type calls (mostly
4161 affecting debugging statements).
4163 13. Added optional readline() support for -be (dynamically loaded).
4165 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4166 same clock tick as a message's arrival, so that its received time was the
4167 same as the "first fail" time on the retry record, and that message
4168 remained on the queue past the ultimate address timeout, every queue runner
4169 would try a delivery (because it was past the ultimate address timeout) but
4170 after another failure, the ultimate address timeout, which should have then
4171 bounced the address, did not kick in. This was a "< instead of <=" error;
4172 in most cases the first failure would have been in the next clock tick
4173 after the received time, and all would be well.
4175 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4176 being recognized when the domain list was tested by the match_domain
4177 condition in an expansion string.
4179 16. Added the ${str2b64: operator.
4181 17. Exim was always calling setrlimit() to set a large limit for the number of
4182 processes, without checking whether the existing limit was already
4183 adequate. (It did check for the limit on file descriptors.) Furthermore,
4184 errors from getrlimit() and setrlimit() were being ignored. Now they are
4185 logged to the main and panic logs, but Exim does carry on, to try to do its
4186 job under whatever limits there are.
4188 18. Imported PCRE 5.0.
4190 19. Trivial typo in log message " temporarily refused connection" (the leading
4193 20. If the log selector return_path_on_delivery was set and an address was
4194 redirected to /dev/null, the delivery process crashed because it assumed
4195 that a return path would always be set for a "successful" delivery. In this
4196 case, the whole delivery is bypassed as an optimization, and therefore no
4199 21. Internal re-arrangement: the function for sending a challenge and reading
4200 a response while authentication was assuming a zero-terminated challenge
4201 string. It's now changed to take a pointer and a length, to allow for
4202 binary data in such strings.
4204 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4206 23. Exim was not respecting finduser_retries when seeking the login of the
4207 uid under which it was called; it was always trying 10 times. (The default
4208 setting of finduser_retries is zero.) Also, it was sleeping after the final
4209 failure, which is pointless.
4211 24. Implemented tls_on_connect_ports.
4213 25. Implemented acl_smtp_predata.
4215 26. If the domain in control=submission is set empty, Exim assumes that the
4216 authenticated id is a complete email address when it generates From: or
4217 Sender: header lines.
4219 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4220 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4221 chown and chgrp in /bin and hostname in /usr/bin.
4223 28. Exim was keeping the "process log" file open after each use, just as it
4224 does for the main log. This opens the possibility of it remaining open for
4225 long periods when the USR1 signal hits a daemon. Occasional processlog
4226 errors were reported, that could have been caused by this. Anyway, it seems
4227 much more sensible not to leave this file open at all, so that is what now
4230 29. The long-running daemon process does not normally write to the log once it
4231 has entered its main loop, and it closes the log before doing so. This is
4232 so that log files can straightforwardly be renamed and moved. However,
4233 there are a couple of unusual error situations where the daemon does write
4234 log entries, and I had neglected to close the log afterwards.
4236 30. The text of an SMTP error response that was received during a remote
4237 delivery was being truncated at 512 bytes. This is too short for some of
4238 the long messages that one sometimes sees. I've increased the limit to
4241 31. It is now possible to make retry rules that apply only when a message has a
4242 specific sender, in particular, an empty sender.
4244 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4245 it possible to be selective about when SMTP synchronization is enforced.
4247 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4249 32. Implemented hosts_connection_nolog.
4251 33. Added an ACL for QUIT.
4253 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4256 35. Added mailbox_size and mailbox_filecount to appendfile.
4258 36. Added control = no_multiline_responses to ACLs.
4260 37. There was a bug in the logic of the code that waits for the clock to tick
4261 in the case where the clock went backwards by a substantial amount such
4262 that the microsecond fraction of "now" was more than the microsecond
4263 fraction of "then" (but the whole seconds number was less).
4265 38. Added support for the libradius Radius client library this is found on
4266 FreeBSD (previously only the radiusclient library was supported).
4272 1. When certain lookups returned multiple values in the form name=value, the
4273 quoting of the values was not always being done properly. Specifically:
4274 (a) If the value started with a double quote, but contained no whitespace,
4276 (b) If the value contained whitespace other than a space character (i.e.
4277 tabs or newlines or carriage returns) it was not quoted.
4278 This fix has been applied to the mysql and pgsql lookups by writing a
4279 separate quoting function and calling it from the lookup code. The fix
4280 should probably also be applied to nisplus, ibase and oracle lookups, but
4281 since I cannot test any of those, I have not disturbed their existing code.
4283 2. A hit in the callout cache for a specific address caused a log line with no
4284 reason for rejecting RCPT. Now it says "Previous (cached) callout
4285 verification failure".
4287 3. There was an off-by-one bug in the queryprogram router. An over-long
4288 return line was truncated at 256 instead of 255 characters, thereby
4289 overflowing its buffer with the terminating zero. As well as fixing this, I
4290 have increased the buffer size to 1024 (and made a note to document this).
4292 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4293 when Exim is waiting for an SMTP response from a remote server, Exim
4294 restarts its select() call on the socket, thereby resetting its timeout.
4295 This is not a problem when such interrupts are rare. Somebody set up a cron
4296 job to run exiwhat every 2 minutes, which is less than the normal select()
4297 timeout (5 or 10 minutes). This meant that the select() timeout never
4298 kicked in because it was always reset. I have fixed this by comparing the
4299 time when an interrupt arrives with the time at the start of the first call
4300 to select(). If more time than the timeout has elapsed, the interrupt is
4301 treated as a timeout.
4303 5. Some internal re-factoring in preparation for the addition of Sieve
4304 extensions (by MH). In particular, the "personal" test is moved to a
4305 separate function, and given an option for scanning Cc: and Bcc: (which is
4306 not set for Exim filters).
4308 6. When Exim created an email address using the login of the caller as the
4309 local part (e.g. when creating a From: or Sender: header line), it was not
4310 quoting the local part when it contained special characters such as @.
4312 7. Installed new OpenBSD configuration files.
4314 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4315 try to make them clearer.
4317 9. Callout options, other than the timeout value, were being ignored when
4318 verifying sender addresses in header lines. For example, when using
4320 verify = header_sender/callout=no_cache
4322 the cache was (incorrectly) being used.
4324 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4325 only the Cygwin environment.
4327 11. When return_path_on_delivery was set as a log selector, if different remote
4328 addresses in the same message used different return paths and parallel
4329 remote delivery occurred, the wrong values would sometimes be logged.
4330 (Whenever a remote delivery process finished, the return path value from
4331 the most recently started remote delivery process was logged.)
4333 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4334 lines when AUTH and/or TLS are in use. This is the "received protocol"
4335 field. Exim used to use "asmtp" for authenticated SMTP, without any
4336 indication (in the protocol name) for TLS use. Now it follows the RFC and
4337 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4338 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4339 names appear in log lines as well as in Received: header lines.
4341 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4342 extensions, and comparison tests, and to fix some bugs.
4344 14. Changes to the "personal" filter test:
4346 (1) The test was buggy in that it was just doing the equivalent of
4347 "contains" tests on header lines. For example, if a user's address was
4348 anne@some.where, the "personal" test would incorrectly be true for
4350 To: susanne@some.where
4352 This test is now done by extracting each address from the header in turn,
4353 and checking the entire address. Other tests that are part of "personal"
4354 are now done using regular expressions (for example, to check local parts
4355 of addresses in From: header lines).
4357 (2) The list of non-personal local parts in From: addresses has been
4358 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4359 taken from the Sieve specification recommendations.
4361 (3) If the message contains any header line starting with "List-" it is
4362 treated as non-personal.
4364 (4) The test for "circular" in the Subject: header line has been removed
4365 because it now seems ill-conceived.
4367 15. Minor typos in src/EDITME comments corrected.
4369 16. Installed latest exipick from John Jetmore.
4371 17. If headers_add on a router specified a text string that was too long for
4372 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4373 of string_sprintf() is now avoided.
4375 18. $message_body_size was not set (it was always zero) when running the DATA
4376 ACL and the local_scan() function.
4378 19. For the "mail" command in an Exim filter, no default was being set for
4379 the once_repeat time, causing a random time value to be used if "once" was
4380 specified. (If the value happened to be <= 0, no repeat happened.) The
4381 default is now 0s, meaning "never repeat". The "vacation" command was OK
4382 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4383 (I found it when inspecting the code).
4385 20. There is now an overall timeout for performing a callout verification. It
4386 defaults to 4 times the callout timeout, which applies to individual SMTP
4387 commands during the callout. The overall timeout applies when there is more
4388 than one host that can be tried. The timeout is checked before trying the
4389 next host. This prevents very long delays if there are a large number of
4390 hosts and all are timing out (e.g. when the network connections are timing
4391 out). The value of the overall timeout can be changed by specifying an
4392 additional sub-option for "callout", called "maxwait". For example:
4394 verify = sender/callout=5s,maxwait=20s
4396 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4397 to the end before writing, but this should make it even safer).
4399 22. Exim was forgetting that it had advertised PIPELINING for the second and
4400 subsequent messages on an SMTP connection. It was also not resetting its
4401 memory on STARTTLS and an internal HELO.
4403 23. When Exim logs an SMTP synchronization error within a session, it now
4404 records whether PIPELINING has been advertised or not.
4406 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4407 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4408 rather than long int.
4410 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4412 26. Added the never_mail option to autoreply.
4418 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4419 crash if the getsockname() call failed; this can happen if a connection is
4420 closed very soon after it is established. The problem was simply in the
4421 order in which certain operations were done, causing Exim to try to write
4422 to the SMTP stream before it had set up the file descriptor. The bug has
4423 been fixed by making things happen in the correct order.
4429 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4430 before the connection was closed, thus losing the rejection response.
4432 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4433 some early Solaris releases, but causes trouble in current releases where
4434 socklen_t is defined.
4436 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4439 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4442 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4443 that releasing the top of it at the end releases what was used for sub-
4444 expansions (unless the block got too big). However, discard this block if
4445 the first thing is a variable or header, so that we can use its block when
4446 it is dynamic (useful for very large $message_headers, for example).
4448 6. Lookups now cache *every* query, not just the most recent. A new, separate
4449 store pool is used for this. It can be recovered when all lookup caches are
4450 flushed. Lookups now release memory at the end of their result strings.
4451 This has involved some general refactoring of the lookup sources.
4453 7. Some code has been added to the store_xxx() functions to reduce the amount
4454 of flapping under certain conditions.
4456 8. log_incoming_interface used to affect only the <= reception log lines. Now
4457 it causes the local interface and port to be added to several more SMTP log
4458 lines, for example "SMTP connection from", and rejection lines.
4460 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4462 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4464 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4465 could be overwritten at the end of the current message (or the start of a
4466 new message if it was set in a HELO ACL). The value is now preserved for
4467 the duration of the SMTP connection.
4469 12. If a transport had a headers_rewrite setting, and a matching header line
4470 contained an unqualified address, that address was qualified, even if it
4471 did not match any rewriting rules. The underlying bug was that the values
4472 of the flags that permit the existence of unqualified sender and recipient
4473 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4474 non-local messages, and by -bnq for local messages) were not being
4475 preserved with the message after it was received.
4477 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4478 "next input=" as part of the text comprising the host identity instead of
4479 the correct text. The code was using the same buffer for two different
4480 strings. However, depending on which order the printing function evaluated
4481 its arguments, the bug did not always show up. Under Linux, for example, my
4482 test suite worked just fine.
4484 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4485 doesn't work with some older versions of Perl. It has been changed to "my",
4486 which in any case is probably the better facility to use.
4488 15. A really picky compiler found some instances of statements for creating
4489 error messages that either had too many or two few arguments for the format
4492 16. The size of the buffer for calls to the DNS resolver has been increased
4493 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4494 for addresses that have a lot of PTR records. This alleviates a problem; it
4495 does not fully solve it.
4497 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4498 buffer now truncates the list and logs the incident, which is the same
4499 action as happens when Exim is looking up a host name and its aliases.
4500 Previously in this situation something unpredictable would happen;
4501 sometimes it was "internal error: store_reset failed".
4503 18. If a server dropped the connection unexpectedly when an Exim client was
4504 using GnuTLS and trying to read a response, the client delivery process
4505 crashed while trying to generate an error log message.
4507 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4508 string, for example:
4510 warn message = H1: something\nH2: something
4512 the text was added as a single header line from Exim's point of view
4513 though it ended up OK in the delivered message. However, searching for the
4514 second and subsequent header lines using $h_h2: did not work. This has been
4515 fixed. Similarly, if a system filter added multiple headers in this way,
4516 the routers could not see them.
4518 20. Expanded the error message when iplsearch is called with an invalid key to
4519 suggest using net-iplsearch in a host list.
4521 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4522 ACLs are no longer actually imposed (and a message to that effect is
4525 22. If a "gecos" field in a passwd entry contained escaped characters, in
4526 particular, if it contained a \" sequence, Exim got it wrong when building
4527 a From: or a Sender: header from that name. A second bug also caused
4528 incorrect handling when an unquoted " was present following a character
4529 that needed quoting.
4531 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4532 was not being matched caselessly.
4534 24. Arranged for all hyphens in the exim.8 source to be escaped with
4537 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4538 properly. Recipient callout cache records were still being keyed to include
4539 the sender, even when use_sender was set false. This led to far more
4540 callouts that were necessary. The sender is no longer included in the key
4541 when use_sender is false.
4543 26. Added "control = submission" modifier to ACLs.
4545 27. Added the ${base62d: operator to decode base 62 numbers.
4547 28. dnsdb lookups can now access SRV records.
4549 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4550 the configuration file.
4552 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4553 (-v) mode. This makes the output for a verbose queue run more intelligible.
4555 31. Added a use_postmaster feature to recipient callouts.
4557 32. Added the $body_zerocount variable, containing the number of binary zero
4558 bytes in the message body.
4560 33. The time of last modification of the "new" subdirectory is now used as the
4561 "mailbox time last read" when there is a quota error for a maildir
4564 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4566 35. Added +ignore_unknown as a special item in host lists.
4568 36. Code for decoding IPv6 addresses in host lists is now included, even if
4569 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4570 address was recognized as an IP address, but was then not correctly decoded
4571 into binary, causing unexpected and incorrect effects when compared with
4578 1. Very minor rewording of debugging text in manualroute to say "list of
4579 hosts" instead of "hostlist".
4581 2. If verify=header_syntax was set, and a header line with an unqualified
4582 address (no domain) and a large number of spaces between the end of the
4583 name and the colon was received, the reception process suffered a buffer
4584 overflow, and (when I tested it) crashed. This was caused by some obsolete
4585 code that should have been removed. The fix is to remove it!
4587 3. When running in the test harness, delay a bit after writing a bounce
4588 message to get a bit more predictability in the log output.
4590 4. Added a call to search_tidyup() just before forking a reception process. In
4591 theory, someone could use a lookup in the expansion of smtp_accept_max_
4592 per_host which, without the tidyup, could leave open a database connection.
4594 5. Added the variables $recipient_data and $sender_data which get set from a
4595 lookup success in an ACL "recipients" or "senders" condition, or a router
4596 "senders" option, similar to $domain_data and $local_part_data.
4598 6. Moved the writing of debug_print from before to after the "senders" test
4601 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4602 problems for message scanning, either using a data ACL, or using
4603 local_scan() because the Received: header was not generated till after they
4604 were called (in order to set the time as the time of reception completion).
4605 I have revised the way this works. The header is now generated after the
4606 body is received, but before the ACL or local_scan() are called. After they
4607 are run, the timestamp in the header is updated.
4613 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4614 before starting a queue runner without re-exec. This happened only when
4615 deliver_drop_privilege was set or when the Exim user was set to root. The
4616 effect of the bug was that timeouts during subsequent deliveries caused
4617 crashes instead of being properly handled. The handler is now left at its
4618 default (and expected) setting.
4620 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4621 message, again when deliver_drop_privilege is set or Exim is run as root.
4622 The bug described in (1) was not present in this case, but the tidying up
4623 of the other signals was missing. I have made the two cases consistent.
4625 3. The ignore_target_hosts setting on a manualroute router was being ignored
4626 for hosts that were looked up using the /MX notation.
4628 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4631 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4632 operated on the sender address. After changing the $sender_address to <>
4633 for the sender address verify, Exim was re-instated it as the original
4634 (before rewriting) address, but remembering that it had rewritten it, so it
4635 wasn't rewriting it again. This bug also had the effect of breaking the
4636 sender address verification caching when the sender address was rewritten.
4638 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4639 This has been changed so that if the ip literal address matches
4640 ignore_target_hosts, the router declines.
4642 7. Added expansion conditions match_domain, match_address, and match_local_
4643 part (NOT match_host).
4645 8. The placeholder for the Received: header didn't have a length field set.
4647 9. Added code to Exim itself and to exim_lock to test for a specific race
4648 condition that could lead to file corruption when using MBX delivery. The
4649 issue is with the lockfile that is created in /tmp. If this file is removed
4650 after a process has opened it but before that process has acquired a lock,
4651 there is the potential for a second process to recreate the file and also
4652 acquire a lock. This could lead to two Exim processes writing to the file
4653 at the same time. The added code performs the same test as UW imapd; it
4654 checks after acquiring the lock that its file descriptor still refers to
4655 the same named file.
4657 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4658 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4661 11. Added the smtp_active_hostname option. If used, this will typically be made
4662 to depend on the incoming interface address. Because $interface_address is
4663 not set up until the daemon has forked a reception process, error responses
4664 that can happen earlier (such as "too many connections") no longer contain
4667 12. If an expansion in a condition on a "warn" statement fails because a lookup
4668 defers, the "warn" statement is abandoned, and the next ACL statement is
4669 processed. Previously this caused the whole ACL to be aborted.
4671 13. Added the iplsearch lookup type.
4673 14. Added ident_timeout as a log selector.
4675 15. Added tls_certificate_verified as a log selector.
4677 16. Added a global option tls_require_ciphers (compare the smtp transport
4678 option of the same name). This controls incoming TLS connections.
4680 17. I finally figured out how to make tls_require_ciphers do a similar thing
4681 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4682 before starting the TLS session.
4684 18. Tabs are now shown as \t in -bP output.
4686 19. If the log selector return_path_on_delivery was set, Exim crashed when
4687 bouncing a message because it had too many Received: header lines.
4689 20. If two routers both had headers_remove settings, and the first one included
4690 a superfluous trailing colon, the final name in the first list and the
4691 first name in the second list were incorrectly joined into one item (with a
4692 colon in the middle).
4698 1. Added -C and -D options to the exinext utility, mainly to make it easier
4699 to include in the automated testing, but these could be helpful when
4700 multiple configurations are in use.
4702 2. The exinext utility was not formatting the output nicely when there was
4703 an alternate port involved in the retry record key, nor when there was a
4704 message id as well (for retries that were specific to a specific message
4705 and a specific host). It was also confused by IPv6 addresses, because of
4706 the additional colons they contain. I have fixed the IPv4 problem, and
4707 patched it up to do a reasonable job for IPv6.
4709 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4710 delivery, the log line now contains "pipelined" if PIPELINING was used.
4712 4. An SMTP transport process used to panic and die if the bind() call to set
4713 an explicit outgoing interface failed. This has been changed; it is now
4714 treated in the same way as a connect() failure.
4716 5. A reference to $sender_host_name in the part of a conditional expansion
4717 that was being skipped was still causing a DNS lookup. This no longer
4720 6. The def: expansion condition was not recognizing references to header lines
4721 that used bh_ and bheader_.
4723 7. Added the _cache feature to named lists.
4725 8. The code for checking quota_filecount in the appendfile transport was
4726 allowing one more file than it should have been.
4728 9. For compatibility with Sendmail, the command line option
4736 and sets the incoming protocol and host name (for trusted callers). The
4737 host name and its colon can be omitted when only the protocol is to be set.
4738 Note the Exim already has two private options, -pd and -ps, that refer to
4739 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4740 "s", but I don't think that's a major issue.
4742 10. A number of refactoring changes to the code, none of which should affect
4745 (a) The number of logging options was getting close to filling up the
4746 32-bit word that was used as a bit map. I have split them into two classes:
4747 those that are passed in the argument to log_write(), and those that are
4748 only ever tested independently outside of that function. These are now in
4749 separate 32-bit words, so there is plenty of room for expansion again.
4750 There is no change in the user interface or the logging behaviour.
4752 (b) When building, for example, log lines, the code previously used a
4753 macro that called string_cat() twice, in order to add two strings. This is
4754 not really sufficiently general. Furthermore, there was one instance where
4755 it was actually wrong because one of the argument was used twice, and in
4756 one call a function was used. (As it happened, calling the function twice
4757 did not affect the overall behaviour.) The macro has been replaced by a
4758 function that can join an arbitrary number of extra strings onto a growing
4761 (c) The code for expansion conditions now uses a table and a binary chop
4762 instead of a serial search (which was left over from when there were very
4763 few conditions). Also, it now recognizes conditions like "pam" even when
4764 the relevant support is not compiled in: a suitably worded error message is
4765 given if an attempt is made to use such a condition.
4767 11. Added ${time_interval:xxxxx}.
4769 12. A bug was causing one of the ddress fields not to be passed back correctly
4770 from remote delivery subprocesses. The field in question was not being
4771 subsequently used, so this caused to problems in practice.
4773 13. Added new log selectors queue_time and deliver_time.
4775 14. Might have fixed a bug in maildirsizefile handling that threw up
4776 "unexpected character" debug warnings, and recalculated the data
4777 unnecessarily. In any case, I expanded the warning message to give more
4780 15. Added the message "Restricted characters in address" to the statements in
4781 the default ACL that block characters like @ and % in local parts.
4783 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4784 Three changes have been made:
4786 (a) There was a serious bug; a negative response to MAIL caused the whole
4787 recipient domain to be cached as invalid, thereby blocking all messages
4788 to all local parts at the same domain, from all senders. This bug has
4789 been fixed. The domain is no longer cached after a negative response to
4790 MAIL if the sender used is not empty.
4792 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4795 (c) A new callout option, "use_sender" has been added for people who want
4796 the modified behaviour.
4802 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4805 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4806 indeed breaks things for older releases.
4808 3. Added additional logging to the case where there is a problem reading data
4809 from a filter that is running in a subprocess using a pipe, in order to
4810 try to track down a specific problem.
4812 4. Testing facility fudge: when running in the test harness and attempting
4813 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4814 getting "No route to host". Convert this to a timeout.
4816 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4819 6. Some OS don't have socklen_t but use size_t instead. This affects the
4820 fifth argument of getsockopt() amongst other things. This is now
4821 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4822 can be set for individual OS. I have set it for SunOS5, OSF1, and
4823 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4824 some earlier ones do not.
4826 7. Change 4.30/15 was not doing the test caselessly.
4828 8. The standard form for an IPv6 address literal was being rejected by address
4829 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4830 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4831 this, as well as the form without the "IPv6" on the front (but only when
4832 address literals are enabled, of course).
4834 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4836 10. Exim crashed if a message with an empty sender address specified by -f
4837 encountered a router with an errors_to setting. This could be provoked only
4838 by a command such as
4842 where an empty string was supplied; "<>" did not hit this bug.
4844 11. Installed PCRE release 4.5.
4846 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4847 remained set. It is now erased.
4849 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4850 times from message ids (which are base 36 rather than the normal 62).
4852 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4853 were being counted as actual protocol errors, and logged if the log
4854 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4855 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4856 and DATA following a set of rejected RCPTs do not count as protocol errors.
4857 In other words, Exim assumes they were pipelined, though this may not
4858 actually be the case. Of course, in all cases the client gets an
4859 appropriate error code.
4861 15. If a lookup fails in an ACL condition, a message about the failure may
4862 be available; it is used if testing the ACL cannot continue, because most
4863 such messages specify what the cause of the deferral is. However, some
4864 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4865 that caused an old message to be retained and used if a later statement
4866 caused a defer, replacing the real cause of the deferral.
4868 16. If an IP address had so many PTR records that the DNS lookup buffer
4869 was not large enough to hold them, Exim could crash while trying to process
4870 the truncated data. It now detects and logs this case.
4872 17. Further to 4.21/58, another change has been made: if (and only if) the
4873 first line of a message (the first header line) ends with CRLF, a bare LF
4874 in a subsequent header line has a space inserted after it, so as not to
4875 terminate the header.
4877 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4878 unnecessarily, used atoi() instead of strtol(), and didn't check the
4879 termination when getting file sizes from file names by regex.
4881 19. Completely re-implemented the support for maildirsize files, in the light
4882 of a number of problems with the previous contributed implementation
4883 (4.30/29). In particular:
4885 . If the quota is zero, the maildirsize file is maintained, but no quota is
4888 . If the maildir directory does not exist, it is created before any attempt
4889 to write a maildirsize file.
4891 . The quota value in the file is just a cache; if the quota is changed in
4892 the transport, the new value overrides.
4894 . A regular expression is available for excluding directories from the
4897 20. The autoreply transport checks the characters in options that define the
4898 message's headers; it allows continued headers, but it was checking with
4899 isspace() after an embedded newline instead of explicitly looking for a
4902 21. If all the "regular" hosts to which an address was routed had passed their
4903 expiry times, and had not reached their retry times, the address was
4904 bounced, even if fallback hosts were defined. Now Exim should go on to try
4907 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4908 equivalent code in the SMTP transport. Some hosts send humungous responses
4909 to HELO/EHLO, more than 1024 it seems.
4911 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4912 gives compiler warnings in some environments. I've now done it "properly",
4915 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4916 (because of problems with the built-in one) was declared to return uschar *
4917 instead of char *, causing compiler failure.
4919 25. Fixed a file descriptor leak when processing alias/forward files.
4921 26. Fixed a minor format string issue in dbfn.c.
4923 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4925 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4926 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4929 29. When a sender address is verified, it is cached, to save repeating the test
4930 when there is more than one recipient in a message. However, when the
4931 verification involves a callout, it is possible for different callout
4932 options to be set for different recipients. It is too complicated to keep
4933 track of this in the cache, so now Exim always runs a verification when a
4934 callout is required, relying on the callout cache for the optimization.
4935 The overhead is duplication of the address routing, but this should not be
4938 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4939 to be verified with callout=postmaster, and the main callout worked but the
4940 postmaster check failed, the verification correctly failed. However, if a
4941 subsequent RCPT command asked for sender verification *without* the
4942 postmaster check, incorrect caching caused this verification also to fail,
4945 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4946 it was not caching the DNS options (qualify_single, search_parents) that
4947 were used when the lookup failed. A subsequent lookup with different
4948 options therefore always gave the same answer, though there were cases
4949 where it should not have. (Example: a "domains = !$mx_any" option on a
4950 dnslookup router: the "domains" option is always processed without any
4951 widening, but the router might have qualify_single set.) Now Exim uses the
4952 cached value only when the same options are set.
4954 32. Added John Jetmore's "exipick" utility to the distribution.
4956 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4957 than a timeout (e.g. a certificate is required, and is not provided), an
4958 Exim server now closes the connection immediately. Previously it waited for
4959 the client to close - but if the client is SSL, it seems that they each
4960 wait for each other, leading to a delay before one of them times out.
4962 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4963 maintained 0.8.x compatibility because I don't think many are using it, and
4964 it is clearly obsolete.
4966 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4969 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4970 client certificate was expired. A simple patch fixes this, though I don't
4971 understand the full logic of why the verify callback is called multiple
4974 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4975 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4976 which causes problems with some clients (such as the Certicom SSL Plus
4977 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4978 disables the coutermeasure allowing Eudora to connect."
4980 38. Exim was not checking that a write() to a log file succeeded. This could
4981 lead to Bad Things if a log got too big, in particular if it hit a file
4982 size limit. Exim now panics and dies if it cannot write to a log file, just
4983 as it does if it cannot open a log file.
4985 39. Modified OS/Makefile-Linux so that it now contains
4987 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4989 The two -D definitions ensure that Exim is compiled with large file
4990 support, which makes it possible to handle log files that are bigger than
4993 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4994 instance) a domain was checked against a named list that involved a lookup,
4995 causing $domain_data to be set, then another domain was checked against the
4996 same list, then the first domain was re-checked, the value of $domain_data
4997 after the final check could be wrong. In particular, if the second check
4998 failed, it could be set empty. This bug probably also applied to
5001 41. The strip_trailing_dot option was not being applied to the address given
5002 with the -f command-line option.
5004 42. The code for reading a message's header from the spool was incrementing
5005 $received_count, but never initializing it. This meant that the value was
5006 incorrect (doubled) while delivering a message in the same process in which
5007 it was received. In the most common configuration of Exim, this never
5008 happens - a fresh exec is done - but it can happen when
5009 deliver_drop_privilege is set.
5011 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5012 it now includes up to 150 characters of the unexpected data in the log
5015 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5016 and building data strings. The size of both of these buffers was 10 000
5017 bytes - far larger than anybody would *ever* want, thought I. Needless to
5018 say, somebody hit the limit. I have increased the maximum line length to
5019 20 000 and the maximum data length of concatenated lines to 100 000. I have
5020 also fixed two bugs, because there was no checking on these buffers. Tsk,
5021 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5022 buffer is too small.
5024 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5025 lsearch lookups. Now it does.
5027 46. When parsing a route_list item in a manualroute router, a fixed-length
5028 buffer was used for the list of hosts. I made this 1024 bytes long,
5029 thinking that nobody would ever have a list of hosts that long. Wrong.
5030 Somebody had a whole pile of complicated expansion conditions, and the
5031 string was silently truncated, leading to an expansion error. It turns out
5032 that it is easier to change to an unlimited length (owing to other changes
5033 that have happened since this code was originally written) than to build
5034 structure for giving a limitation error. The length of the item that
5035 expands into the list of hosts is now unlimited.
5037 47. The lsearch lookup could not handle data where the length of text line was
5038 more than 4095 characters. Such lines were truncated, leading to shortened
5039 data being returned. It should now handle lines of any length.
5041 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5042 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5045 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5046 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5047 no longer happens in comment lines. A list of replacements is now placed
5048 at the head of all of the source files, except those whose only change is
5049 to replace PERL_COMMAND in the very first #! line.
5051 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5052 messages on the queue, with a bottom-up merge sort, using code contributed
5053 by Michael Haardt. This should make operations like -bp somewhat faster on
5054 large queues. It won't affect queue runners, except when queue_run_in_order
5057 51. Installed eximstats 1.31 in the distribution.
5059 52. Added support for SRV lookups to the dnslookup router.
5061 53. If an ACL referred to $message_body or $message_body_end, the value was not
5062 reset for any messages that followed in the same SMTP session.
5064 54. The store-handling optimization for building very long strings was not
5065 differentiating between the different store pools. I don't think this
5066 actually made any difference in practice, but I've tidied it.
5068 55. While running the routers to verify a sender address, $sender_address
5069 was still set to the sender address. This is wrong, because when routing to
5070 send a bounce to the sender, it would be empty. Therefore, I have changed
5071 it so that, while verifying a sender address, $sender_address is set to <>.
5072 (There is no change to what happens when verifying a recipient address.)
5074 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5075 target A or AAAA records (if not already returned) without resetting the
5076 qualify_single or search_parents options of the DNS resolver. These are
5077 inappropriate in this case because the targets of MX and SRV records must
5078 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5079 target that, when qualified, matched something in the local domain. These
5080 two options are now turned off when doing these lookups.
5082 57. It seems that at least some releases of Reiserfs (which does not have the
5083 concept of a fixed number of inodes) returns zero and not -1 for the
5084 number of available inodes. This interacted badly with check_spool_inodes,
5085 which assumed that -1 was the "no such thing" setting. What I have done is
5086 to check that the total number of inodes is greater than zero before doing
5087 the test of how many are available.
5089 58. When a "warn" ACL statement has a log_message modifier, the message is
5090 remembered, and not repeated. This is to avoid a lot of repetition when a
5091 message has many recipients that cause the same warning to be written.
5092 Howewer, Exim was preserving the list of already written lines for an
5093 entire SMTP session, which doesn't seem right. The memory is now reset if a
5094 new message is started.
5096 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5097 debugging output unless log_rewrite was also set.
5099 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5100 of (int)(handle) when we know that handle contains (void *)(-1).
5102 61. The Exim daemon panic-logs an error return when it closes the incoming
5103 connection. However "connection reset by peer" seems to be common, and
5104 isn't really an error worthy of noting specially, so that particular error
5107 62. When Exim is trying to find all the local interfaces, it used to panic and
5108 die if the ioctl to get the interface flags failed. However, it seems that
5109 on at least one OS (Solaris 9) it is possible to have an interface that is
5110 included in the list of interfaces, but for which you get a failure error
5111 for this call. This happens when the interface is not "plumbed" into a
5112 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5113 failure of the "get flags" call assumes that the interface is down.
5115 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5116 makes life easier for people who are doing arithmetic on fields extracted
5117 from dates, where you often get leading zeros that should not be
5118 interpreted as octal.
5120 64. Added qualify_domain to the redirect router, to override the global
5123 65. If a pathologically long header line contained very many addresses (the
5124 report of this problem mentioned 10 000) and each of them was rewritten,
5125 Exim could use up a very large amount of memory. (It kept on making new
5126 copies of the header line as it rewrote, and never released the old ones.)
5127 At the expense of a bit more processing, the header rewriting function has
5128 been changed so that it no longer eats memory in this way.
5130 66. The generation of the Received: header has been moved from the time that a
5131 message starts to be received, to the time that it finishes. The timestamp
5132 in the Received: header should now be very close to that of the <= log
5133 line. There are two side-effects of this change:
5135 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5136 logged header lines no longer include the local Received: line, because
5137 it has not yet been created. The same applies to a copy of the message
5138 that is returned to a non-SMTP sender when a message is rejected.
5140 (b) When a filter file is tested using -bf, no additional Received: header
5141 is added to the test message. After some thought, I decided that this
5144 This change does not affect the value of $received_for. It is still set
5145 after address rewriting, but before local_scan() is called.
5147 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5149 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5150 gave an unhelpful panic error message, and a defer error. I have managed to
5151 change this behaviour so that it now rejects any supplied certificate,
5152 which seems right, as the list of acceptable certificates is empty.
5154 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5155 gave an unhelpful defer error. I have not managed to make this reject any
5156 supplied certificates, but the error message it gives is "no certificate
5157 supplied", which is not helpful.
5159 70. exigrep's output now also includes lines that are not associated with any
5160 message, but which match the given pattern. Implemented by a patch from
5161 Martin Sluka, which also tidied up the Perl a bit.
5163 71. Recipient callout verification, like sender verification, was using <> in
5164 the MAIL FROM command. This isn't really the right thing, since the actual
5165 sender may affect whether the remote host accepts the recipient or not. I
5166 have changed it to use the actual sender in the callout; this means that
5167 the cache record is now keyed on a recipient/sender pair, not just the
5168 recipient address. There doesn't seem to be a real danger of callout loops,
5169 since a callout by the remote host to check the sender would use <>.
5170 [SEE ABOVE: changed after hitting problems.]
5172 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5173 temporary errors. However, in the case of such a code being given after
5174 the end of a data transmission (i.e. after ".") Exim was failing to write
5175 a retry record for the message. (Yes, there was some broken host that was
5176 actually sending 8xx at this point.)
5178 73. An unknown lookup type in a host list could cause Exim to panic-die when
5179 the list was checked. (An example that provoked this was putting <; in the
5180 middle of a list instead of at the start.) If this happened during a DATA
5181 ACL check, a -D file could be left lying around. This kind of configuration
5182 error no longer causes Exim to die; instead it causes a defer errror. The
5183 incident is still logged to the main and panic logs.
5185 74. Buglet left over from Exim 3 conversion. The message "too many messages
5186 in one connection" was written to the rejectlog but not the mainlog, except
5187 when address rewriting (yes!) was being logged.
5189 75. Added write_rejectlog option.
5191 76. When a system filter was run not as root (that is, when system_filter_user
5192 was set), the values of the $n variables were not being returned to the
5193 main process; thus, they were not subsequently available in the $sn
5196 77. Added +return_path_on_delivery log selector.
5198 78. A connection timeout was being treated differently from recipients deferred
5199 when testing hosts_max_try with a message that was older than the host's
5200 retry timeout. (The host should not be counted, thus allowing all hosts to
5201 be tried at least once before bouncing.) This may have been the cause of an
5202 occasionally reported bug whereby a message would remain on the queue
5203 longer than the retry timeout, but would be bounced if a delivery was
5204 forced. I say "may" because I never totally pinned down the problem;
5205 setting up timeout/retry tests is difficult. See also the next item.
5207 79. The ultimate address timeout was not being applied to errors that involved
5208 a combination of host plus message (for example, a timeout on a MAIL
5209 command). When an address resolved to a number of possible hosts, and they
5210 were not all tried for each delivery (e.g. because of hosts_max_try), a
5211 message could remain on the queue longer than the retry timeout.
5213 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5216 81. Fixed an obscure SMTP outgoing bug which required at least the following
5217 conditions: (a) there was another message waiting for the same server;
5218 (b) the server returned 5xx to all RCPT commands in the first message so
5219 that the message was not completed; (c) the server dropped the connection
5220 or gave a negative response to the RSET that Exim sends to abort the
5221 transaction. The observed case was a dropped connection after DATA that had
5222 been sent in pipelining mode. That is, the server had advertised PIPELINING
5223 but was not implementing it correctly. The effect of the bug was incorrect
5224 behaviour, such as trying another host, and this could lead to a crash.
5230 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5231 and daemon.c were passed as pointers to ints; they should have been
5232 pointers to socklen_t variables (which are typically unsigned ints).
5234 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5237 3. Fixed a really odd bug that affected only the testing scheme; patching a
5238 certain fixed string in the binary changed the value of another string that
5239 happened to be identical to the end of the original first string.
5241 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5242 name", it returns that address as the IP address. On some operating
5243 systems (e.g. Solaris), it also passes back the IP address string as the
5244 "host name". However, on others (e.g. Linux), it passes back an empty
5245 string. Exim wasn't checking for this, and was changing the host name to an
5246 empty string, assuming it had been canonicized.
5248 5. Although rare, it is permitted to have more than one PTR record for a given
5249 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5250 all the names associated with an address, because they do in Solaris.
5251 However, it seems that they do not in Linux for data that comes from the
5252 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5253 I found this out when I moved to a new Linux workstation and tried to run
5254 the Exim test suite.
5256 To get round this problem I have changed the code so that it now does its
5257 own call to the DNS to look up PTR records when searching for a host name.
5258 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5259 addresses that are only in /etc/hosts are still found.
5261 This behaviour is, however, controlled by an option called host_lookup_
5262 order, which defaults to "bydns:byaddr". If people want to use the other
5263 order, or indeed, just use one or the other means of lookup, they can
5264 specify it in this variable.
5266 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5267 some operating systems, this comes back from gethostbyaddr() as an empty
5268 string, and this is what Exim used to test for. However, it seems that in
5269 other systems, "." is yielded. Exim now tests for this case too.
5271 7. The values of check_spool_space and check_log_space are now held internally
5272 as a number of kilobytes instead of an absolute number of bytes. If a
5273 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5274 kilobyte. This means that much larger values can be stored.
5276 8. Exim monitor: an attempt to get the action menu when not actually pointing
5277 at a message produces an empty menu entitled "No message selected". This
5278 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5279 no entries in it ("Shell widget menu has zero width and/or height"). So I
5280 have added a single, blank menu entry in this case.
5282 9. Added ${quote_local_part.
5284 10. MIME decoding is now applied to the contents of Subject: header lines when
5287 11. Now that a reference to $sender_host_address automatically causes a reverse
5288 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5289 host lookup before query-style lookups in lists that might use this
5290 variable. This has therefore been abolished, and the "net-" prefix is no
5291 longer necessary for query-style lookups.
5293 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5294 have been a typo for LFLAGS, so it has been changed.
5296 13. The install script calls Exim with "-C /dev/null" in order to find the
5297 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5298 to be output. Howeve, since Exim outputs its version number before the
5299 error, it didn't break the script. It just looked ugly. I fixed this by
5300 always allowing "-C /dev/null" if the caller is root.
5302 14. Ignore overlarge ACL variable number when reading spool file - insurance
5303 against a later release with more variables having written the file.
5305 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5306 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5307 this, as well as the form without the "IPv6" on the front.
5309 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5310 OS/Makefile-Darwin file.
5312 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5313 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5315 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5316 variable. However, some OS use ints for inodes. I've added cast to long int
5317 to get rid of the compiler warning.
5319 19. I had forgotten to lock out "/../" in configuration file names when
5320 ALT_CONFIG_PREFIX was set.
5322 20. Routers used for verification do not need to specify transports. However,
5323 if such a router generated a host list, and callout was configured, Exim
5324 crashed, because it could not find a port number from the (non-existent)
5325 transport. It now assumes port 25 in this circumstance.
5327 21. Added the -t option to exigrep.
5329 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5330 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5331 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5332 them is set, LOOKUP_LSEARCH is forced.
5334 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5336 24. Added sender and host information to the "rejected by local_scan()" log
5337 line; previously there was no indication of these.
5339 25. Added .include_if_exists.
5341 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5342 Linux. It turns out that not all file systems support this. Apparently some
5343 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5344 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5345 not supported on the file descriptor, is now ignored when Exim is trying to
5346 sync a directory. This applies only to Linux.
5348 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5350 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5351 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5354 29. Added support for maildirsize files from supplied patch (modified a bit).
5356 30. The use of :fail: followed by an empty string could lead Exim to respond to
5357 sender verification failures with (e.g.):
5359 550 Verification failed for <xxx>
5360 550 Sender verify failed
5362 where the first response line was missing the '-' that indicates it is not
5363 the final line of the response.
5365 31. The loop for finding the name of the user that called Exim had a hardwired
5366 limit of 10; it now uses the value of finduser_retries, which is used for
5367 all other user lookups.
5369 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5372 33. Exim was neglecting to zero errno before one call of strtol() when
5373 expanding a string and expecting an integer value. On some systems this
5374 resulted in spurious "integer overflow" errors. Also, it was casting the
5375 result into an int without checking.
5377 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5378 did not work. The code looks as if it has *never* worked, though it appears
5379 to have been documented since at least releast 1.62. I have made it work.
5381 35. The "timeout_DNS" error in retry rules, also documented since at least
5382 1.62, also never worked. As it isn't clear exactly what this means, and
5383 clearly it isn't a major issue, I have abolished the feature by treating it
5384 as "timeout", and writing a warning to the main and panic logs.
5386 36. The display of retry rules for -brt wasn't always showing the error code
5389 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5390 timeout_connect_A, timeout_connect_MX.
5392 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5393 to the empty sender.
5395 39. The daemon was not analysing the content of -oX till after it had closed
5396 stderr and disconnected from the controlling terminal. This meant that any
5397 syntax errors were only noted on the panic log, and the return code from
5398 the command was 0. By re-arranging the code a little, I've made the
5399 decoding happen first, so such errors now appear on stderr, and the return
5400 code is 1. However, the actual setting up of the sockets still happens in
5401 the disconnected process, so errors there are still only recorded on the
5404 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5405 connections (as happens on some IP stacks) was logged at start up time as
5406 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5407 it from "IPv6 and IPv4", which means that two separate sockets are being
5410 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5411 says whether AF_INET or AF_INET6 was passed as an argument.
5413 42. Exiwhat output was messed up when time zones were included in log
5416 43. Exiwhat now gives more information about the daemon's listening ports,
5417 and whether -tls-on-connect was used.
5419 44. The "port" option of the smtp transport is now expanded.
5421 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5422 silently ignored. Now an error message is written to the main and panic
5425 46. There's a new ACL modifier called "logwrite" which writes to a log file
5426 as soon as it is encountered.
5428 47. Added $local_user_uid and $local_user_gid at routing time.
5430 48. Exim crashed when trying to verify a sender address that was being
5433 49. Exim was recognizing only a space character after ".include". It now also
5434 recognizes a tab character.
5436 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5437 extracting the relevant information from the specification. The man page no
5438 longer contains scrambled data for the -d option, and I've added a section
5439 at the front about calling Exim under different names.
5441 51. Added "extra_headers" argument to the "mail" command in filter files.
5443 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5446 53. Installed eximstats 1.29.
5448 54. Added transport_filter_timeout as a generic transport option.
5450 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5451 Cc: header lines. This was required by RFC 822, but it not required by RFC
5454 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5455 incoming messages that did not have them. Now it does so only if the
5456 message originates locally, that is, if there is no associated remote host
5457 address. When Resent- header lines are present, this applies to the Resent-
5458 lines rather than the non-Resent- lines.
5460 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5461 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5463 58. Messages for configuration errors now include the name of the main
5464 configuration file - useful now that there may be more than one file in a
5465 list (.included file names were always shown).
5467 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5468 for those rare installations that do not start the daemon as root or run it
5469 setuid root. I've cut out the call to initgroups() if the daemon is not
5472 60. The Exim user and group can now be bound into the binary as text strings
5473 that are looked up at the start of Exim's processing.
5475 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5477 62. Added $mailstore_basename variable.
5479 63. Installed patch to sieve.c from Michael Haardt.
5481 64. When Exim failed to open the panic log after failing to open the main log,
5482 the original message it was trying to log was written to stderr and debug
5483 output, but if they were not available (the usual case in production), it
5484 was lost. Now it is written to syslog before the two lines that record the
5485 failures to open the logs.
5487 65. Users' Exim filters run in subprocesses under the user's uid. It is
5488 possible for a "deliver" command or an alias in a "personal" command to
5489 provoke an address rewrite. If logging of address rewriting is configured,
5490 this fails because the process is not running as root or exim. There may be
5491 a better way of dealing with this, but for the moment (because 4.30 needs
5492 to be released), I have disabled address rewrite logging when running a
5493 filter in a non-root, non-exim process.
5499 1. The buildconfig auxiliary program wasn't quoting the value set for
5500 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5501 not defined. This bug was masked in 4.22 by the effect that was fixed in
5504 2. Some messages that were rejected after a message id was allocated were
5505 shown as "incomplete" by exigrep. It no longer does this for messages that
5506 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5508 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5509 have allow_domain_literals set, the ID did not get logged in the <= line.
5510 Domain literals are now always recognized in Message-ID: header lines.
5512 4. The first argument for a ${extract expansion item is the key name or field
5513 number. Leading and trailing spaces in this item were not being ignored,
5514 causing some misleading effects.
5516 5. When deliver_drop_privilege was set, single queue runner processes started
5517 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5518 same command in the process it spins off) were not dropping privilege.
5520 6. When the daemon running as "exim" started a queue runner, it always
5521 re-executed Exim in the spun-off process. This is a waste of effort when
5522 deliver_drop_privilege is set. The new process now just calls the
5523 queue-runner function directly.
5529 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5532 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5533 ignored. Though the use of -oP was forcing the writing of a pid file, it
5534 was always written to the default place.
5536 3. If the message "no IP address found for host xxxx" is generated during
5537 incoming verification, it is now followed by identification of the incoming
5538 connection (so you can more easily find what provoked it).
5540 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5542 5. Added some features to "harden" Exim a bit more against certain attacks:
5544 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5545 be put in Local/Makefile. This is like the never_users runtime option,
5546 but it cannot be overridden. The default setting is "root".
5548 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5549 prefix string with which any file named in a -C command line option
5552 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5553 is retained for -C and -D only if the caller of Exim is root. Without
5554 it, the exim user may also use -C and -D and retain privilege.
5556 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5557 command line option is disabled.
5559 6. Macro names set by the -D option must start with an upper case letter, just
5560 like macro names defined in the configuration file.
5562 7. Added "dereference=" facility to LDAP.
5564 8. Two instances of the typo "uknown" in the source files are fixed.
5566 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5567 the Configure-Makefile script screwed up while processing it.
5569 10. Incorporated PCRE 4.4.
5571 11. The SMTP synchronization check was not operating right at the start of an
5572 SMTP session. For example, it could not catch a HELO sent before the client
5573 waited for the greeting. There is now a check for outstanding input at the
5574 point when the greeting is written. Because of the duplex, asynchronous
5575 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5576 way, but not yet received, when the check is performed.
5578 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5579 on TCP/IP sockets, because this apparently causes some broken clients to
5582 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5583 unchanged) from the Cygwin maintainer.
5585 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5586 was supported instead of testing for mbx. Effectively a typo.
5588 15. The spa authenticator server code was not checking that the input it
5589 received was valid base64.
5591 16. The debug output line for the "set" modifier in ACLs was not showing the
5592 name of the variable that was being set.
5594 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5596 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5597 lookup on its own. Something else was needed to trigger the lookup. For
5598 example, a match in host_lookup or the need for a host name in a host list.
5599 Now, if $sender_host_name is referenced and the host name has not yet been
5600 looked up, a lookup is performed. If the lookup fails, the variable remains
5601 empty, and $host_lookup_failed is set to "1".
5603 19. Added "eqi" as a case-independent comparison operator.
5605 20. The saslauthd authentication condition could segfault if neither service
5606 nor realm was specified.
5608 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5609 error message that was logged was misleading, and incoming SMTP
5610 connections were dropped. The message is now more accurate, and temporary
5611 errors are given to SMTP connections.
5613 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5614 (except RSET) with a 421 error, until QUIT is received. However, it was
5615 failing to send a response to QUIT.
5617 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5618 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5619 ACL is now run after the helo_try_verify_hosts code.
5621 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5622 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5623 case-independent, so other case variants are also recognized). Apparently
5624 some systems use these upper case variants.
5626 25. If more than two messages were waiting for the same host, and a transport
5627 filter was specified for the transport, Exim sent two messages over the
5628 same TCP/IP connection, and then failed with "socket operation on non-
5629 socket" when it tried to send the third.
5631 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5633 27. The extern definition of crypt16() in expand.c was not being excluded when
5634 the OS had its own crypt16() function.
5636 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5637 as a preferred synonym for return_size_limit, both as an option and as an
5640 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5642 30. Changed the default configuration ACL to relax the local part checking rule
5643 for addresses that are not in any local domains. For these addresses,
5644 slashes and pipe symbols are allowed within local parts, but the sequence
5645 /../ is explicitly forbidden.
5647 31. SPA server authentication was not clearing the challenge buffer before
5650 32. log_message in a "warn" ACL statement was writing to the reject log as
5651 well as to the main log, which contradicts the documentation and doesn't
5652 seem right (because no rejection is happening). So I have stopped it.
5654 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5655 However, I am unable to do any testing of this.
5657 34. Fixed an infelicity in the appendfile transport. When checking directories
5658 for a mailbox, to see if any needed to be created, it was accidentally
5659 using path names with one or more superfluous leading slashes; tracing
5660 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5662 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5663 contents are added to the log line that is written for every discarded
5664 recipient. (Previously a log_message setting was ignored.)
5666 36. The ${quote: operator now quotes the string if it is empty.
5668 37. The install script runs exim in order to find its version number. If for
5669 some reason other than non-existence or emptiness, which it checks, it
5670 could not run './exim', it was installing it with an empty version number,
5671 i.e. as "exim-". This error state is now caught, and the installation is
5674 38. An argument was missing from the function that creates an error message
5675 when Exim fails to connect to the socket for saslauthd authentication.
5676 This could cause Exim to crash, or give a corrupted message.
5678 39. Added isip, isip4, and isip6 to ${if conditions.
5680 40. The ACL variables $acl_xx are now saved with the message, and can be
5681 accessed later in routers, transports, and filters.
5683 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5684 strings in the file are not string-expanded.
5686 42. If a MAIL command specified a SIZE value that was too large to fit into an
5687 int variable, the check against message_size_limit failed. Such values are
5688 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5689 day this will have to be increased, but I don't think I want to be around
5690 when emails are that large.
5697 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5698 iconv() is not standard in FreeBSD.
5700 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5701 IPv6 enabled. The observed symptom was a segmentation fault on return from
5702 the function os_common_find_running_interfaces() in src/os.c.
5704 3. In the check_special_case() function in daemon.c I had used "errno" as an
5705 argument name, which causes warnings on some systems. This was basically a
5706 typo, since it was named "eno" in the comments!
5708 4. The code that waits for the clock to tick (at a resolution of some fraction
5709 of a second) so as to ensure message-id uniqueness was always waiting for
5710 at least one whole tick, when it could have waited for less. [This is
5711 almost certainly not relevant at current processor speeds, where it is
5712 unlikely to ever wait at all. But we try to future-proof.]
5714 5. The function that sleeps for a time interval that includes fractions of a
5715 second contained a race. It did not block SIGALRM between setting the
5716 timer, and suspending (a couple of lines later). If the interval was short
5717 and the sigsuspend() was delayed until after it had expired, the suspension
5718 never ended. On busy systems this could lead to processes getting stuck for
5721 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5722 process, before it forks any delivery processes. The open lookup caching
5723 mechanism meant that the open file or database connection was passed into
5724 the delivery process. The problem was that delivery processes always tidy
5725 up cached lookup data. This could cause a problem for the next delivery
5726 process started by the queue runner, because the external queue runner
5727 process does not know about the closure. So the next delivery process
5728 still has data in the lookup cache. In the case of a file lookup, there was
5729 no problem because closing a file descriptor in a subprocess doesn't affect
5730 the parent. However, if the lookup was caching a connection to a database,
5731 the connection was closed, and the second delivery process was likely to
5732 see errors such as "PGSQL: query failed: server closed the connection
5733 unexpectedly". The problem has been fixed by closing all cached lookups
5734 in a queue runner before running a delivery process.
5736 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5737 seem to have the "const" qualifier which it has on other OS. I've
5740 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5741 *of the same type* (client or server) with the same public name that an
5742 error should be diagnosed.
5744 9. When Exim looked up a host name for an IP address, but failed to find the
5745 original IP address when looking up the host name (a safety check), it
5746 output the message "<ip address> does not match any IP for NULL", which was
5747 confusing, to say the least. The bug was that the host name should have
5748 appeared instead of "NULL".
5750 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5751 user that is built into the binary, and the -C or -D options is used, root
5752 privilege is dropped before the configuration file is read. In addition,
5753 logging is switched to stderr instead of the normal log files. If the
5754 configuration then re-defines the Exim user, the unprivileged environment
5755 is probably not what is expected, so Exim logs a panic warning message (but
5758 However, if deliver_drop_privilege is set, the unprivileged state may well
5759 be exactly what is intended, so the warning has been cut out in that case,
5760 and Exim is allowed to try to write to its normal log files.
5766 1. smtp_return_error_details was not giving details for temporary sender
5767 or receiver verification errors.
5769 2. Diagnose a configuration error if two authenticators have the same public
5772 3. Exim used not to create the message log file for a message until the first
5773 delivery attempt. This could be confusing when incoming messages were held
5774 for policy or load reasons. The message log file is now created at the time
5775 the message is received, and an initial "Received" line is written to it.
5777 4. The automatically generated man page for command line options had a minor
5778 bug that caused no ill effects; however, a more serious problem was that
5779 the procedure for building the man page automatically didn't always
5780 operate. Consequently, release 4.20 contains an out-of-date version. This
5781 shouldn't happen again.
5783 5. When building Exim with embedded Perl support, the script that builds the
5784 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5785 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5787 6. The freeze_tell option was not being used for messages that were frozen on
5788 arrival, either by an ACL or by local_scan().
5790 7. Added the smtp_incomplete_transaction log selector.
5792 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5793 was accepting AUTH without a new EHLO.
5795 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5796 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5799 10. Logging of TCP/IP connections (when configured) now happens in the main
5800 daemon process instead of the child process, so that the TCP/IP connection
5801 count is more accurate (but it can never be perfect).
5803 11. The use of "drop" in a nested ACL was not being handled correctly in the
5804 outer ACL. Now, if condition failure induced by the nested "drop" causes
5805 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5806 or "require"), the connection is dropped.
5808 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5809 that yield "discard" can now be used with an "accept" or a "discard" verb,
5810 but an error is generated for any others (because I can't see a useful way
5811 to define what should happen).
5813 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5814 are now processed in the same way as lines in the Exim configuration file.
5815 In particular, continuation lines are supported.
5817 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5819 15. Added -ti meaning -t -i.
5821 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5822 domains, and warn by logging if others are found.
5824 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5825 returned from GIFCONF when Exim is trying to find the list of interfaces on
5826 a host. The code in os.c has been modified to copy each ifreq to an aligned
5827 structure in all cases.
5829 Also, in some cases, the returned ifreq's were being copied to a 'struct
5830 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5831 means the last couple of bytes of an IPv6 address could be chopped if the
5832 ifreq contained only a normal sockaddr (14 bytes storage).
5834 18. Named domain lists were not supported in the hosts_treat_as_local option.
5835 An entry such as +xxxx was not recognized, and was treated as a literal
5838 19. Ensure that header lines added by a DATA ACL are included in the reject log
5839 if the ACL subsequently rejects the message.
5841 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5842 MD5 (which is deprecated).
5844 21. When testing a filter file using -bf, Exim was writing a message when it
5845 took the sender from a "From " line in the message, but it was not doing so
5846 when it took $return_path from a Return-Path: header line. It now does.
5848 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5849 with a valid header line field name (a series of printing characters
5850 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5852 23. Changed "disc" in the source to "disk" to conform to the documentation and
5853 the book and for uniformity.
5855 24. Ignore Sendmail's -Ooption=value command line item.
5857 25. When execve() failed while trying to run a command in a pipe transport,
5858 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5859 could be confused with a return value of 69 from the command itself. This
5860 has been changed to 127, the value the shell returns if it is asked to run
5861 a non-existent command. The wording for the related log line suggests a
5862 non-existent command as the problem.
5864 26. If received_header_text expands to an empty string, do not add a Received:
5865 header line to the message. (Well, it adds a token one on the spool, but
5866 marks it "old" so that it doesn't get used or transmitted.)
5868 27. Installed eximstats 1.28 (addition of -nt option).
5870 28. There was no check for failure on the call to getsockname() in the daemon
5871 code. This can fail if there is a shortage of resources on the system, with
5872 ENOMEM, for example. A temporary error is now given on failure.
5874 29. Contrary to the C standard, it seems that in some environments, the
5875 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5876 program. Exim now does this explicitly; it affects the formatting of
5877 timestamps using strftime().
5879 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5880 complaints. I've now initialized all the variables, to avoid this.
5882 32. Header lines added by a system filter were not being "seen" during
5883 transport-time rewrites.
5885 33. The info_callback() function passed to OpenSSL is set up with type void
5886 (*)(SSL *, int, int), as described somewhere. However, when calling the
5887 function (actually a macro) that sets it up, the type void(*)() is
5888 expected. I've put in a cast to prevent warnings from picky compilers.
5890 34. If a DNS black list lookup found a CNAME record, but there were no A
5891 records associated with the domain it pointed at, Exim crashed.
5893 35. If a DNS black list lookup returned more than one A record, Exim ignored
5894 all but the first. It now scans all returned addresses if a particular IP
5895 value is being sought. In this situation, the contents of the
5896 $dnslist_value variable are a list of all the addresses, separated by a
5899 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5900 to accept a host name and all its aliases if the forward lookup for any of
5901 them yielded the IP address of the incoming connection. Now it accepts only
5902 those names whose forward lookup yields the correct IP address. Any other
5903 names are discarded. This closes a loophole whereby a rogue DNS
5904 administrator could create reverse DNS records to break through a
5905 wildcarded host restriction in an ACL.
5907 37. If a user filter or a system filter that ran in a subprocess used any of
5908 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5909 the wrong values were passed to the pipe command ($thisaddress had the
5910 value of $0, $0 had the value of $1, etc). This bug was introduced by
5911 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5913 38. Improved the line breaking for long SMTP error messages from ACLs.
5914 Previously, if there was no break point between 40 and 75 characters, Exim
5915 left the rest of the message alone. Two changes have been made: (a) I've
5916 reduced the minimum length to 35 characters; (b) if it can't find a break
5917 point between 35 and 75 characters, it looks ahead and uses the first one
5918 that it finds. This may give the occasional overlong line, but at least the
5919 remaining text gets split now.
5921 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5922 file descriptors might be low, and that setting 1000 would always raise it.
5923 It turns out that in some environments, the limit is already over 1000 and
5924 that lowering it causes trouble. So now Exim takes care not to decrease it.
5926 40. When delivering a message, the value of $return_path is set to $sender_
5927 address at the start of routing (routers may change the value). By an
5928 oversight, this default was not being set up when an address was tested by
5929 -bt or -bv, which affected the outcome if any router or filter referred to
5932 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5933 log only when "message" or "log_message" are set. However, if one of the
5934 conditions was an address verification, or a call to a nested ACL, the
5935 messages generated by the underlying test were being passed through. This
5936 no longer happens. The underlying message is available in $acl_verify_
5937 message for both "message" and "log_message" expansions, so it can be
5938 passed through if needed.
5940 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5941 new expansion $bh_ to give the encoded byte string without charset
5942 translation. Translation happens only if iconv() is available; HAVE_ICONV
5943 indicates this at build time. HEADERS_CHARSET gives the charset to
5944 translate to; headers_charset can change it in the configuration, and
5945 "headers charset" can change it in an individual filter file.
5947 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5948 that creates RFC 2047 encoded "words" labels them as that charset instead
5949 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5950 expansion operator; (ii) when Exim creates a From: line for a local
5951 message; (iii) when a header line is rewritten to include a "phrase" part.
5953 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5954 buggy, causing it to skip the first lines of messages whose message ID
5955 ended in 'D'. This would not have bitten before Exim release 4.14, because
5956 message IDs were unlikely to end in 'D' before then. The effect was to have
5957 incorrect size information for certain domains.
5959 45. #include "config.h" was missing at the start of the crypt16.c module. This
5960 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5963 46. If there was a timeout during a "random" callout check, Exim treated it as
5964 a failure of the random address, and carried on sending RSET and the real
5965 address. If the delay was just some slowness somewhere, the response to the
5966 original RCPT would be taken as a response to RSET and so on, causing
5967 mayhem of various kinds.
5969 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5970 when I implemented it. It didn't allow for the fact that some option values
5971 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5972 the right test for positive values.
5974 48. Domain names in DNS records are case-independent. Exim always looks them up
5975 in lower case. Some resolvers return domain names in exactly the case they
5976 appear in the zone file, that is, they may contain uppercase letters. Not
5977 all resolvers do this - some return always lower case. Exim was treating a
5978 change of case by a resolver as a change of domain, similar to a widening
5979 of a domain abbreviation. This triggered its re-routing code and so it was
5980 trying to route what was effectively the same domain again. This normally
5981 caused routing to fail (because the router wouldn't handle the domain
5982 twice). Now Exim checks for this case specially, and just changes the
5983 casing of the domain that it ultimately uses when it transmits the message
5986 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5989 50. If a filter generated a file delivery with a non-absolute name (possible if
5990 no home directory exists for the router), the forbid_file option was not
5993 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5994 the existing equality matching.
5996 52. Exim was using ints instead of ino_t variables in some places where it was
5997 dealing with inode numbers.
5999 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6000 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6001 called TMPDIR, and if it finds it is different, it changes its value.
6003 54. The smtp_printf() function is now made available to local_scan() so
6004 additional output lines can be written before returning. There is also an
6005 smtp_fflush() function to enable the detection of a dropped connection.
6006 The variables smtp_input and smtp_batched_input are exported to
6009 55. Changed the default runtime configuration: the message "Unknown user"
6010 has been removed from the ACL, and instead placed on the localuser router,
6011 using the cannot_route_message feature. This means that any verification
6012 failures that generate their own messages won't get overridden. Similarly,
6013 the "Unrouteable address" message that was in the ACL for unverifiable
6014 relay addresses has also been removed.
6016 56. Added hosts_avoid_esmtp to the smtp transport.
6018 57. The exicyclog script was not checking for the esoteric option
6019 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6020 will work only if exicyclog is run under the appropriate euid.
6022 58. Following a discussion on the list, the rules by which Exim recognises line
6023 endings on incoming messages have been changed. The -dropcr and drop_cr
6024 options are now no-ops, retained only for backwards compatibility. The
6025 following line terminators are recognized: LF CRLF CR. However, special
6026 processing applies to CR:
6028 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6029 nor a local message in the state where . is a terminator.
6031 (ii) If a bare CR is encountered in a header line, an extra space is added
6032 after the line terminator so as not to end the header. The reasoning
6033 behind this is that bare CRs in header lines are most likely either
6034 to be mistakes, or people trying to play silly games.
6036 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6037 was being incorrectly given as 18 bytes larger than it should have been.
6038 This is a VOB (very old bug).
6040 60. This may never have affected anything current, but just in case it has:
6041 When the local host is found other than at the start of a list of hosts,
6042 the local host, those with the same MX, and any that follow, are discarded.
6043 When the list in question was part of a longer list of hosts, the following
6044 hosts (not currently being processed) were also being discarded. This no
6045 longer happens. I'm not sure if this situation could ever has previously
6048 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6051 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6052 header. This is something that is recommended in a new Internet Draft, and
6053 is something that is documented as being done by Sendmail. There are two
6054 possible values. For messages generated by the autoreply transport, Exim
6057 Auto-Submitted: auto-replied
6059 whereas for all other generated messages (e.g. bounces) it adds
6061 Auto-Submitted: auto-generated
6063 63. The "personal" condition in filters now includes a test for the
6064 Auto-Submitted: header. If it contains the string "auto-" the message it
6065 not considered personal.
6067 64. Added rcpt_include_affixes as a generic transport option.
6069 65. Added queue_only_override (default true).
6071 66. Added the syslog_duplication option.
6073 67. If what should have been the first header line of a message consisted of
6074 a space followed by a colon, Exim was mis-interpreting it as a header line.
6075 It isn't of course - it is syntactically invalid and should therefore be
6076 treated as the start of the message body. The misbehaviour could have
6077 caused a number of strange effects, including loss of data in subsequent
6078 header lines, and spool format errors.
6080 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6081 client host had authenticated. This control can now be exercised by an ACL
6082 for more flexibility.
6084 69. By default, callouts do not happen when testing with -bh. There is now a
6085 variant, -bhc, which does actually run the callout code, including
6086 consulting and updating the callout cache.
6088 70. Added support for saslauthd authentication, courtesy of Alexander
6091 71. If statvfs() failed on the spool or log directories while checking their
6092 size for availability, Exim confusingly gave the error "space shortage".
6093 Furthermore, in debugging mode it crashed with a floating point exception.
6094 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6095 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6096 serious problem, Exim now writes to the main and panic logs when this
6097 happens, with details of the failure. It then refuses to accept the
6098 incoming message, giving the message "spool directory problem" or "log
6099 directory problem" with a 421 code for SMTP messages.
6101 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6102 0, 1, and 2 exist, because some OS complain for execs without them (see
6103 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6104 descriptors. However, the code omitted to check that the open succeeded,
6105 causing mysterious errors if for some reason the permissions on /dev/null
6106 got screwed. Now Exim writes a message to the main and panic logs, and
6107 bombs out if it can't open /dev/null.
6109 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6110 interact so that it is all more flexible. It is supposed to remain
6111 backwards compatible. Also added extra_local_interfaces.
6113 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6114 to bomb out with an assertion failure - to the client this appears as a
6115 connection drop. This problem occurs in the part of the code that was taken
6116 from the Samba project. Fortunately, the assertion is in a very simple
6117 function, so I have fixed this by reproducing the function inline in the
6118 one place where it is called, and arranging for authentication to fail
6119 instead of killing the process with assert().
6121 75. The SPA client code was not working when the server requested OEM rather
6122 than Unicode encoding.
6124 76. Added code to make require_files with a specific uid setting more usable in
6125 the case where statting the file as root fails - usually a non-root-mounted
6126 NFS file system. When this happens and the failure is EACCES, Exim now
6127 forks a subprocess and does the per-uid checking as the relevant uid.
6129 77. Added process_log_path.
6131 78. If log_file_path was not explicitly set, a setting of check_log_space or
6132 check_log_inodes was ignored.
6134 79. If a space check for the spool or log partitions fails, the incident is now
6135 logged. Of course, in the latter case the data may get lost...
6137 80. Added the %p formatting code to string_format() so that it can be used to
6138 print addresses in debug_print(). Adjusted all the address printing in the
6139 debugging in store.c to use %p rather than %d.
6141 81. There was a concern that a line of code in smtp_in.c could overflow a
6142 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6143 initially expressed, the concern was not well-founded, because trailing
6144 spaces are removed early. However, if the trailing spaces were followed by
6145 a NULL, they did not get removed, so the overflow was possible. Two fixes
6148 (a) I re-wrote the offending code in a cleaner fashion.
6149 (b) If an incoming SMTP command contains a NULL character, it is rejected
6152 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6153 runs initgroups(), so that if the Exim user is in any additional groups,
6154 they will be used during message reception.
6160 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob