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.
63 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
64 checkers that were complaining about end of non-void function with no
67 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
68 This was a regression intruduced in 4.83 by another bugfix.
70 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
72 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
73 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
79 TF/01 Correctly close the server side of TLS when forking for delivery.
81 When a message was received over SMTP with TLS, Exim failed to clear up
82 the incoming connection properly after forking off the child process to
83 deliver the message. In some situations the subsequent outgoing
84 delivery connection happened to have the same fd number as the incoming
85 connection previously had. Exim would try to use TLS and fail, logging
86 a "Bad file descriptor" error.
88 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
89 utilities have not been installed.
91 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
92 temporary space as the ACL may create new global variables.
94 TL/01 LDAP support uses per connection or global context settings, depending
95 upon the detected version of the libraries at build time.
97 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
98 to extract and use the src ip:port in logging and expansions as if it
99 were a direct connection from the outside internet. PPv2 support was
100 updated based on HAProxy spec change in May 2014.
102 JH/02 Add ${listextract {number}{list}{success}{fail}}.
104 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
105 Properly escape header and check for NULL return.
107 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
110 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
112 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
113 characters in header names, implemented as a verify condition.
114 Contributed by Michael Fischer v. Mollard.
116 TL/05 Rename SPF condition results err_perm and err_temp to standardized
117 results permerror and temperror. Previous values are deprecated but
118 still accepted. In a future release, err_perm and err_temp will be
119 completely removed, which will be a backward incompatibility if the
120 ACL tests for either of these two old results. Patch contributed by
121 user bes-internal on the mailing list.
123 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
125 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
126 selectors, in both main and reject logs.
128 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
131 JH/07 Add malware type "sock" for talking to simple daemon.
133 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
135 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
136 routers/transports under cutthrough routing.
138 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
139 numbers. Touch up "bool" conditional to keep the same definition.
141 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
143 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
145 JH/12 Expand items in router/transport headers_add or headers_remove lists
146 individually rather than the list as a whole. Bug 1452.
148 Required for reasonable handling of multiple headers_ options when
149 they may be empty; requires that headers_remove items with embedded
150 colons must have them doubled (or the list-separator changed).
152 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
153 view the policy declared in the DMARC record. Currently, $dmarc_status
154 is a combined value of both the record presence and the result of the
157 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
159 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
160 dnslookup router and the smtp transport (applying to the forward
163 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
164 of ldap servers used for a specific lookup. Patch provided by Heiko
167 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
168 New variable $lookup_dnssec_authenticated for observability.
170 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
171 Patch submitted by Lars Timman.
173 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
175 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
176 Requires trusted mode and valid format message id, aborts otherwise.
177 Patch contributed by Heiko Schlichting.
179 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
180 certextract with support for various fields. Bug 1358.
182 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
183 is requested by default, modifiable by smtp transport option
186 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
187 operate on certificate variables to give certificate fingerprints
188 Also new ${sha256:cert_variable}.
190 JH/23 The PRDR feature is moved from being Experimental into the mainline.
192 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
193 Christian Aistleitner.
195 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
197 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
198 file. Patch from Wolfgang Breyha.
200 JH/25 Expand the coverage of the delivery $host and $host_address to
201 client authenticators run in verify callout. Bug 1476.
203 JH/26 Port service names are now accepted for tls_on_connect_ports, to
204 align with daemon_smtp_ports. Bug 72.
206 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
207 support and error reporting did not work properly.
209 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
210 and is readable. Patch from Andrew Colin Kissa.
212 TL/14 Enhance documentation of ${run expansion and how it parses the
213 commandline after expansion, particularly in the case when an
214 unquoted variable expansion results in an empty value.
216 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
218 PP/02 Fix internal collision of T_APL on systems which support RFC3123
219 by renaming away from it. Addresses GH issue 15, reported by
222 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
224 TL/15 SECURITY: prevent double expansion in math comparison functions
225 (can expand unsanitized data). Not remotely exploitable.
232 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
234 PP/02 Make -n do something, by making it not do something.
235 When combined with -bP, the name of an option is not output.
237 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
240 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
241 $sender_host_name and config options to manage this, and basic check
244 PP/05 DSCP support for outbound connections and control modifier for inbound.
246 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
247 (Only plugin which currently uses this is kerberos4, which nobody should
248 be using, but we should make it available and other future plugins might
249 conceivably use it, even though it would break NAT; stuff *should* be
250 using channel bindings instead).
252 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
253 name; added for Sendmail compatibility; requires admin caller.
254 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
255 just ignore it); requires trusted caller.
256 Also parse but ignore: -Ac -Am -X<logfile>
259 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
261 TL/02 Add +smtp_confirmation as a default logging option.
263 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
264 Patch by Magnus Holmgren from 2007-02-20.
266 TL/04 Bugzilla 1281 - Spec typo.
267 Bugzilla 1283 - Spec typo.
268 Bugzilla 1290 - Spec grammar fixes.
270 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
272 TL/06 Add Experimental DMARC support using libopendmarc libraries.
274 TL/07 Fix an out of order global option causing a segfault. Reported to dev
275 mailing list by by Dmitry Isaikin.
277 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
279 JH/02 Support "G" suffix to numbers in ${if comparisons.
281 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
283 NM/01 Bugzilla 1197 - Spec typo
284 Bugzilla 1196 - Spec examples corrections
286 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
288 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
289 gnutls_enable_pkcs11, but renamed to more accurately indicate its
292 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
293 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
295 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
296 "acl {{name}{arg}...}", and optional args on acl condition
299 JH/05 Permit multiple router/transport headers_add/remove lines.
301 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
303 JH/07 Avoid using a waiting database for a single-message-only transport.
304 Performance patch from Paul Fisher. Bugzilla 1262.
306 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
309 JH/09 Add $headers_added variable, with content from use of ACL modifier
310 add_header (but not yet added to the message). Bugzilla 199.
312 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
313 Pulled from Bugzilla 817 by Wolfgang Breyha.
315 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
317 (nb: this is the same fix as in Exim 4.80.1)
319 JH/11 Add A= logging on delivery lines, and a client_set_id option on
322 JH/12 Add optional authenticated_sender logging to A= and a log_selector
325 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
327 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
328 advertise SMTP AUTH mechanism to us, instead of a generic
329 protocol violation error. Also, make Exim more robust to bad
330 data from the Dovecot auth socket.
332 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
334 When a queue runner is handling a message, Exim first routes the
335 recipient addresses, during which it prunes them based on the retry
336 hints database. After that it attempts to deliver the message to
337 any remaining recipients. It then updates the hints database using
340 So if a recipient address works intermittently, it can get repeatedly
341 deferred at routing time. The retry hints record remains fresh so the
342 address never reaches the final cutoff time.
344 This is a fairly common occurrence when a user is bumping up against
345 their storage quota. Exim had some logic in its local delivery code
346 to deal with this. However it did not apply to per-recipient defers
347 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
349 This change adds a proper retry rule check during routing so that the
350 final cutoff time is checked against the message's age. We only do
351 this check if there is an address retry record and there is not a
352 domain retry record; this implies that previous attempts to handle
353 the address had the retry_use_local_parts option turned on. We use
354 this as an approximation for the destination being like a local
355 delivery, as in LMTP.
357 I suspect this new check makes the old local delivery cutoff check
358 redundant, but I have not verified this so I left the code in place.
360 TF/02 Correct gecos expansion when From: is a prefix of the username.
362 Test 0254 submits a message to Exim with the header
366 When I ran the test suite under the user fanf2, Exim expanded
367 the header to contain my full name, whereas it should have added
368 a Resent-Sender: header. It erroneously treats any prefix of the
369 username as equal to the username.
371 This change corrects that bug.
373 GF/01 DCC debug and logging tidyup
374 Error conditions log to paniclog rather than rejectlog.
375 Debug lines prefixed by "DCC: " to remove any ambiguity.
377 TF/03 Avoid unnecessary rebuilds of lookup-related code.
379 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
380 Bug spotted by Jeremy Harris; was flawed since initial commit.
381 Would have resulted in OCSP responses post-SNI triggering an Exim
382 NULL dereference and crash.
384 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
386 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
387 Bug detection, analysis and fix by Samuel Thibault.
388 Bugzilla 1331, Debian bug #698092.
390 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
392 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
393 Server implementation by Todd Lyons, client by JH.
394 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
395 config variable "prdr_enable" controls whether the server
396 advertises the facility. If the client requests PRDR a new
397 acl_data_smtp_prdr ACL is called once for each recipient, after
398 the body content is received and before the acl_smtp_data ACL.
399 The client is controlled by bolth of: a hosts_try_prdr option
400 on the smtp transport, and the server advertisement.
401 Default client logging of deliveries and rejections involving
402 PRDR are flagged with the string "PRDR".
404 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
405 fclose(). Diagnosis by Todd Lyons.
407 PP/17 Update configure.default to handle IPv6 localhost better.
408 Patch by Alain Williams (plus minor tweaks).
411 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
412 This is now consistent with GnuTLS, and is now documented: the
413 previous undocumented portable approach to treating the option as
414 unset was to force an expansion failure. That still works, and
415 an empty string is now equivalent.
417 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
418 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
419 not performing validation itself.
421 PP/20 Added force_command boolean option to pipe transport.
422 Patch from Nick Koston, of cPanel Inc.
424 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
427 TF/04 Added udpsend ACL modifer and hexquote expansion operator
429 PP/21 Fix eximon continuous updating with timestamped log-files.
430 Broken in a format-string cleanup in 4.80, missed when I repaired the
431 other false fix of the same issue.
432 Report and fix from Heiko Schlichting.
435 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
436 Report from Prashanth Katuri.
438 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
439 It's SecureTransport, so affects any MacOS clients which use the
440 system-integrated TLS libraries, including email clients.
442 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
443 using a MIME ACL for non-SMTP local injection.
444 Report and assistance in diagnosis by Warren Baker.
446 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
448 JH/16 Fix comparisons for 64b. Bugzilla 1385.
450 TL/09 Add expansion variable $authenticated_fail_id to keep track of
451 last id that failed so it may be referenced in subsequent ACL's.
453 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
456 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
457 ldap library initialization, allowing self-signed CA's to be
458 used. Also properly sets require_cert option later in code by
459 using NULL (global ldap config) instead of ldap handle (per
460 session). Bug diagnosis and testing by alxgomz.
462 TL/12 Enhanced documentation in the ratelimit.pl script provided in
463 the src/util/ subdirectory.
465 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
466 renamed to Transport Post Delivery Action by Jeremy Harris, as
469 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
470 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
471 redis_servers = needs to be configured which will be used by the redis
472 lookup. Patch from Warren Baker, of The Packet Hub.
474 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
476 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
477 hostname or reverse DNS when processing a host list. Used suggestions
478 from multiple comments on this bug.
480 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
482 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
483 Missed a few lines, added it to make the runtest require no keyboard
486 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
487 contains upper case chars. Make router use caseful_local_part.
489 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
490 support when GnuTLS has been built with p11-kit.
496 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
498 This, or similar/improved, will also be change PP/11 of 4.82.
504 PP/01 Handle short writes when writing local log-files.
505 In practice, only affects FreeBSD (8 onwards).
506 Bugzilla 1053, with thanks to Dmitry Isaikin.
508 NM/01 Bugzilla 949 - Documentation tweak
510 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
513 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
515 PP/02 Implemented gsasl authenticator.
517 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
519 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
520 `pkg-config foo` for cflags/libs.
522 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
523 with rest of GSASL and with heimdal_gssapi.
525 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
526 `pkg-config foo` for cflags/libs for the TLS implementation.
528 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
529 properties get this fed in as external SSF. A number of robustness
530 and debugging improvements to the cyrus_sasl authenticator.
532 PP/08 cyrus_sasl server now expands the server_realm option.
534 PP/09 Bugzilla 1214 - Log authentication information in reject log.
535 Patch by Jeremy Harris.
537 PP/10 Added dbmjz lookup type.
539 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
541 PP/12 MAIL args handles TAB as well as SP, for better interop with
542 non-compliant senders.
543 Analysis and variant patch by Todd Lyons.
545 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
546 Bug report from Lars Müller <lars@samba.org> (via SUSE),
547 Patch from Dirk Mueller <dmueller@suse.com>
549 PP/13 tls_peerdn now print-escaped for spool files.
550 Observed some $tls_peerdn in wild which contained \n, which resulted
551 in spool file corruption.
553 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
554 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
555 or write after TLS renegotiation, which otherwise led to messages
558 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
559 as a tracking header (ie: a signed header comes before the signature).
560 Patch from Wolfgang Breyha.
562 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
563 comma-sep list; embedded commas doubled.
565 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
567 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
569 Report and patch from Dmitry Banschikov.
571 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
572 Removed SSL_clear() after SSL_new() which led to protocol negotiation
573 failures. We appear to now support TLS1.1+ with Exim.
575 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
576 lets Exim select keys and certificates based upon TLS SNI from client.
577 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
578 before an outbound SMTP session. New log_selector, +tls_sni.
580 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
581 NULL dereference. Report and patch from Alun Jones.
583 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
584 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
585 Not seeing resolver debug output on NetBSD, but suspect this is a
586 resolver implementation change.
588 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
589 Left warnings. Added "eximon gdb" invocation mode.
591 PP/21 Defaulting "accept_8bitmime" to true, not false.
593 PP/22 Added -bw for inetd wait mode support.
595 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
596 locate the relevant includes and libraries. Made this the default.
598 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
599 Bugzilla 1246, report and most of solution from Tomasz Kusy.
601 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
602 This may cause build issues on older platforms.
604 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
605 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
606 gnutls_require_mac & gnutls_require_protocols (no longer supported).
607 Added SNI support via GnuTLS too.
608 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
610 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
612 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
613 Applied second patch from Janne, implementing suggestion to default
614 multiple-strings-in-record handling to match SPF spec.
616 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
618 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
619 read-only, out of scope).
620 Patch from Wolfgang Breyha, report from Stuart Northfield.
622 PP/29 Fix three issues highlighted by clang analyser static analysis.
623 Only crash-plausible issue would require the Cambridge-specific
624 iplookup router and a misconfiguration.
625 Report from Marcin Mirosław.
627 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
629 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
630 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
631 As part of this, removing so much warning spew let me fix some minor
632 real issues in debug logging.
634 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
635 assignment on my part. Fixed.
637 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
638 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
639 Janne Snabb (who went above and beyond: thank you).
641 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
642 string otherwise requires a connection and a bunch more work and it's
643 relatively easy to get wrong. Should also expose TLS library linkage
646 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
647 64-bit ${eval} (JH/03).
649 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
650 GNU libc to support some of the 64-bit stuff, should not lead to
651 conflicts. Defined before os.h is pulled in, so if a given platform
652 needs to override this, it can.
654 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
655 protection layer was required, which is not implemented.
656 Bugzilla 1254, patch from Wolfgang Breyha.
658 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
659 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
660 tls_dhparam take prime identifiers. Also unbreak combination of
661 OpenSSL+DH_params+TLSSNI.
663 PP/39 Disable SSLv2 by default in OpenSSL support.
669 PP/01 Solaris build fix for Oracle's LDAP libraries.
670 Bugzilla 1109, patch from Stephen Usher.
672 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
674 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
677 TF/02 Fix a couple more cases where we did not log the error message
678 when unlink() failed. See also change 4.74-TF/03.
680 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
681 lock up or crash if it happened to be inside a call to libc when it
682 got a SIGUSR1 from exiwhat.
684 The SIGUSR1 handler appends the current process status to the process
685 log which is later printed by exiwhat. It used to use the general
686 purpose logging code to do this, but several functions it calls are
687 not safe for signals.
689 The new output code in the SIGUSR1 handler is specific to the process
690 log, and simple enough that it's easy to inspect for signal safety.
691 Removing some special cases also simplifies the general logging code.
692 Removing the spurious timestamps from the process log simplifies
695 TF/04 Improved ratelimit ACL condition.
697 The /noupdate option has been deprecated in favour of /readonly which
698 has clearer semantics. The /leaky, /strict, and /readonly update modes
699 are mutually exclusive. The update mode is no longer included in the
700 database key; it just determines when the database is updated. (This
701 means that when you upgrde Exim will forget old rate measurements.)
703 Exim now checks that the per_* options are used with an update mode that
704 makes sense for the current ACL. For example, when Exim is processing a
705 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
706 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
707 must specify per_mail/readonly. If you omit the update mode it defaults to
708 /leaky where that makes sense (as before) or /readonly where required.
710 The /noupdate option is now undocumented but still supported for
711 backwards compatibility. It is equivalent to /readonly except that in
712 ACLs where /readonly is required you may specify /leaky/noupdate or
713 /strict/noupdate which are treated the same as /readonly.
715 A useful new feature is the /count= option. This is a generalization
716 of the per_byte option, so that you can measure the throughput of other
717 aggregate values. For example, the per_byte option is now equivalent
718 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
720 The per_rcpt option has been generalized using the /count= mechanism
721 (though it's more complicated than the per_byte equivalence). When it is
722 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
723 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
724 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
725 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
726 Note that using per_rcpt with a non-readonly update mode in more than
727 one ACL will cause the recipients to be double-counted. (The per_mail
728 and per_byte options don't have this problem.)
730 The handling of very low rates has changed slightly. If the computed rate
731 is less than the event's count (usually one) then this event is the first
732 after a long gap. In this case the rate is set to the same as this event's
733 count, so that the first message of a spam run is counted properly.
735 The major new feature is a mechanism for counting the rate of unique
736 events. The new per_addr option counts the number of different
737 recipients that someone has sent messages to in the last time period. It
738 behaves like per_rcpt if all the recipient addresses are different, but
739 duplicate recipient addresses do not increase the measured rate. Like
740 the /count= option this is a general mechanism, so the per_addr option
741 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
742 example, measure the rate that a client uses different sender addresses
743 with the options per_mail/unique=$sender_address. There are further
744 details in the main documentation.
746 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
748 TF/06 Removed a few PCRE remnants.
750 TF/07 Automatically extract Exim's version number from tags in the git
751 repository when doing development or release builds.
753 PP/02 Raise smtp_cmd_buffer_size to 16kB.
754 Bugzilla 879. Patch from Paul Fisher.
756 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
757 Heavily based on revision 40f9a89a from Simon Arlott's tree.
760 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
762 PP/05 Variable $av_failed, true if the AV scanner deferred.
763 Bugzilla 1078. Patch from John Horne.
765 PP/06 Stop make process more reliably on build failure.
766 Bugzilla 1087. Patch from Heiko Schlittermann.
768 PP/07 Make maildir_use_size_file an _expandable_ boolean.
769 Bugzilla 1089. Patch from Heiko Schlittermann.
771 PP/08 Handle ${run} returning more data than OS pipe buffer size.
772 Bugzilla 1131. Patch from Holger Weiß.
774 PP/09 Handle IPv6 addresses with SPF.
775 Bugzilla 860. Patch from Wolfgang Breyha.
777 PP/10 GnuTLS: support TLS 1.2 & 1.1.
779 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
782 PP/11 match_* no longer expand right-hand-side by default.
783 New compile-time build option, EXPAND_LISTMATCH_RHS.
784 New expansion conditions, "inlist", "inlisti".
786 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
788 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
790 PP/14 fix log_write() format string regression from TF/03.
791 Bugzilla 1152. Patch from Dmitry Isaikin.
797 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
799 PP/02 Harmonised TLS library version reporting; only show if debugging.
800 Layout now matches that introduced for other libraries in 4.74 PP/03.
802 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
804 PP/04 New "dns_use_edns0" global option.
806 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
809 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
810 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
812 TK/01 Updated PolarSSL code to 0.14.2.
813 Bugzilla 1097. Patch from Andreas Metzler.
815 PP/07 Catch divide-by-zero in ${eval:...}.
818 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
821 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
822 format-string attack -- SECURITY: remote arbitrary code execution.
824 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
825 time unintentionally subject to list matching rules, letting the header
826 cause arbitrary Exim lookups (of items which can occur in lists, *not*
827 arbitrary string expansion). This allowed for information disclosure.
829 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
830 INT_MIN/-1 -- value coerced to INT_MAX.
836 NM/01 Workround for PCRE version dependency in version reporting
839 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
840 This fixes portability to compilers other than gcc, notably
841 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
843 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
844 makefiles for portability to HP-UX and POSIX correctness.
846 PP/01 Permit LOOKUP_foo enabling on the make command-line.
847 Also via indented variable definition in the Makefile.
848 (Debugging by Oliver Heesakkers).
850 PP/02 Restore caching of spamd results with expanded spamd_address.
851 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
853 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
854 Improves build reliability. Fix from: Frank Elsner
856 NM/02 Fix wide character breakage in the rfc2047 coding
857 Fixes bug 1064. Patch from Andrey N. Oktyabrski
859 NM/03 Allow underscore in dnslist lookups
860 Fixes bug 1026. Patch from Graeme Fowler
862 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
863 Code patches from Adam Ciarcinski of NetBSD.
865 NM/04 Fixed exiqgrep to cope with mailq missing size issue
868 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
869 is logged, to avoid truncation. Patch from John Horne.
871 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
872 Patch from Jakob Hirsch.
874 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
875 SQL string expansion failure details.
876 Patch from Andrey Oktyabrski.
878 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
879 Patch from Simon Arlott.
881 PP/09 New lookups functionality failed to compile on old gcc which rejects
882 extern declarations in function scope.
883 Patch from Oliver Fleischmann
885 PP/10 Use sig_atomic_t for flags set from signal handlers.
886 Check getgroups() return and improve debugging.
887 Fixed developed for diagnosis in bug 927 (which turned out to be
890 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
891 Patch from Mark Zealey.
893 PP/12 Bugzilla 1056: Improved spamd server selection.
894 Patch from Mark Zealey.
896 PP/13 Bugzilla 1086: Deal with maildir quota file races.
897 Based on patch from Heiko Schlittermann.
899 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
900 Patch from Uwe Doering, sign-off by Michael Haardt.
902 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
903 variable declaration deep within a block. Bug and patch from
906 PP/15 lookups-Makefile IRIX compatibilty coercion.
908 PP/16 Make DISABLE_DKIM build knob functional.
910 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
911 Patch by Simon Arlott
913 TF/03 Fix valgrind.h portability to C89 compilers that do not support
914 variable argument macros. Our copy now differs from upstream.
920 TF/01 Failure to get a lock on a hints database can have serious
921 consequences so log it to the panic log.
923 TF/02 Log LMTP confirmation messages in the same way as SMTP,
924 controlled using the smtp_confirmation log selector.
926 TF/03 Include the error message when we fail to unlink a spool file.
928 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
929 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
930 for maintaining out-of-tree patches for some time.
932 PP/01 Bugzilla 139: Documentation and portability issues.
933 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
934 Handle per-OS dynamic-module compilation flags.
936 PP/02 Let /dev/null have normal permissions.
937 The 4.73 fixes were a little too stringent and complained about the
938 permissions on /dev/null. Exempt it from some checks.
939 Reported by Andreas M. Kirchwitz.
941 PP/03 Report version information for many libraries, including
942 Exim version information for dynamically loaded libraries. Created
943 version.h, now support a version extension string for distributors
944 who patch heavily. Dynamic module ABI change.
946 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
947 privilege escalation vulnerability whereby the Exim run-time user
948 can cause root to append content of the attacker's choosing to
951 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
954 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
955 If dropping privileges for untrusted macros, we disabled normal logging
956 on the basis that it would fail; for the Exim run-time user, this is not
957 the case, and it resulted in successful deliveries going unlogged.
958 Fixed. Reported by Andreas Metzler.
964 PP/01 Date: & Message-Id: revert to normally being appended to a message,
965 only prepend for the Resent-* case. Fixes regression introduced in
966 Exim 4.70 by NM/22 for Bugzilla 607.
968 PP/02 Include check_rfc2047_length in configure.default because we're seeing
969 increasing numbers of administrators be bitten by this.
971 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
973 PP/03 Bugzilla 994: added openssl_options main configuration option.
975 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
977 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
979 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
981 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
982 without a peer certificate, leading to a segfault because of an
983 assumption that peers always have certificates. Be a little more
984 paranoid. Problem reported by Martin Tscholak.
986 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
987 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
988 NB: ClamAV planning to remove STREAM in "middle of 2010".
989 CL also introduces -bmalware, various -d+acl logging additions and
990 more caution in buffer sizes.
992 PP/09 Implemented reverse_ip expansion operator.
994 PP/10 Bugzilla 937: provide a "debug" ACL control.
996 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
998 PP/12 Bugzilla 973: Implement --version.
1000 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1002 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1004 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1006 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1007 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1008 ignore trailing whitespace.
1010 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1012 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1015 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1016 Notification from Dr Andrew Aitchison.
1018 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1019 ExtendedDetectionInfo response format.
1020 Notification from John Horne.
1022 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1025 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1026 XSL and documented dependency on system catalogs, with examples of how
1029 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1032 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1033 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1034 configuration file which is writeable by the Exim user or group.
1036 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1037 of configuration files to cover files specified with the -C option if
1038 they are going to be used with root privileges, not just the default
1041 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1042 option (effectively making it always true).
1044 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1045 files to be used while preserving root privileges.
1047 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1048 that rogue child processes cannot use them.
1050 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1051 run-time user, instead of root.
1053 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1054 Exim run-time user without dropping privileges.
1056 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1057 result string, instead of calling string_vformat() twice with the same
1060 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1061 for other users. Others should always drop root privileges if they use
1062 -C on the command line, even for a whitelisted configure file.
1064 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1066 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1072 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1073 $data_path, and $header_path variables; fixed documentation bugs and
1076 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1077 exipick to access non-standard spools, including the "frozen" queue
1080 NM/01 Bugzilla 965: Support mysql stored procedures.
1081 Patch from Alain Williams
1083 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1085 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1086 Patch from Andreas Metzler
1088 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1089 Patch from Kirill Miazine
1091 NM/05 Bugzilla 671: Added umask to procmail example.
1093 JJ/03 installed exipick 20100323.0, fixing doc bug
1095 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1096 directory. Notification and patch from Dan Rosenberg.
1098 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1100 TK/02 Improve log output when DKIM signing operation fails.
1102 MH/01 Treat the transport option dkim_domain as a colon separated
1103 list, not as a single string, and sign the message with each element,
1104 omitting multiple occurences of the same signer.
1106 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1107 Bugzilla 985, 986. Patch by Simon Arlott
1109 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1110 Patch by Simon Arlott
1112 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1113 MBX locking. Notification from Dan Rosenberg.
1119 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1121 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1123 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1125 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1127 NM/04 Bugzilla 915: Flag broken perl installation during build.
1133 TK/01 Added patch by Johannes Berg that expands the main option
1134 "spamd_address" if it starts with a dollar sign.
1136 TK/02 Write list of recipients to X-Envelope-Sender header when building
1137 the mbox-format spool file for content scanning (suggested by Jakob
1140 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1141 (http://www.dcc-servers.net/) support via dccifd. Activated by
1142 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1144 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1145 by Mark Daniel Reidel <mr@df.eu>.
1147 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1148 When building exim an external PCRE library is now needed -
1149 PCRE is a system library on the majority of modern systems.
1150 See entry on PCRE_LIBS in EDITME file.
1152 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1153 conversation. Added nologin parameter to request.
1154 Patch contributed by Kirill Miazine.
1156 TF/01 Do not log submission mode rewrites if they do not change the address.
1158 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1160 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1161 log files in place. Contributed by Roberto Lima.
1163 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1165 TF/03 Bugzilla 615: When checking the local_parts router precondition
1166 after a local_part_suffix or local_part_prefix option, Exim now
1167 does not use the address's named list lookup cache, since this
1168 contains cached lookups for the whole local part.
1170 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1171 Robert Millan. Documentation is in experimental-spec.txt.
1173 TF/04 Bugzilla 668: Fix parallel build (make -j).
1175 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1177 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1178 Patch provided by Jan Srzednicki.
1180 TF/05 Leading white space used to be stripped from $spam_report which
1181 wrecked the formatting. Now it is preserved.
1183 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1184 that they are available at delivery time.
1186 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1188 TF/08 TLS error reporting now respects the incoming_interface and
1189 incoming_port log selectors.
1191 TF/09 Produce a more useful error message if an SMTP transport's hosts
1192 setting expands to an empty string.
1194 NM/06 Bugzilla 744: EXPN did not work under TLS.
1195 Patch provided by Phil Pennock.
1197 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1198 Patch provided by Richard Godbee.
1200 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1201 acl_smtp_notquit, added index entry.
1203 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1204 Patch provided by Eugene Bujak.
1206 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1207 accept(). Patch provided by Maxim Dounin.
1209 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1210 Patch provided by Phil Pennock.
1212 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1214 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1215 Patch provided by Brad "anomie" Jorsch.
1217 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1218 Patch provided by Dean Brooks.
1220 TK/05 Add native DKIM support (does not depend on external libraries).
1222 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1223 Patch provided by Graeme Fowler.
1225 NM/16 Bugzilla 851: Documentation example syntax fix.
1227 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1229 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1232 NM/19 Bugzilla 745: TLS version reporting.
1233 Patch provided by Phil Pennock.
1235 NM/20 Bugzilla 167: bool: condition support.
1236 Patch provided by Phil Pennock.
1238 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1239 clients. Patch provided by Phil Pennock.
1241 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1242 Patch provided by Brad "anomie" Jorsch.
1244 NM/23 Bugzilla 687: Fix misparses in eximstats.
1245 Patch provided by Heiko Schlittermann.
1247 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1248 Patch provided by Heiko Schlittermann.
1250 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1251 plus update to original patch.
1253 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1255 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1256 Patch provided by David Brownlee.
1258 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1260 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1262 NM/30 Bugzilla 888: TLS documentation bugfixes.
1264 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1266 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1267 Unlike the original bugzilla I have changed all shell scripts in src tree.
1269 NM/33 Bugzilla 898: Transport filter timeout fix.
1270 Patch by Todd Rinaldo.
1272 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1273 Patch by Serge Demonchaux.
1275 NM/35 Bugzilla 39: Base64 decode bug fixes.
1276 Patch by Jakob Hirsch.
1278 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1280 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1282 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1284 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1290 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1291 ALT-N's libdkim that I have put here:
1292 http://duncanthrax.net/exim-experimental/
1294 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1295 were called 'true' and it seems that C99 defines that as a reserved
1296 keyword to be used with 'bool' variable types. That means you could
1297 not include C99-style headers which use bools without triggering
1298 build errors in sieve.c.
1300 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1301 as mailq or other aliases. Changed the --help handling significantly
1302 to do whats expected. exim_usage() emits usage/help information.
1304 SC/01 Added the -bylocaldomain option to eximstats.
1306 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1308 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1310 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1316 PH/01 Another patch from the Sieve maintainer.
1318 PH/02 When an IPv6 address is converted to a string for single-key lookup
1319 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1320 dots are used instead of colons so that keys in lsearch files need not
1321 contain colons. This was done some time before quoting was made available
1322 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1323 (notated using the quote facility) so as to distinguish them from IPv4
1324 keys. This meant that lookups for IP addresses in host lists did not work
1325 for iplsearch lookups.
1327 This has been fixed by arranging for IPv6 addresses to be expressed with
1328 colons if the lookup type is iplsearch. This is not incompatible, because
1329 previously such lookups could never work.
1331 The situation is now rather anomolous, since one *can* have colons in
1332 ordinary lsearch keys. However, making the change in all cases is
1333 incompatible and would probably break a number of configurations.
1335 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1338 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1339 conversion specification without a maximum field width, thereby enabling
1340 a rogue spamd server to cause a buffer overflow. While nobody in their
1341 right mind would setup Exim to query an untrusted spamd server, an
1342 attacker that gains access to a server running spamd could potentially
1343 exploit this vulnerability to run arbitrary code as the Exim user.
1345 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1346 $primary_hostname instead of what libspf2 thinks the hosts name is.
1348 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1349 a directory entry by the name of the lookup key. Previously, if a
1350 symlink pointed to a non-existing file or a file in a directory that
1351 Exim lacked permissions to read, a lookup for a key matching that
1352 symlink would fail. Now it is enough that a matching directory entry
1353 exists, symlink or not. (Bugzilla 503.)
1355 PH/03 The body_linecount and body_zerocount variables are now exported in the
1358 PH/04 Added the $dnslist_matched variable.
1360 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1361 This means they are set thereafter only if the connection becomes
1364 PH/06 Added the client_condition to authenticators so that some can be skipped
1365 by clients under certain conditions.
1367 PH/07 The error message for a badly-placed control=no_multiline_responses left
1368 "_responses" off the end of the name.
1370 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1372 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1373 (without spaces) instead of just copying the configuration text.
1375 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1377 PH/11 Added $max_received_linelength.
1379 PH/12 Added +ignore_defer and +include_defer to host lists.
1381 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1382 way in which PCRE > 7.0 is built.
1384 PH/14 Implemented queue_only_load_latch.
1386 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1387 MAIL command. The effect was to mangle the value on 64-bit systems.
1389 PH/16 Another patch from the Sieve maintainer.
1391 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1393 PH/18 If a system quota error occurred while trying to create the file for
1394 a maildir delivery, the message "Mailbox is full" was not appended to the
1395 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1396 only to a quota excession during the actual writing of the file.
1398 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1399 characters?) which causes problems in log lines. The DN values are now
1400 passed through string_printing() before being added to log lines.
1402 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1403 and InterBase are left for another time.)
1405 PH/21 Added message_body_newlines option.
1407 PH/22 Guard against possible overflow in moan_check_errorcopy().
1409 PH/23 POSIX allows open() to be a macro; guard against that.
1411 PH/24 If the recipient of an error message contained an @ in the local part
1412 (suitably quoted, of course), incorrect values were put in $domain and
1413 $local_part during the evaluation of errors_copy.
1419 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1420 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1423 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1424 issue a MAIL command.
1426 PH/02 In an ACL statement such as
1428 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1430 if a client was not listed at all, or was listed with a value other than
1431 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1432 the condition was not true (as it should be), so access was not denied.
1433 The bug was that the ! inversion was incorrectly passed on to the second
1434 item. This has been fixed.
1436 PH/03 Added additional dnslists conditions == and =& which are different from
1437 = and & when the dns lookup returns more than one IP address.
1439 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1440 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1442 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1443 FSYNC, which compiles an option called disable_fsync that allows for
1444 bypassing fsync(). The documentation is heavily laced with warnings.
1446 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1448 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1449 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1450 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1451 including adding "make clean"; (3) Added -fPIC when compiling the test
1452 dynamically loaded module, to get rid of a warning.
1454 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1455 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1456 = 0s. The bug is otherwise harmless.
1458 PH/07 There was a bug in the dovecot authenticator such that the value of
1459 $auth1 could be overwritten, and so not correctly preserved, after a
1460 successful authentication. This usually meant that the value preserved by
1461 the server_setid option was incorrect.
1463 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1465 PH/09 Installed PCRE release 7.0.
1467 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1468 run for batched SMTP input. It is now run at the start of every message
1469 in the batch. While fixing this I discovered that the process information
1470 (output by running exiwhat) was not always getting set for -bs and -bS
1471 input. This is fixed, and it now also says "batched" for BSMTP.
1473 PH/11 Added control=no_pipelining.
1475 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1476 patch, slightly modified), and move the expansion of helo_data till after
1477 the connection is made in the smtp transport (so it can use these
1480 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1482 PH/14 Added log_selector = +pid.
1484 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1486 PH/16 Add ${if forany and ${if forall.
1488 PH/17 Added dsn_from option to vary the From: line in DSNs.
1490 PH/18 Flush SMTP output before performing a callout, unless control =
1491 no_callout_flush is set.
1493 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1494 was true (the default) a successful delivery failed to delete the retry
1495 item, thus causing premature timeout of the address. The bug is now
1498 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1500 PH/21 Long custom messages for fakedefer and fakereject are now split up
1501 into multiline reponses in the same way that messages for "deny" and
1502 other ACL rejections are.
1504 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1505 with slight modification.
1507 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1508 draft, changing the syntax and factoring some duplicate code".
1510 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1511 for deliveries of the second and subsequent messages over the same SMTP
1514 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1515 ${reduce, with only minor "tidies".
1517 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1519 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1520 expansion side effects.
1522 PH/27 When a message times out after an over-quota error from an Exim-imposed
1523 quota, the bounce message says "mailbox is full". This message was not
1524 being given when it was a system quota that was exceeded. It now should
1527 MH/03 Made $recipients available in local_scan(). local_scan() already has
1528 better access to the recipient list through recipients_list[], but
1529 $recipients can be useful in postmaster-provided expansion strings.
1531 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1532 in the case of a MAIL command with additional options following the
1533 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1534 were accidentally chopped off.
1536 PH/29 SMTP synchronization checks are implemented when a command is read -
1537 there is a check that no more input is waiting when there shouldn't be
1538 any. However, for some commands, a delay in an ACL can mean that it is
1539 some time before the response is written. In this time, more input might
1540 arrive, invalidly. So now there are extra checks after an ACL has run for
1541 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1542 pipelining has not been advertised.
1544 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1546 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1547 returned in the SMTP response when smtp_return_error_details was false.
1548 This has been fixed.
1550 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1551 instead of the C I/O that was originally supplied, because problems were
1552 reported on Solaris.
1554 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1555 Exim which did not show up earlier: it was assuming that a call to
1556 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1557 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1558 was a macro that became an assignment, so it seemed to work. This has
1559 changed to a proper function call with a void return, hence the compile
1560 error. Exim's code has been fixed.
1562 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1565 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1567 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1569 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1570 because it turns out to be too big or there is a timeout) there is no
1571 "Completed" line in the log. When some messages of this type were
1572 selected by exigrep, they were listed as "not completed". Others were
1573 picked up by some special patterns. I have improved the selection
1574 criteria to be more general.
1576 PH/38 The host_find_failed option in the manualroute router can now be set
1577 to "ignore", to completely ignore a host whose IP address cannot be
1578 found. If all hosts are ignored, the behaviour is controlled by the new
1579 host_all_ignored option.
1581 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1582 homing or because of multiple MX records with /mx) generated more than
1583 one IP address, and the following item turned out to be the local host,
1584 all the secondary addresses of the first item were incorrectly removed
1585 from the list, along with the local host and any following hosts (which
1586 is what is supposed to happen).
1588 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1589 whoever called Exim into the -H file. In the case of the daemon it was
1590 behaving confusingly. When first started, it used values for whoever
1591 started the daemon, but after a SIGHUP it used the Exim user (because it
1592 calls itself on a restart). I have changed the code so that it now always
1595 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1596 message are rejected with the same error (e.g. no authentication or bad
1597 sender address), and a DATA command is nevertheless sent (as can happen
1598 with PIPELINING or a stupid MUA), the error message that was given to the
1599 RCPT commands is included in the rejection of the DATA command. This is
1600 intended to be helpful for MUAs that show only the final error to their
1603 PH/42 Another patch from the Sieve maintainer.
1605 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1606 Eximstats - Fixed some broken HTML links and added missing column headers
1608 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1609 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1611 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1613 PH/43 Yet another patch from the Sieve maintainer.
1615 PH/44 I found a way to check for a TCP/IP connection going away before sending
1616 the response to the final '.' that terminates a message, but only in the
1617 case where the client has not sent further data following the '.'
1618 (unfortunately, this is allowed). However, in many cases there won't be
1619 any further data because there won't be any more messages to send. A call
1620 to select() can be used: if it shows that the input is "ready", there is
1621 either input waiting, or the socket has been closed. An attempt to read
1622 the next input character can distinguish the two cases. Previously, Exim
1623 would have sent an OK response which the client would never have see.
1624 This could lead to message repetition. This fix should cure that, at
1625 least in a lot of common cases.
1627 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1628 advertised in response to EHLO.
1634 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1635 fixed by 4.65/MH/01 (is this a record?) are fixed:
1637 (i) An empty string was always treated as zero by the numeric comparison
1638 operators. This behaviour has been restored.
1640 (ii) It is documented that the numeric comparison operators always treat
1641 their arguments as decimal numbers. This was broken in that numbers
1642 starting with 0 were being interpreted as octal.
1644 While fixing these problems I realized that there was another issue that
1645 hadn't been noticed. Values of message_size_limit (both the global option
1646 and the transport option) were treated as octal if they started with 0.
1647 The documentation was vague. These values are now always treated as
1648 decimal, and I will make that clear in the documentation.
1654 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1655 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1658 MH/01 Don't check that the operands of numeric comparison operators are
1659 integers when their expansion is in "skipping" mode (fixes bug
1660 introduced by 4.64-PH/07).
1662 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1663 child addresses, Exim now panics and dies. Previously, because the count
1664 is held in a short int, deliveries were likely to be lost. As such a
1665 large number of recipients for a single message is ridiculous
1666 (performance will be very, very poor), I have chosen to impose a limit
1667 rather than extend the field.
1673 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1674 leftover -K file (the existence of which was triggered by #402).
1675 While we were at it, introduced process PID as part of the -K
1676 filename. This should rule out race conditions when creating
1679 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1680 processing considerably. Previous code took too long for large mails,
1681 triggering a timeout which in turn triggers #401.
1683 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1684 in the DK code in transports.c. sendfile() is not really portable,
1685 hence the _LINUX specificness.
1687 TF/01 In the add_headers option to the mail command in an Exim filter,
1688 there was a bug that Exim would claim a syntax error in any
1689 header after the first one which had an odd number of characters
1692 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1693 callout verification, Exim cached a "reject" for the entire domain. This
1694 is correct for most verifications, but it is not correct for a recipient
1695 verification with use_sender or use_postmaster set, because in that case
1696 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1697 case of MAIL FROM:<> rejection from other early rejections (e.g.
1698 rejection of HELO). When verifying a recipient using a non-null MAIL
1699 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1700 Whatever the result of the callout, the value of the domain cache is
1701 left unchanged (for any other kind of callout, getting as far as trying
1702 RCPT means that the domain itself is ok).
1704 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1707 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1708 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1709 session, the daemon ignores ECONNECT errors and logs others; it now
1710 ignores EPIPE as well.
1712 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1713 (quoted-printable decoding).
1715 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1716 later the small subsequent patch to fix an introduced bug.
1718 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1720 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1722 PH/08 An error is now given if message_size_limit is specified negative.
1724 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1725 to be given (somewhat) arbitrary names.
1727 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1730 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1731 miscellaneous code fixes
1733 PH/10 Added the log_reject_target ACL modifier to specify where to log
1736 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1737 hostname. This is wrong, because it relates to the incoming message (and
1738 probably the interface on which it is arriving) and not to the outgoing
1739 callout (which could be using a different interface). This has been
1740 changed to use the value of the helo_data option from the smtp transport
1741 instead - this is what is used when a message is actually being sent. If
1742 there is no remote transport (possible with a router that sets up host
1743 addresses), $smtp_active_hostname is used.
1745 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1746 tweaks were necessary in order to get it to work (see also 21 below):
1747 (a) The code assumed that strncpy() returns a negative number on buffer
1748 overflow, which isn't the case. Replaced with Exim's string_format()
1750 (b) There were several signed/unsigned issues. I just did the minimum
1751 hacking in of casts. There is scope for a larger refactoring.
1752 (c) The code used strcasecmp() which is not a standard C function.
1753 Replaced with Exim's strcmpic() function.
1754 (d) The code set only $1; it now sets $auth1 as well.
1755 (e) A simple test gave the error "authentication client didn't specify
1756 service in request". It would seem that Dovecot has changed its
1757 interface. Fortunately there's a specification; I followed it and
1758 changed what the client sends and it appears to be working now.
1760 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1763 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1764 address A is aliased to B and C, where B exists and C does not. Without
1765 -v the output is "A verified" because verification stops after a
1766 successful redirection if more than one address is generated. However,
1767 with -v the child addresses are also verified. Exim was outputting "A
1768 failed to verify" and then showing the successful verification for C,
1769 with its parentage. It now outputs "B failed to verify", showing B's
1770 parentage before showing the successful verification of C.
1772 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1773 look up a TXT record in a specific list after matching in a combined
1776 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1777 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1778 they consult the DNS. I had assumed they would set it the way they
1779 wanted; and indeed my experiments on Linux seem to show that in some
1780 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1781 To be on the safe side, however, I have now made the interface to
1782 host_find_byname() similar to host_find_bydns(), with an argument
1783 containing the DNS resolver options. The host_find_byname() function now
1784 sets these options at its start, just as host_find_bydns() does. The smtp
1785 transport options dns_qualify_single and dns_search_parents are passed to
1786 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1787 of host_find_byname() use the default settings of RES_DEFNAMES
1788 (qualify_single) but not RES_DNSRCH (search_parents).
1790 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1791 spool_read_header() do less string testing, by means of a preliminary
1792 switch on the second character of optional "-foo" lines. (This is
1793 overdue, caused by the large number of possibilities that now exist.
1794 Originally there were few.) While I was there, I also converted the
1795 str(n)cmp tests so they don't re-test the leading "-" and the first
1796 character, in the hope this might squeeze out yet more improvement.
1798 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1799 flag allowing group syntax was set by the header_syntax check but not
1800 turned off, possible causing trouble later; (2) The flag was not being
1801 set at all for the header_verify test, causing "group"-style headers to
1802 be rejected. I have now set it in this case, and also caused header_
1803 verify to ignore an empty address taken from a group. While doing this, I
1804 came across some other cases where the code for allowing group syntax
1805 while scanning a header line wasn't quite right (mostly, not resetting
1806 the flag correctly in the right place). These bugs could have caused
1807 trouble for malformed header lines. I hope it is now all correct.
1809 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1810 with the "reply" argument non-NULL. The code, however (which originally
1811 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1812 but it didn't always do it. This confused somebody who was copying the
1813 code for some other use. I have removed all the tests.
1815 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1816 feature that was used to support insecure browsers during the U.S. crypto
1817 embargo. It requires special client support, and Exim is probably the
1818 only MTA that supported it -- and would never use it because real RSA is
1819 always available. This code has been removed, because it had the bad
1820 effect of slowing Exim down by computing (never used) parameters for the
1821 RSA_EXPORT functionality.
1823 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1824 authenticator to fail if there's a tab character in the incoming data
1825 (there should never be unless someone is messing about, as it's supposed
1826 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1827 if the connection is using TLS or if the remote IP is the same as the
1828 local IP, and the "valid-client-cert option" if a client certificate has
1831 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1832 authenticators. This can be used for authorization after authentication
1833 succeeds. (In the case of plaintext, it servers for both authentication
1836 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1837 if any retry times were supplied.
1839 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1840 connection, where there is no client IP address to check. In this
1841 situation, the verify now always succeeds.
1843 PH/25 Applied John Jetmore's -Mset patch.
1845 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1847 PH/27 In a string expansion for a processed (not raw) header when multiple
1848 headers of the same name were present, leading whitespace was being
1849 removed from all of them, but trailing whitespace was being removed only
1850 from the last one. Now trailing whitespace is removed from each header
1851 before concatenation. Completely empty headers in a concatenation (as
1852 before) are ignored.
1854 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1855 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1857 PH/29 [Removed. This was a change that I later backed out, and forgot to
1858 correct the ChangeLog entry (that I had efficiently created) before
1859 committing the later change.]
1861 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1862 address errors (4xx response to RCPT) over the same connection as other
1863 messages routed to the same hosts. Such deliveries are always "forced",
1864 so retry times are not inspected. This resulted in far too many retries
1865 for the affected addresses. The effect occurred only when there were more
1866 hosts than the hosts_max_try setting in the smtp transport when it had
1867 the 4xx errors. Those hosts that it had tried were not added to the list
1868 of hosts for which the message was waiting, so if all were tried, there
1869 was no problem. Two fixes have been applied:
1871 (i) If there are any address or message errors in an SMTP delivery, none
1872 of the hosts (tried or untried) are now added to the list of hosts
1873 for which the message is waiting, so the message should not be a
1874 candidate for sending over the same connection that was used for a
1875 successful delivery of some other message. This seems entirely
1876 reasonable: after all the message is NOT "waiting for some host".
1877 This is so "obvious" that I'm not sure why it wasn't done
1878 previously. Hope I haven't missed anything, but it can't do any
1879 harm, as the worst effect is to miss an optimization.
1881 (ii) If, despite (i), such a delivery is accidentally attempted, the
1882 routing retry time is respected, so at least it doesn't keep
1883 hammering the server.
1885 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1886 in ${readsocket because some servers need this prod.
1888 PH/32 Added some extra debug output when updating a wait-xxx database.
1890 PH/33 The hint "could be header name not terminated by colon", which has been
1891 given for certain expansion errors for a long time, was not being given
1892 for the ${if def:h_colon_omitted{... case.
1894 PH/34 The spec says: "With one important exception, whenever a domain list is
1895 being scanned, $domain contains the subject domain." There was at least
1896 one case where this was not true.
1898 PH/35 The error "getsockname() failed: connection reset by peer" was being
1899 written to the panic log as well as the main log, but it isn't really
1900 panic-worthy as it just means the connection died rather early on. I have
1901 removed the panic log writing for the ECONNRESET error when getsockname()
1904 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1905 runs only) independently of the message's sender address. This meant
1906 that, if the 4xx error was in fact related to the sender, a different
1907 message to the same recipient with a different sender could confuse
1908 things. In particualar, this can happen when sending to a greylisting
1909 server, but other circumstances could also provoke similar problems.
1910 I have changed the default so that the retry time for these errors is now
1911 based a combination of the sender and recipient addresses. This change
1912 can be overridden by setting address_retry_include_sender=false in the
1915 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1916 remote server are returned as part of bounce messages. This was not
1917 happening for LMTP over a pipe (the lmtp transport), but now it is the
1918 same for both kinds of LMTP.
1920 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1921 in header lines that were in fact CNAMEs. This is no longer the case.
1923 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1924 and queue runs started by the daemon processed all messages. This has
1925 been fixed so that -R and -S can now usefully be given with -q<time>.
1927 PH/40 Import PCRE release 6.7 (fixes some bugs).
1929 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1931 PH/42 Give an error if -q is specified more than once.
1933 PH/43 Renamed the variables $interface_address and $interface_port as
1934 $received_ip_address and $received_port, to make it clear that these
1935 values apply to message reception, and not to the outgoing interface when
1936 a message is delivered. (The old names remain recognized, of course.)
1938 PH/44 There was no timeout on the connect() call when using a Unix domain
1939 socket in the ${readsocket expansion. There now is.
1941 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1942 be meaningful with "accept".
1944 SC/01 Eximstats V1.43
1945 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1947 SC/02 Eximstats V1.44
1948 Use a glob alias rather than an array ref in the generated
1949 parser. This improves both readability and performance.
1951 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1952 Collect SpamAssassin and rejection statistics.
1953 Don't display local sender or destination tables unless
1954 there is data to show.
1955 Added average volumes into the top table text output.
1957 SC/04 Eximstats V1.46
1958 Collect data on the number of addresses (recipients)
1959 as well as the number of messages.
1961 SC/05 Eximstats V1.47
1962 Added 'Message too big' to the list of mail rejection
1963 reasons (thanks to Marco Gaiarin).
1965 SC/06 Eximstats V1.48
1966 Mainlog lines which have GMT offsets and are too short to
1967 have a flag are now skipped.
1969 SC/07 Eximstats V1.49 (Alain Williams)
1970 Added the -emptyok flag.
1972 SC/08 Eximstats V1.50
1973 Fixes for obtaining the IP address from reject messages.
1975 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1976 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1977 whitesspace changes from 4.64-PH/27
1979 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1982 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1983 are found, allow negative numbers in numeric criteria)
1985 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1987 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1988 to match changes made in 4.64-PH/43
1990 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1992 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1993 "domain" is actually an IP address, and force a failure. This locks out
1994 those revolvers/nameservers that support "A-for-A" lookups, in
1995 contravention of the specifications.
1997 PH/48 When a host name was looked up from an IP address, and the subsequent
1998 forward lookup of the name timed out, the host name was left in
1999 $sender_host_name, contrary to the specification.
2001 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2002 restricted to single-key lookups, Exim was not diagnosing an error if
2003 * or *@ was used with a query-style lookup.
2005 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2007 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2008 long ago, but noone interested enough thought of it. Let's just say that
2009 the "1.1" means that there are some new functions that weren't there at
2010 some point in the past.
2012 PH/51 Error processing for expansion failure of helo_data from an smtp
2013 transport during callout processing was broken.
2015 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2016 tested/used via the -bh/-bhc/-bs options.
2018 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2019 bug, fixed in subsequent PCRE releases).
2021 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2022 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2024 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2030 SC/01 Use a glob alias rather than an array ref in eximstats generated
2031 parser. This improves both readability and performance.
2033 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2034 Don't display local sender or destination tables in eximstats unless
2035 there is data to show.
2036 Added average volumes into the eximstats top table text output.
2038 SC/03 Collect data on the number of addresses (recipients) as well
2039 as the number of messages in eximstats.
2041 TF/01 Correct an error in the documentation for the redirect router. Exim
2042 does (usually) call initgroups() when daemonizing.
2044 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2045 with consistent privilege compared to when running as a daemon.
2047 TF/03 Note in the spec that $authenticated_id is not set for local
2048 submissions from trusted users.
2050 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2051 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2053 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2054 by adding some example configuration directives to the default
2055 configuration file. A little bit of work is required to uncomment the
2056 directives and define how usernames and passwords are checked, but
2057 there is now a framework to start from.
2059 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2060 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2061 without this. I don't know how relevant this is to other LDAP libraries.
2063 PH/02 Add the verb name to the "unknown ACL verb" error.
2065 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2067 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2069 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2070 directory not expanded when it should be if an expanded home directory
2071 was set for the address (which is overridden by the transport).
2073 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2076 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2077 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2078 because it is too late at that time, and has no effect.
2080 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2081 security issue with \' (bugzilla #107). I could not use the
2082 PQescapeStringConn() function, because it needs a PGconn value as one of
2085 PH/08 When testing addresses using -bt, indicate those final addresses that
2086 are duplicates that would not cause an additional delivery. At least one
2087 person was confused, thinking that -bt output corresponded to deliveries.
2088 (Suppressing duplicates isn't a good idea as you lose the information
2089 about possibly different redirections that led to the duplicates.)
2091 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2092 systems where poll() doesn't work, in particular OS X.
2094 PH/10 Added more information to debugging output for retry time not reached.
2096 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2097 operations in malware.c.
2099 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2102 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2103 syslog_duplication set false, log lines that would normally be written
2104 both the the main log and to the reject log were not written to syslog at
2107 PH/14 In the default configuration, change the use of "message" in ACL warn
2108 statements to "add_header".
2110 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2111 not followed by a command (e.g. "seen endif").
2113 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2114 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2117 PH/17 Added extra conditions to the default value of delay_warning_condition
2121 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2122 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2123 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2126 The Auto-Submitted: and various List- headers are standardised, whereas I
2127 don't think Precedence: ever was.
2129 PH/18 Refactored debugging code in route_finduser() to show more information,
2130 in particular, the error code if getpwnam() issues one.
2132 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2133 This is apparently needed in addition to the PH/07 change above to avoid
2134 any possible encoding problems.
2136 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2137 but not after initializing Perl.
2139 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2140 output them only if debugging. By default they are written stderr,
2141 apparently, which is not desirable.
2143 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2146 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2149 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2151 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2152 authenticated or an ident call has been made. Suppress the default
2153 values for $authenticated_id and $authenticated_sender (but permit -oMai
2154 and -oMas) when testing with -bh.
2156 PH/24 Re-jigged the order of the tests in the default configuration so that the
2157 tests for valid domains and recipients precede the DNS black list and CSA
2158 tests, on the grounds that those ones are more expensive.
2160 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2161 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2162 This bug exists in every version of Exim that I still have, right back to
2165 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2166 However, an attempt to turn on case-sensitivity in a regex key by
2167 including (?-i) didn't work because the subject string was already
2168 lowercased, and the effects were non-intuitive. It turns out that a
2169 one-line patch can be used to allow (?-i) to work as expected.
2175 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2176 other effects) broke the use of negated acl sub-conditions.
2178 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2181 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2182 "Deny" causes Exim to reject the incoming connection with a 554 error.
2183 Unfortunately, if there is a major crisis, such as a disk failure,
2184 tcp-wrappers gives "deny", whereas what one would like would be some
2185 kind of temporary error. A kludge has been added to help with this.
2186 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2187 554 error is used if errno is still zero or contains ENOENT (which occurs
2188 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2191 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2193 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2194 errors. Otherwise a message that provokes a temporary error (when other
2195 messages do not) can cause a whole host to time out.
2197 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2198 addresses were routed directly to files or pipes from a redirect router.
2199 File deliveries just didn't batch; pipe deliveries might have suffered
2202 PH/06 A failure to get a lock for a hints database would erroneously always say
2203 "Failed to get write lock", even when it was really a read lock.
2205 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2206 of 0600. This has been changed to use the value of the lockfile_mode
2207 option (which defaults to 0600).
2209 PH/08 Applied small patch from the Sieve maintainer.
2211 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2212 folder from quota calculations, a direct delivery into this folder messed
2213 up the contents of the maildirsize file. This was because the regex was
2214 used only to exclude .Trash (or whatever) when the size of the mailbox
2215 was calculated. There was no check that a delivery was happening into an
2216 excluded directory. This bug has been fixed by ignoring all quota
2217 processing for deliveries into excluded directories.
2219 PH/10 Added the maildirfolder_create_regex option to appendfile.
2225 PH/01 The code for finding all the local interface addresses on a FreeBSD
2226 system running IPv6 was broken. This may well have applied to all BSD
2227 systems, as well as to others that have similar system calls. The broken
2228 code found IPv4 interfaces correctly, but gave incorrect values for the
2229 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2230 that it would not match correctly against @[] and not recognize the IPv6
2233 PH/02 The ipliteral router was not recognizing addresses of the form user@
2234 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2236 PH/03 Added disable_ipv6.
2238 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2239 decoded form, because it is most often used to construct To: headers
2240 lines in autoreplies, and the decoded form may well be syntactically
2241 invalid. However, $reply_address has leading white space removed, and all
2242 newlines turned into spaces so that the autoreply transport does not
2245 PH/05 If group was specified without a user on a router, and no group or user
2246 was specified on a transport, the group from the router was ignored.
2248 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2249 for visible compile-time settings that can be used to change these
2250 numbers, for those that want even more. Backwards compatibility with old
2251 spool files has been maintained. However, going back to a previous Exim
2252 release will lost any variables that are in spool files.
2254 PH/07 Two small changes when running in the test harness: increase delay when
2255 passing a TCP/IP connection to a new process, in case the original
2256 process has to generate a bounce, and remove special handling of
2257 127.0.0.2 (sic), which is no longer necessary.
2259 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2260 be the same on different OS.
2262 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2265 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2266 whether --show-vars was specified or not
2268 JJ/02 exipick: Added support for new ACL variable spool format introduced
2271 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2272 syntactically invalid From: or Reply-to: line, and a filter used this to
2273 generate an autoreply, and therefore failed to obtain an address for the
2274 autoreply, Exim could try to deliver to a non-existent relative file
2275 name, causing unrelated and misleading errors. What now happens is that
2276 it logs this as a hard delivery error, but does not attempt to create a
2279 PH/11 The exinext utility has a -C option for testing purposes, but although
2280 the given file was scanned by exinext itself; it wasn't being passed on
2281 when Exim was called.
2283 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2284 an end-of-file indication when reading a command response.
2286 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2287 compiled. In many other places in Exim, IPv6 addresses are always
2288 recognized, so I have changed this. It also means that IPv4 domain
2289 literals of the form [IPV4:n.n.n.n] are now always recognized.
2291 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2292 used if the router is not running as root, for example, when verifying at
2293 ACL time, or when using -bh. The debugging output from this situation was
2294 non-existent - all you got was a failure to exec. I have made two
2297 (a) Failures to set uid/gid, the current directory, or a process leader
2298 in a subprocess such as that created by queryprogram now generate
2299 suitable debugging ouput when -d is set.
2301 (b) The queryprogram router detects when it is not running as root,
2302 outputs suitable debugging information if -d is set, and then runs
2303 the subprocess without attempting to change uid/gid.
2305 PH/15 Minor change to Makefile for building test_host (undocumented testing
2308 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2309 additional section of a DNS packet that returns MX or SRV records.
2310 Instead, it always explicitly searches for A/AAAA records. This avoids
2311 major problems that occur when a DNS server includes only records of one
2312 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2313 fixed another bug: if SRV records were looked up and the corresponding
2314 address records were *not* found in the additional section, the port
2315 values from the SRV records were lost.
2317 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2318 using the correct key (the original address) when searching the retry
2319 rules in order to find which one to use for generating the retry hint.
2321 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2322 adding the default one. Similarly, if it contains a Reply-To: header, the
2323 errors_reply_to option, if set, is not used.
2325 PH/19 When calculating a retry time, Exim used to measure the "time since
2326 failure" by looking at the "first failed" field in the retry record. Now
2327 it does not use this if it is later than than the arrival time of the
2328 message. Instead it uses the arrival time. This makes for better
2329 behaviour in cases where some deliveries succeed, thus re-setting the
2330 "first failed" field. An example is a quota failure for a huge message
2331 when small messages continue to be delivered. Without this change, the
2332 "time since failure" will always be short, possible causing more frequent
2333 delivery attempts for the huge message than are intended.
2334 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2336 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2337 $1, $2, $3) because the numerical variables can be reset during some
2338 expansion items (e.g. "match"), thereby losing the authentication data.
2340 PH/21 Make -bV show the size of off_t variables so that the test suite can
2341 decide whether to run tests for quotas > 2G.
2343 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2344 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2345 filecount value is greater than 2G or if a quota value is greater than 2G
2346 on a system where the size of off_t is not greater than 4, a panic error
2349 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2350 never match. The debug and -bh output now contains an explicit error
2351 message indicating a malformed IPv4 address or mask.
2353 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2354 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2355 PH/23 above applies.
2357 PH/25 Do not write to syslog when running in the test harness. The only
2358 occasion when this arises is a failure to open the main or panic logs
2359 (for which there is an explicit test).
2361 PH/26 Added the /no_tell option to "control=freeze".
2363 PH/27 If a host name lookup failed very early in a connection, for example, if
2364 the IP address matched host_lookup and the reverse lookup yielded a name
2365 that did not have a forward lookup, an error message of the form "no IP
2366 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2367 could be logged. Now it outputs the IP address instead of "NULL".
2369 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2370 allows the sender and the authenticated sender to be set when
2371 submitting a message from within Exim. Since child_open_exim() is
2372 documented for local_scan(), the new function should be too.
2374 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2375 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2376 results in an empty string is now treated as unset.
2378 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2380 PH/31 Added sender_verify_fail logging option.
2382 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2383 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2384 tidied the source and removed it altogether.
2386 PH/33 When a queue run was abandoned because the load average was too high, a
2387 log line was always written; now it is written only if the queue_run log
2388 selector is set. In addition, the log line for abandonment now contains
2389 information about the queue run such as the pid. This is always present
2390 in "start" and "stop" lines but was omitted from the "abandon" line.
2392 PH/34 Omit spaces between a header name and the colon in the error message that
2393 is given when verify = headers_syntax fails (if there are lots of them,
2394 the message gets confusing).
2396 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2397 names, as there are now some PTR records that contain slashes. This check
2398 is only to protect against broken name servers that fall over on strange
2399 characters, so the fact that it applies to all lookups doesn't matter.
2401 PH/36 Now that the new test suite is complete, we can remove some of the
2402 special code in Exim that was needed for the old test suite. For example,
2403 sorting DNS records because real resolvers return them in an arbitrary
2404 order. The new test suite's fake resolver always returns records in the
2407 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2408 bounces) except when queue_only is set, to avoid logging races between
2409 the different processes.
2411 PH/38 Panic-die if .include specifies a non-absolute path.
2413 PH/39 A tweak to the "H" retry rule from its user.
2415 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2416 a label. They prevented compilation on older perls.
2418 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2419 a warning to be raised on newish perls.
2421 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2422 on queue. Changes to match documented behaviour of showing count of
2423 messages matching specified criteria.
2425 PH/40 Changed the default ident timeout from 30s to 5s.
2427 PH/41 Added support for the use of login_cap features, on those BSD systems
2428 that have them, for controlling the resources used by pipe deliveries.
2430 PH/42 The content-scanning code uses fopen() to create files in which to put
2431 message data. Previously it was not paying any attention to the mode of
2432 the files. Exim runs with umask(0) because the rest of the code creates
2433 files with open(), and sets the required mode explicitly. Thus, these
2434 files were ending up world-writeable. This was not a big issue, because,
2435 being within the spool directory, they were not world-accessible. I have
2436 created a function called modefopen, which takes an additional mode
2437 argument. It sets umask(777), creates the file, chmods it to the required
2438 mode, then resets the umask. All the relevant calls to fopen() in the
2439 content scanning code have been changed to use this function.
2441 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2442 to 24 hours. This avoids potential overflow problems when processing G
2443 and H retry rules. I suspect nobody ever tinkers with this value.
2445 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2447 PH/45 When the plaintext authenticator is running as a client, the server's
2448 challenges are checked to ensure they are valid base64 strings. By
2449 default, the authentication attempt is cancelled if an invalid string is
2450 received. Setting client_ignore_invalid_base64 true ignores these errors.
2451 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2452 they are received. Thus, the responses can be made to depend on the
2453 challenges. If an invalid string is ignored, an empty string is placed in
2456 PH/46 Messages that are created by the autoreply transport now contains a
2457 References: header, in accordance with RFCs 2822 and 3834.
2459 PH/47 Added authenticated_sender_force to the smtp transport.
2461 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2463 PH/49 Installed latest patch from the Sieve maintainer.
2465 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2466 was also set, the appendfile transport was unnecessarily scanning a
2467 directory of message files (e.g. for maildir delivery) to find the count
2468 of files (along with the size), even though it did not need this
2469 information. It now does the scan only if it needs to find either the
2470 size of the count of files.
2472 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2474 PH/52 Two bugs concerned with error handling when the smtp transport is
2477 (i) Exim was not creating retry information for temporary errors given
2478 for individual recipients after the DATA command when the smtp transport
2479 was used in LMTP mode. This meant that they could be retried too
2480 frequently, and not timed out correctly.
2482 (ii) Exim was setting the flag that allows error details to be returned
2483 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2484 recipients that were returned after the DATA command.
2486 PH/53 This is related to PH/52, but is more general: for any failing address,
2487 when detailed error information was permitted to be returned to the
2488 sender, but the error was temporary, then after the final timeout, only
2489 "retry timeout exceeded" was returned. Now it returns the full error as
2490 well as "retry timeout exceeded".
2492 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2493 do this, and (what is worse) MTAs that accept it.
2495 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2496 will now be deprecated.
2498 PH/56 New os.c-cygwin from the Cygwin maintainer.
2500 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2501 formats (previously only available in exim formats via -bpr, -bpru,
2502 and -bpra. Now also available in native and exiqgrep formats)
2504 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2505 with very large, slow to parse queues
2507 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2509 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2511 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2512 responses to authentication challenges, though it was showing the
2513 challenges; (ii) I've removed the CR characters from the debug output for
2516 PH/58 Allow for the insertion of a newline as well as a space when a string
2517 is turned into more than one encoded-word during RFC 2047 encoding. The
2518 Sieve code now uses this.
2520 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2521 data_4xx, lost_connection, tls_required.
2523 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2524 message was being sent as an SMTP response.
2526 PH/61 Add -l and -k options to exicyclog.
2528 PH/62 When verifying, if an address was redirected to one new address, so that
2529 verification continued, and the new address failed or deferred after
2530 having set something in $address_data, the value of $address_data was not
2531 passed back to the ACL. This was different to the case when no
2532 redirection occurred. The value is now passed back in both cases.
2534 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2535 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2536 use login_cap.h, so on its own it isn't the distinguishing feature. The
2537 new name refers directly to the setclassresources() function.
2539 PH/65 Added configuration files for NetBSD3.
2541 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2543 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2544 is preferred over IPv4.
2546 PH/68 The bounce_return_message and bounce_return_body options were not being
2547 honoured for bounces generated during the reception of non-SMTP messages.
2548 In particular, this applied to messages rejected by the ACL. This bug has
2549 been fixed. However, if bounce_return_message is true and bounce_return_
2550 body is false, the headers that are returned for a non-SMTP message
2551 include only those that have been read before the error was detected.
2552 (In the case of an ACL rejection, they have all been read.)
2554 PH/69 The HTML version of the specification is now built in a directory called
2555 spec_html instead of spec.html, because the latter looks like a path with
2556 a MIME-type, and this confuses some software.
2558 PH/70 Catch two compiler warnings in sieve.c.
2560 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2561 function verify_get_ident() calls ip_connect() to connect a socket, but
2562 if the "connect()" function timed out, ip_connect() used to close the
2563 socket. However, verify_get_ident() also closes the socket later, and in
2564 between Exim writes to the log, which may get opened at this point. When
2565 the socket was closed in ip_connect(), the log could get the same file
2566 descriptor number as the socket. This naturally causes chaos. The fix is
2567 not to close the socket in ip_connect(); the socket should be closed by
2568 the function that creates it. There was only one place in the code where
2569 this was missing, in the iplookup router, which I don't think anybody now
2570 uses, but I've fixed it anyway.
2572 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2573 well as to direct DNS lookups. Otherwise the handling of names in host
2574 lists is inconsistent and therefore confusing.
2580 PH/01 Two changes to the default runtime configuration:
2582 (1) Move the checks for relay_from_hosts and authenticated clients from
2583 after to before the (commented out) DNS black list checks.
2585 (2) Add control=submission to the relay_from_hosts and authenticated
2586 clients checks, on the grounds that messages accepted by these
2587 statements are most likely to be submissions.
2589 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2591 (1) Generate an error if the third argument for the ${prvs expansion is
2594 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2597 (3) Reset the variables that are obtained from the first argument of
2598 ${prvscheck and used in the second argument before leaving the code,
2599 because their memory is reclaimed, so using them afterwards may do
2602 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2603 one (it's much easier than Tom thought :-).
2605 (5) Because of (4), we can now allow for the use of $prvscheck_result
2606 inside the third argument.
2608 PH/03 For some reason, the default setting of PATH when running a command from
2609 a pipe transport was just "/usr/bin". I have changed it to
2612 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2613 anything to be listed in the output from -bV.
2615 PH/05 When a filter generated an autoreply, the entire To: header line was
2616 quoted in the delivery log line, like this:
2618 => >A.N.Other <ano@some.domain> <original@ddress> ...
2620 This has been changed so that it extracts the operative address. There
2621 may be more than one such address. If so, they are comma-separated, like
2624 => >ano@some.domain,ona@other.domain <original@ddress> ...
2626 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2627 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2628 not being looked up in the rDNS to get a host name, Exim was showing the
2629 IP address twice in Received: lines, even though the IP addresses were
2630 identical. For example:
2632 Received: from [1.2.3.4] (helo=[1.2.3.4])
2634 However, if the real host name was known, it was omitting the HELO data
2635 if it matched the actual IP address. This has been tidied up so that it
2636 doesn't show the same IP address twice.
2638 PH/07 When both +timestamp and +memory debugging was on, the value given by
2639 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2640 called by the debug printing, thereby overwriting the timestamp buffer.
2641 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2643 PH/08 When the original message was included in an autoreply transport, it
2644 always said "this is a copy of the message, including all the headers",
2645 even if body_only or headers_only was set. It now gives an appropriate
2648 PH/09 Applied a patch from the Sieve maintainer which:
2650 o fixes some comments
2651 o adds the (disabled) notify extension core
2652 o adds some debug output for the result of if/elsif tests
2653 o points to the current vacation draft in the documentation
2654 and documents the missing references header update
2658 o fixes a bug in processing the envelope test (when testing
2659 multiple envelope elements, the last element determinted the
2662 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2663 Electronic Mail") by including:
2665 Auto-submitted: auto-generated
2667 in the messages that it generates (bounce messages and others, such as
2668 warnings). In the case of bounce messages for non-SMTP mesages, there was
2669 also a typo: it was using "Auto_submitted" (underscore instead of
2670 hyphen). Since every message generated by Exim is necessarily in response
2671 to another message, thes have all been changed to:
2673 Auto-Submitted: auto-replied
2675 in accordance with these statements in the RFC:
2677 The auto-replied keyword:
2679 - SHOULD be used on messages sent in direct response to another
2680 message by an automatic process,
2682 - MUST NOT be used on manually-generated messages,
2684 - MAY be used on Delivery Status Notifications (DSNs) and Message
2685 Disposition Notifications (MDNs),
2687 - MUST NOT be used on messages generated by automatic or periodic
2688 processes, except for messages which are automatic responses to
2691 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2692 to the default Received: header definition.
2694 PH/12 Added log selector acl_warn_skipped (default on).
2696 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2697 variables because (a) they are in the wrong storage pool and (b) even if
2698 they were copied, it wouldn't work properly because of the caching.
2700 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2701 checking when decoding. Apparently there are clients that generate
2702 overlong encoded strings. Why am I not surprised?
2704 PH/15 If the first argument of "${if match_address" was not empty, but did not
2705 contain an "@" character, Exim crashed. Now it writes a panic log message
2706 and treats the condition as false.
2708 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2710 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2711 extension "envelope-auth". The code is finished and in agreement with
2712 other implementations, but there is no documentation so far and in fact,
2713 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2714 not changing the active code.
2716 Print executed "if" and "elsif" statements when debugging is used. This
2717 helps a great deal to understand what a filter does.
2719 Document more things not specified clearly in RFC3028. I had all this
2720 sorted out, when out of a sudden new issues came to my mind. Oops."
2722 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2725 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2726 canonical form (as documented). However, after a host name lookup from
2727 the IP address, check_host() was doing a simple string comparison with
2728 addresses acquired from the DNS when checking that the found name did
2729 have the original IP as one of its addresses. Since any found IPv6
2730 addresses are likely to be in abbreviated form, the comparison could
2731 fail. Luckily, there already exists a function for doing the comparison
2732 by converting both addresses to binary, so now that is used instead of
2733 the text comparison.
2735 PH/20 There was another similar case to PH/19, when a complete host name was
2736 given in a host list; looking up its IP address could give an abbreviated
2737 form, whereas the current host's name might or might not be abbreviated.
2738 The same fix has been applied.
2744 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2745 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2748 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2749 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2751 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2753 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2754 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2755 Also, for "Auto-Submitted", treat anything other than "no" as
2756 non-personal, in accordance with RFC 3834. (Previously it treated
2757 anything starting "auto-" as non-personal.)
2759 TF/01 The control=submission/name=... option had a problem with syntax
2760 errors if the name included a slash character. The /name= option
2761 now slurps the rest of the string, so it can include any characters
2762 but it must come last in the list of options (after /sender_retain
2765 PH/05 Some modifications to the interface to the fake nameserver for the new
2773 TK/01 Added the "success_on_redirect" address verification option. See
2774 NewStuff for rationale and an example.
2776 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2778 PH/02 Patch to exigrep to allow it to work on syslog lines.
2780 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2781 fread() to skip over the body file's header line, because in Cygwin the
2782 header line is locked and is inaccessible.
2784 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2785 co-exist for some time) to make it clear that it is the Exim ID that is
2786 referenced, not the Message-ID: header line.
2788 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2789 string_format() function, because snprintf() does not exist on all
2792 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2793 ${stat: expansion item.
2795 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2796 protocol synchronization error", to keep the pedants happy.
2798 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2799 well as for IRIX systems, when gcc is being used. See the host.c source
2802 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2804 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2807 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2808 transport and to the smtp transport in LMTP mode.
2810 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2812 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2813 run a filter in a subprocess. This could lead to confusion in subsequent
2814 lookups in the parent process. There should also be a search_tidyup() at
2815 the end of the subprocess.
2817 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2818 only if the host matched helo_try_verify_hosts, which caused the
2819 verification to occur when the EHLO/HELO command was issued. The ACL just
2820 tested the remembered result. Now, if a previous verification attempt has
2821 not happened, "verify = helo" does it there and then.
2823 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2825 TK/03 Fix log output including CR from clamd.
2827 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2828 exist provoked a memory error which could cause a segfault.
2830 PH/15 Installed PCRE 6.2
2832 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2834 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2835 of the problem. Specifically, suggested +O2 rather than +O1 for the
2838 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2840 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2841 record, the port was not correctly logged when the outgoing_port log
2842 selector was set (it logged the transort's default port).
2844 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2845 fallback_hosts, and "hosts" in the smtp transport.
2847 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2848 host errors such as "Connection refused".
2850 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2851 authentication with radiusclient 0.4.9:
2853 - Error returned from rc_read_config was caught wrongly
2854 - Username/password not passed on to radius server due to wrong length.
2856 The presumption is that some radiusclient API changes for 4.51/PH/17
2857 were not taken care of correctly. The code is still untested by me (my
2858 Linux distribution still has 0.3.2 of radiusclient), but it was
2859 contributed by a Radius user.
2861 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2862 expanding the "port" option of the smtp transport.
2864 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2865 while reading a MIME header. Thanks to Tom Hughes for a patch.
2867 PH/24 Include config.h inside local_scan.h so that configuration settings are
2870 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2871 that in an ACL for RCPT (for example), you can examine exactly what was
2874 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2875 commands, but it was not correctly comparing the address with the actual
2876 client host address. Thus, it would show the EHLO address in Received:
2877 header lines when this was not necessary.
2879 PH/27 Added the % operator to ${eval:}.
2881 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2882 it should be ignoring failures (because with -C, for example, it has lost
2883 privilege). It wasn't ignoring creation failures other than "already
2886 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2889 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2890 because an input line was too long, either on its own, or by virtue of
2891 too many continuations, the temporary file was not being removed, and the
2892 return code was incorrect.
2894 PH/31 Missing "BOOL" in function definition in filtertest.c.
2896 PH/32 Applied Sieve patches from the maintainer.
2898 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2900 PH/33 Added "verify = not_blind".
2902 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2903 Local/Makefile (with some defaults set). These are used in built scripts
2904 such as exicyclog, but they have never been used in the exim_install
2905 script (though there are many overriding facilities there). I have
2906 arranged that the exim_install script now takes note of these two
2909 PH/35 Installed configuration files for Dragonfly.
2911 PH/36 When a locally submitted message by a trusted user did not contain a
2912 From: header, and the sender address was obtained from -f or from an SMTP
2913 MAIL command, and the trusted user did not use -F to supply a sender
2914 name, $originator_name was incorrectly used when constructing a From:
2915 header. Furthermore, $originator_name was used for submission mode
2916 messages from external hosts without From: headers in a similar way,
2917 which is clearly wrong.
2919 PH/37 Added control=suppress_local_fixups.
2921 PH/38 When log_selector = +received_sender was set, and the addition of the
2922 sender made the log line's construction buffer exactly full, or one byte
2923 less than full, an overflow happened when the terminating "\n" was
2926 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2927 when the result of a list match is failure because a DNS lookup failed.
2929 PH/40 RM_COMMAND is now used in the building process.
2931 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2932 the "build-* directories that it finds.
2934 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2935 address in a domain literal was a prefix of an interface address.
2937 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2938 when verifying a sender address, unless rewrite_headers is false.
2940 PH/44 Wrote a long comment about why errors_to addresses are verified as
2941 recipients, not senders.
2943 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2944 the ratelimit ACL was added.
2946 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2948 PH/46 Added extra information about PostgreSQL errors to the error string.
2950 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2951 avoiding the need to install special zones in a real server. This is
2952 backwards compatible; if it can't find the fake resolver, it drops back.
2953 Thus, both old and new test suites can be run.
2955 TF/02 Added util/ratelimit.pl
2957 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2958 clock is set back in time.
2960 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2961 Candler <B.Candler@pobox.com>.
2963 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2964 OK for addresses that are the result of redirections.
2966 PH/48 A number of further additions for the benefit of the new test suite,
2967 including a fake gethostbyname() that interfaces to the fake DNS resolver
2970 TF/06 The fix for widen_domains has also been applied to qualify_single and
2971 search_parents which are the other dnslookup options that can cause
2974 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2977 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2979 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2980 service" if it encountered an MX record with an empty target hostname.
2981 The message is now "an MX or SRV record indicated no SMTP service".
2983 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2984 if the DNS of the sending site is misconfigured. This is quite a
2985 common situation. This change restores the behaviour of treating a
2986 helo verification defer as a failure.
2988 PH/51 If self=fail was set on a router, the bounce message did not include the
2989 actual error message.
2995 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2997 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2998 command itself ended in error, the underlying message about the transport
2999 filter timeout was being overwritten with the pipe command error. Now the
3000 underlying error message should be appended to the second error message.
3002 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3004 PH/02 Reduce the amount of output that "make" produces by default. Full output
3005 can still be requested.
3007 PH/03 The warning log line about a condition test deferring for a "warn" verb
3008 was being output only once per connection, rather than after each
3009 occurrence (because it was using the same function as for successful
3010 "warn" verbs). This seems wrong, so I have changed it.
3012 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3013 it should not have, which might have caused a crash in the right
3014 circumstances, but probably never did.
3016 PH/04 Installed a modified version of Tony Finch's patch to make submission
3017 mode fix the return path as well as the Sender: header line, and to
3018 add a /name= option so that you can make the user's friendly name appear
3021 TF/03 Added the control = fakedefer ACL modifier.
3023 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3024 Mark Lowes for thorough testing.
3026 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3028 TK/03 Merged latest SRS patch from Miles Wilton.
3030 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3031 with the definition in sysexits.h (which is #included earlier).
3032 Fortunately, Exim does not actually use EX_OK. The code used to try to
3033 preserve the sysexits.h value, by assumimg that macro definitions were
3034 scanned for macro replacements. I have been disabused of this notion,
3035 so now the code just undefines EX_OK before #including unistd.h.
3037 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3038 in the smtp transport. When a block could not be written in a single
3039 write() function, the timeout was being re-applied to each part-write.
3040 This seems wrong - if the receiver was accepting one byte at a time it
3041 would take for ever. The timeout is now adjusted when this happens. It
3042 doesn't have to be particularly precise.
3044 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3045 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3047 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3048 without a domain if the check to <postmaster@domain> fails.
3050 SC/01 Eximstats: added -xls and the ability to specify output files
3051 (patch written by Frank Heydlauf).
3053 SC/02 Eximstats: use FileHandles for outputing results.
3055 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3057 SC/04 Eximstats: fixed display of large numbers with -nvr option
3059 SC/05 Eximstats: fixed merging of reports with empty tables.
3061 SC/06 Eximstats: added the -include_original_destination flag
3063 SC/07 Eximstats: removed tabs and trailing whitespace.
3065 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3067 TK/06 MBOX spool code: Add real "From " MBOX separator line
3068 so the .eml file is really in mbox format (even though
3069 most programs do not really care). Patch from Alex Miller.
3071 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3072 The latter is generated from $received_to and is only set if the
3073 message has one envelope recipient. SA can use these headers,
3074 obviously out-of-the-box. Patch from Alex Miller.
3076 PH/08 The ${def test on a variable was returning false if the variable's
3077 value was "0", contrary to what the specification has always said!
3078 The result should be true unless the variable is empty.
3080 PH/09 The syntax error of a character other than { following "${if
3081 def:variable_name" (after optional whitespace) was not being diagnosed.
3082 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3083 accidental colon was present, for example, could give incorrect results.
3085 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3086 result is used (not including appendfile, where other changes are about
3089 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3090 This involved changing a lot of size variables from int to off_t. It
3091 should work with maildirs and everything.
3093 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3094 spamd dying while we are connected to it.
3096 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3099 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3100 function for BDB 4.3.
3102 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3104 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3105 the sources for PCRE have been re-arranged and more files are now
3108 PH/15 The code I had for printing potentially long long variables in PH/11
3109 above was not the best (it lost precision). The length of off_t variables
3110 is now inspected at build time, and an appropriate printing format (%ld
3111 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3112 to be "long long int" or "long int". This is needed for the internal
3113 formatting function string_vformat().
3115 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3116 the configuration file to be ":syslog", then the script "guesses" where
3117 the logs files are, rather than using the compiled in default. In our
3118 case the guess is not the same as the compiled default, so the script
3119 suddenly stopped working when I started to use syslog. The patch checks
3120 to see if log_file_path is "". If so, it attempts to read it from exim
3121 with no configuration file to get the compiled in version, before it
3122 falls back to the previous guessing code."
3124 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3125 implementing BATV in an Exim configuration. See NewStuff for the gory
3128 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3129 Makefile that are specific to HP-UX.
3131 PH/18 If the "use_postmaster" option was set for a recipient callout together
3132 with the "random" option, the postmaster address was used as the MAIL
3133 FROM address for the random test, but not for the subsequent recipient
3134 test. It is now used for both.
3136 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3137 patch removes a few documentation additions to RFC 3028, because the
3138 latest draft now contains them. It adds the new en;ascii-case comparator
3139 and a new error check for 8bit text in MIME parts. Comparator and
3140 require names are now matched exactly. I enabled the subaddress
3141 extension, but it is not well tested yet (read: it works for me)."
3143 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3144 rework some of the code of TK/09 above to avoid the hardwired use of
3145 "%lld" and "long long". Replaced the call to snprintf() with a call to
3148 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3149 records point to non-existent hosts", "retry timeout exceeded", and
3150 "retry time not reached for any host after a long failure period".
3152 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3153 experimental DomainKeys support:
3155 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3156 (2) On an error such as an illegally used "control", the wrong name for
3157 the control was given.
3159 These problems did NOT occur unless DomainKeys support was compiled.
3161 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3163 PH/24 Added ${if match_ip condition.
3165 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3166 databases so that it will be absolutely obvious if a crash occurs in the
3167 DB library. This is a regular occurrence (often caused by mis-matched
3170 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3171 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3172 picked up on a user's system that detects such things. There doesn't seem
3173 to be a gcc warning option for this - only an attribute that has to be
3174 put on the function's prototype. It seems that in Fedora Core 4 they have
3175 set this on a number of new functions. No doubt there will be more in due
3178 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3179 specify a transport. However, if an address that was verified by such a
3180 router was the subject of a callout, Exim crashed because it tried to
3181 read the rcpt_include_affixes from the non-existent transport. Now it
3182 just assumes that the setting of that option is false. This bug was
3183 introduced by 4.51/PH/31.
3185 PH/28 Changed -d+all to exclude +memory, because that information is very
3186 rarely of interest, but it makes the output a lot bigger. People tend to
3187 do -d+all out of habit.
3189 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3190 code in os-type was giving problems when libc.so lives in lib64, like on
3193 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3194 aren't the modern standard, and it seems that some systems' include files
3195 don't always have them. Exim was already checking for some of the newer
3196 ones like T_AAAA, and defining it itself. I've added checks for all the
3197 record types that Exim uses.
3199 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3200 not automatically generating a new one, as it is supposed to. This
3201 prevented TLS from working. If the file did exist, but contained invalid
3202 data, a new version was generated, as expected. It was only the case of a
3203 non-existent file that was broken.
3205 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3206 with a change in libdomainkeys > 0.64.
3208 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3209 from DNS. If the selector record carries the flag, it now has
3210 precedence over the domain-wide flag.
3212 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3214 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3215 the use of a port name that isn't defined in /etc/services) occurred, the
3216 message was deferred as in a normal delivery, and thus remained on the
3217 spool, instead of being failed because of the mua_wrapper setting. This
3218 is now fixed, and I tidied up some of the mua_wrapper messages at the
3221 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3222 the messages in a hash of arrays rather than using individual hashes.
3223 This is a bit cleaner and results in dramatic memory savings, albeit
3224 at a slight CPU cost.
3226 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3227 as requested by Marc Sherman.
3229 SC/10 Eximstats: added histograms for user specified patterns as requested
3232 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3234 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3235 fopen() in the content-scanning modules that did not already have it.
3241 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3242 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3244 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3246 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3248 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3249 to test Sieve filters that use "vacation".
3251 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3252 that changes the way the GnuTLS parameters are stored in the cache file.
3253 The new format can be generated externally. For backward compatibility,
3254 if the data in the cache doesn't make sense, Exim assumes it has read an
3255 old-format file, and it generates new data and writes a new file. This
3256 means that you can't go back to an older release without removing the
3259 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3260 work if there are any delivery delays because "one_time" forces the
3261 parent to be marked "delivered", so its unseen clone is never tried
3262 again. For this reason, Exim now forbids the simultaneous setting of
3265 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3266 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3267 entry if you want to know the details. The fix, however, neglected to
3268 consider the case when local delivery batching is involved. The test for
3269 "previously delivered" was not happening when checking to see if an
3270 address could be batched with a previous (undelivered) one; under
3271 certain circumstances this could lead to multiple deliveries to the same
3274 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3275 in its include files, and this causes problems building Exim.
3277 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3278 header_syntax) but Exim was just ignoring anything given after a slash.
3279 In particular, this caused confusion with an attempt to use "verify =
3280 reverse_host_lookup/defer_ok". An error is now given when options are
3281 supplied for verify items that do not have them. (Maybe reverse_host_
3282 lookup should have a defer_ok option, but that's a different point.)
3284 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3285 defined by RFC 821) to 2048, because there were problems with some AUTH
3286 commands, and RFC 1869 says the size should be increased for extended
3287 SMTP commands that take arguments.
3289 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3292 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3293 "unknown" error; now it says that the functionality isn't in the binary.
3295 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3296 an address' error message when a string expansion fails (syntax or
3297 whatever). Otherwise the password may appear in the log. Following change
3298 PH/42 below, there is no longer a chance of it appearing in a bounce
3301 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3303 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3304 of its addresses was ever tried. (Bugzilla bug #2.)
3306 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3307 the result incorrectly in the debug output. (It correctly added a newline
3308 to what was transported.)
3310 TF/01 Added $received_time.
3312 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3313 commented out examples of how to interface to a virus scanner and to
3314 SpamAssassin. Also added commented examples of av_scanner and
3315 spamd_address settings.
3317 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3318 and controls are allowed in which ACLs. There were a couple of minor
3319 errors. Some of the entries in the conditions table (which is a table of
3320 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3321 negation of where the condition IS allowed.
3323 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3325 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3326 header file does not have a version number, so I've had to invent a new
3327 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3328 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3329 radiusclient), but it was contributed by a Radius user.
3331 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3332 files or directories, for OpenSSL.
3334 PH/19 When an Exim process that is running as root has to create an Exim log
3335 file, it does so in a subprocess that runs as exim:exim so as to get the
3336 ownership right at creation (otherwise, other Exim processes might see
3337 the file with the wrong ownership). There was no test for failure of this
3338 fork() call, which would lead to the process getting stuck as it waited
3339 for a non-existent subprocess. Forks do occasionally fail when resources
3340 run out. I reviewed all the other calls to fork(); they all seem to check
3343 PH/20 When checking for unexpected SMTP input at connect time (before writing
3344 the banner), Exim was not dealing correctly with a non-positive return
3345 from the read() function. If the client had disconnected by this time,
3346 the result was a log entry for a synchronization error with an empty
3347 string after "input=" when read() returned zero. If read() returned -1
3348 (an event I could not check), uninitialized data bytes were printed.
3349 There were reports of junk text (parts of files, etc) appearing after
3352 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3354 PH/22 Added support for macro redefinition, and (re)definition in between
3355 driver and ACL definitions.
3357 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3358 forgetting to use the resulting value; it was using the unexpanded value.
3360 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3361 hadn't been configured. The fix is from Juergen Kreileder, who
3362 understands it better than I do:
3364 "Here's what I see happening with three configured cyrus_sasl
3365 authenticators configured (plain, login, cram-md5):
3367 On startup auth_cyrus_sasl_init() gets called for each of these.
3368 This means three calls to sasl_listmech() without a specified mech_list.
3369 => SASL tests which mechs of all available mechs actually work
3370 => three warnings about OTP not working
3371 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3373 With the patch, sasl_listmech() also gets called three times. But now
3374 SASL's mech_list option is set to the server_mech specified in the the
3375 authenticator. Or in other words, the answer from sasl_listmech()
3376 gets limited to just the mech you're testing for (which is different
3378 => the return list contains just 'plain' or 'login', 'cram-md5' or
3379 nothing depending on the value of ob->server_mech.
3381 I've just tested the patch: Authentication still works fine,
3382 unavailable mechs specified in the exim configuration are still
3383 caught, and the auth.log warnings about OTP are gone."
3385 PH/25 When debugging is enabled, the contents of the command line are added
3386 to the debugging output, even when log_selector=+arguments is not
3389 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3390 answer is "GNU", and only if the return is "GNU/something" is the answer
3393 PH/27 $acl_verify_message is now set immediately after the failure of a
3394 verification in an ACL, and so is available in subsequent modifiers. In
3395 particular, the message can be preserved by coding like this:
3397 warn !verify = sender
3398 set acl_m0 = $acl_verify_message
3400 Previously, $acl_verify_message was set only while expanding "message"
3401 and "log_message" when a very denied access.
3403 PH/28 Modified OS/os.c-Linux with
3405 -#ifndef OS_LOAD_AVERAGE
3406 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3408 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3409 nomenclature these days.)
3411 PH/29 Installed patch from the Sieve maintainer that adds the options
3412 sieve_useraddress and sieve_subaddress to the redirect router.
3414 PH/30 In these circumstances:
3415 . Two addresses routed to the same list of hosts;
3416 . First host does not offer TLS;
3417 . First host accepts first address;
3418 . First host gives temporary error to second address;
3419 . Second host offers TLS and a TLS session is established;
3420 . Second host accepts second address.
3421 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3422 and peerdn, if requested) that were in fact used only for the second
3425 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3426 attention to any local part prefix or suffix that was matched by the
3427 router that accepted the address. It now behaves in the same way as it
3428 does for delivery: the affixes are removed from the local part unless
3429 rcpt_include_affixes is set on the transport.
3431 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3432 timeout during the DATA phase of an incoming message.
3434 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3435 applied a patch sanctioned by the Sieve maintainer.
3437 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3438 the uid or gid is negative. A case of a negative gid caused this to be
3439 noticed. The fix allows for either to be negative.
3441 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3442 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3445 PH/36 The change PH/12 above was broken. Fixed it.
3447 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3448 the grounds that routing the same address twice would always produce the
3449 same answer. This might have been true once, but it is certainly no
3450 longer true now. Routing a child address may depend on the previous
3451 routing that produced that child. Some complicated redirection strategies
3452 went wrong when messages had multiple recipients, and made Exim's
3453 behaviour dependent on the order in which the addresses were given.
3455 I have moved the duplicate checking until after the routing is complete.
3456 Exim scans the addresses that are assigned to local and remote
3457 transports, and removes any duplicates. This means that more work will be
3458 done, as duplicates will always all be routed, but duplicates are
3459 presumably rare, so I don't expect this is of any significance.
3461 For deliveries to pipes, files, and autoreplies, the duplicate checking
3462 still happens during the routing process, since they are not going to be
3465 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3466 It corrects a timeout issue with spamd. This is Ian's comment: "The
3467 background is that sometimes spamd either never reads data from a
3468 connection it has accepted, or it never writes response data. The exiscan
3469 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3470 blindly assumes that writes won't block so it may never time out."
3472 PH/39 Allow G after quota size as well as K and M.
3474 PH/40 The value set for $authenticated_id in an authenticator may not contain
3475 binary zeroes or newlines because the value is written to log lines and
3476 to spool files. There was no check on this. Now the value is run through
3477 the string_printing() function so that such characters are converted to
3478 printable escape sequences.
3480 PH/41 $message_linecount is a new variable that contains the total number of
3481 lines in the message. Compare $body_linecount, which is the count for the
3484 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3485 bounce and delay warning messages, except in certain special cases, which
3488 (a) An SMTP error message from a remote host;
3489 (b) A message specified in a :fail: redirection;
3490 (c) A message specified in a "fail" command in a system filter;
3491 (d) A message specified in a FAIL return from the queryprogram router;
3492 (e) A message specified by the cannot_route_message router option.
3494 In these cases only, Exim does include the error details in bounce and
3495 warning messages. There are also a few cases where bland messages such
3496 as "unrouteable address" or "local delivery error" are given.
3498 PH/43 $value is now also set for the "else" part of a ${run expansion.
3500 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3501 being worked on, but at least Exim now implements the latest version to
3504 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3505 process to complete was treated as a delivery failure, a timeout while
3506 writing the message to the pipe was logged, but erroneously treated as a
3507 successful delivery. Such timeouts include transport filter timeouts. For
3508 consistency with the overall process timeout, these timeouts are now
3509 treated as errors, giving rise to delivery failures by default. However,
3510 there is now a new Boolean option for the pipe transport called
3511 timeout_defer, which, if set TRUE, converts the failures into defers for
3512 both kinds of timeout. A transport filter timeout is now identified in
3515 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3516 systems where "make" and "gmake" are different, calling "gmake" at top
3517 level broke things. I've arranged for the value of $(MAKE) to be passed
3518 from the Makefile to this script so that it can call the same version of
3522 A note about Exim versions 4.44 and 4.50
3523 ----------------------------------------
3525 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3526 changes of various kinds. As a consequence, a big documentation update was
3527 needed. This delayed the release for rather longer than seemed good, especially
3528 in the light of a couple of (minor) security issues. Therefore, the changes
3529 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3530 release. So 4.44 and 4.50 are in effect two different branches that both start
3533 I have left the 4.50 change log unchanged; it contains all the changes since
3534 4.43. The change log for 4.44 is below; many of its items are identical to
3535 those for 4.50. This seems to be the most sensible way to preserve the
3536 historical information.
3542 1. Minor wording change to the doc/README.SIEVE file.
3544 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3545 computation of the current number of files was incorrect.
3547 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3548 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3551 4. Give more explanation in the error message when the command for a transport
3552 filter fails to execute.
3554 5. There are several places where Exim runs a non-Exim command in a
3555 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3556 was being done only for the command run by the queryprogram router. It is
3557 now done for all such subprocesses. The other cases are: ${run, transport
3558 filters, and the commands run by the lmtp and pipe transports.
3560 6. Added CONFIGURE_GROUP build-time option.
3562 7. Some older OS have a limit of 256 on the maximum number of file
3563 descriptors. Exim was using setrlimit() to set 1000 as a large value
3564 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3565 systems. I've change it so that if it can't get 1000, it tries for 256.
3567 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3568 was an oversight, and furthermore, ever since the addition of extra
3569 controls (e.g. 4.43/32), the checks on when to allow different forms of
3570 "control" were broken. There should now be diagnostics for all cases when a
3571 control that does not make sense is encountered.
3573 9. Added the /retain_sender option to "control=submission".
3575 10. $recipients is now available in the predata ACL (oversight).
3577 11. Tidy the search cache before the fork to do a delivery from a message
3578 received from the command line. Otherwise the child will trigger a lookup
3579 failure and thereby defer the delivery if it tries to use (for example) a
3580 cached ldap connection that the parent has called unbind on.
3582 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3583 of $address_data from the recipient verification was clobbered by the
3584 sender verification.
3586 13. The value of address_data from a sender verification is now available in
3587 $sender_address_data in subsequent conditions in the ACL statement.
3589 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3591 15. Added a new option "connect=<time>" to callout options, to set a different
3594 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3595 was its contents. (It was OK if the option was not defined at all.)
3597 17. A "Completed" log line is now written for messages that are removed from
3598 the spool by the -Mrm option.
3600 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3601 information about exactly what failed.
3603 19. Added -dd to debug only the daemon process.
3605 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3606 handles timeouts, both on the server side and network timeouts. Renamed the
3607 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3609 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3610 It is now set to "smtps".
3612 22. $host_address is now set to the target address during the checking of
3613 ignore_target_hosts.
3615 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3616 being passed; this would have caused $sender_host_name to have been used if
3617 matching the list had actually called for a host name (not very likely,
3618 since this list is usually IP addresses). A host name is now passed as
3621 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3622 code that specifies a non-restarting handler (typically sigaction() in
3623 modern systems) in an attempt to fix a rare and obscure crash bug.
3625 25. Narrowed the window for a race in the daemon that could cause it to ignore
3626 SIGCHLD signals. This is not a major problem, because they are used only to
3627 wake it up if nothing else does.
3629 26. A malformed maildirsize file could cause Exim to calculate negative values
3630 for the mailbox size or file count. Odd effects could occur as a result.
3631 The maildirsize information is now recalculated if the size or filecount
3634 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3635 support for a long time. Removed HAVE_SYS_VFS_H.
3637 28. Installed the latest version of exipick from John Jetmore.
3639 29. In an address list, if the pattern was not a regular expression, an empty
3640 subject address (from a bounce message) matched only if the pattern was an
3641 empty string. Non-empty patterns were not even tested. This was the wrong
3642 because it is perfectly reasonable to use an empty address as part of a
3643 database query. An empty address is now tested by patterns that are
3644 lookups. However, all the other forms of pattern expect the subject to
3645 contain a local part and a domain, and therefore, for them, an empty
3646 address still always fails if the pattern is not itself empty.
3648 30. Exim went into a mad DNS loop when attempting to do a callout where the
3649 host was specified on an smtp transport, and looking it up yielded more
3650 than one IP address.
3652 31. Re-factored the code for checking spool and log partition space into a
3653 function that finds that data and another that does the check. The former
3654 is then used to implement four new variables: $spool_space, $log_space,
3655 $spool_inodes, and $log_inodes.
3657 32. The RFC2047 encoding function was originally intended for short strings
3658 such as real names; it was not keeping to the 75-character limit for
3659 encoded words that the RFC imposes. It now respects the limit, and
3660 generates multiple encoded words if necessary. To be on the safe side, I
3661 have increased the buffer size for the ${rfc2047: expansion operator from
3664 33. It is now permitted to omit both strings after an "if" condition; if the
3665 condition is true, the result is "true". As before, when the second string
3666 is omitted, a false condition yields an empty string. This makes it less
3667 cumbersome to write custom ACL and router conditions.
3669 34. Failure to deliver a bounce message always caused it to be frozen, even if
3670 there was an errors_to setting on the router. The errors_to setting is now
3673 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3674 canonical form (fully expanded) before being placed in
3675 $sender_host_address.
3677 36. The table in the code that translates DNS record types into text (T_A to
3678 "A" for instance) was missing entries for NS and CNAME. It is just possible
3679 that this could have caused confusion if both these types were looked up
3680 for the same domain, because the text type is used as part of Exim's
3681 per-process caching. But the chance of anyone hitting this buglet seems
3684 37. The dnsdb lookup has been extended in a number of ways.
3686 (1) There is a new type, "zns", which walks up the domain tree until it
3687 finds some nameserver records. It should be used with care.
3689 (2) There is a new type, "mxh", which is like "mx" except that it returns
3690 just the host names, not the priorities.
3692 (3) It is now possible to give a list of domains (or IP addresses) to be
3693 looked up. The behaviour when one of the lookups defers can be
3694 controlled by a keyword.
3696 (4) It is now possible to specify the separator character for use when
3697 multiple records are returned.
3699 38. The dnslists ACL condition has been extended: it is now possible to supply
3700 a list of IP addresses and/or domains to be looked up in a particular DNS
3703 39. Added log_selector=+queue_time_overall.
3705 40. When running the queue in the test harness, wait just a tad after forking a
3706 delivery process, to get repeatability of debugging output.
3708 41. Include certificate and key file names in error message when GnuTLS fails
3709 to set them up, because the GnuTLS error message doesn't include the name
3710 of the failing file when there is a problem reading it.
3712 42. Allow both -bf and -bF in the same test run.
3714 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3716 44. The "Exiscan patch" is now merged into the mainline Exim source.
3718 45. Sometimes the final signoff response after QUIT could fail to get
3719 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3720 < 0 before doing a fflush(). This bug looks as though it goes back to the
3721 introduction of TLS in release 3.20, but "sometimes" must have been rare
3722 because the tests only now provoked it.
3724 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3725 (this can affect the format of dates).
3727 47. exim_tidydb, when checking for the continued existence of a message for
3728 which it has found a message-specific retry record, was not finding
3729 messages that were in split spool directories. Consequently, it was
3730 deleting retry records that should have stayed in existence.
3732 48. Steve fixed some bugs in eximstats.
3734 49. The SPA authentication driver was not abandoning authentication and moving
3735 on to the next authenticator when an expansion was forced to fail,
3736 contradicting the general specification for all authenticators. Instead it
3737 was generating a temporary error. It now behaves as specified.
3739 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3740 (the order specifies the preference for clients). The order is now AES256,
3741 AES128, 3DES, ARCFOUR128.
3743 51. Small patch to Sieve code - explicitly set From: when generating an
3746 52. Exim crashed if a remote delivery caused a very long error message to be
3747 recorded - for instance if somebody sent an entire SpamAssassin report back
3748 as a large number of 550 error lines. This bug was coincidentally fixed by
3749 increasing the size of one of Exim's internal buffers (big_buffer) that
3750 happened as part of the Exiscan merge. However, to be on the safe side, I
3751 have made the code more robust (and fixed the comments that describe what
3754 53. Now that there can be additional text after "Completed" in log lines (if
3755 the queue_time_overall log selector is set), a one-byte patch to exigrep
3756 was needed to allow it to recognize "Completed" as not the last thing in
3759 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3760 patch that reportedly fixes this has been added. I am not expert enough to
3761 create a test for it. This is what the patch creator wrote:
3763 "I found a little strange behaviour of ldap code when working with
3764 Windows 2003 AD Domain, where users was placed in more than one
3765 Organization Units. When I tried to give exim partial DN, the exit code
3766 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3767 But simultaneously result of request was absolutely normal ldap result,
3768 so I produce this patch..."
3770 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3771 REFERENCE, so I have modified the code to exclude the patch when that macro
3774 55. Some experimental protocols are using DNS PTR records for new purposes. The
3775 keys for these records are domain names, not reversed IP addresses. The
3776 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3777 leaves it alone. Component reversal etc. now happens only for IP addresses.
3780 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3782 57. Double the size of the debug message buffer (to 2048) so that more of very
3783 long debug lines gets shown.
3785 58. The exicyclog utility now does better if the number of log files to keep
3786 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3788 59. Two changes related to the smtp_active_hostname option:
3790 (1) $smtp_active_hostname is now available as a variable.
3791 (2) The default for smtp_banner uses $smtp_active_hostname instead
3792 of $primary_hostname.
3794 60. The host_aton() function is supposed to be passed a string that is known
3795 to be a valid IP address. However, in the case of IPv6 addresses, it was
3796 not checking this. This is a hostage to fortune. Exim now panics and dies
3797 if the condition is not met. A case was found where this could be provoked
3798 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3799 components; fortuitously, this particular loophole had already been fixed
3800 by change 4.50/55 above.
3802 If there are any other similar loopholes, the new check in host_aton()
3803 itself should stop them being exploited. The report I received stated that
3804 data on the command line could provoke the exploit when Exim was running as
3805 exim, but did not say which command line option was involved. All I could
3806 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3807 running as the user.
3810 61. There was a buffer overflow vulnerability in the SPA authentication code
3811 (which came originally from the Samba project). I have added a test to the
3812 spa_base64_to_bits() function which I hope fixes it.
3815 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3816 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3818 63. The daemon start-up calls getloadavg() while still root for those OS that
3819 need the first call to be done as root, but it missed one case: when
3820 deliver_queue_load_max is set with deliver_drop_privilege. This is
3821 necessary for the benefit of the queue runner, because there is no re-exec
3822 when deliver_drop_privilege is set.
3824 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3825 This has been fixed.
3827 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3828 was in use, was not putting the data itself into the right store pool;
3829 consequently, it could be overwritten for a subsequent message in the same
3830 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3833 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3835 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3836 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3837 the function were treating the return as a boolean value, which happened to
3838 work because 0=false and not-0=true, but is not correct code.
3840 68. The host_aton() function was not handling scoped IPv6 addresses (those
3841 with, for example, "%eth0" on the end) correctly.
3843 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3844 negated items (that is, ~something) in unsigned ints. Some compilers
3845 apparently mutter when there is no cast.
3847 70. If an address verification called from an ACL failed, and did not produce a
3848 user-specific message (i.e. there was only a "system" message), nothing was
3849 put in $acl_verify_message. In this situation, it now puts the system
3852 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3853 session; change 4.31/43 added the unwanted input to the log line - except
3854 that it did not do this in the start of session case. It now does.
3856 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3857 This is wrong and can cause the other end to generate a synchronization
3858 error if it is another Exim or anything else that does the synchronization
3859 check. A QUIT command is no longer sent after a timeout.
3861 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3862 during host lookups.
3864 74. The defer_ok option of callout verification was not working if it was used
3865 when verifying addresses in header lines, that is, for this case:
3867 verify = header_sender/callout=defer_ok
3869 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3870 those file descriptors could be used for SMTP connections. If anything
3871 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3872 could be sent to the SMTP client, causing chaos. The daemon now opens
3873 stdin, stdout, and stderr to /dev/null when it puts itself into the
3876 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3877 log by default. The user can override this with suitable Perl magic.
3879 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3880 the log message when discard triggers, was not working for the DATA ACL or
3881 for the non-SMTP ACL.
3883 78. Error message wording change in sieve.c.
3885 79. If smtp_accept_max_per_host was set, the number of connections could be
3886 restricted to fewer than expected, because the daemon was trying to set up
3887 a new connection before checking whether the processes handling previous
3888 connections had finished. The check for completed processes is now done
3889 earlier. On busy systems, this bug wouldn't be noticed because something
3890 else would have woken the daemon, and it would have reaped the completed
3893 80. If a message was submitted locally by a user whose login name contained one
3894 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3895 It caused a spool format error. I have fixed the spool reading code. A
3896 related problem was that the "from" clause in the Received: line became
3897 illegal because of the space(s). It is now covered by ${quote_local_part.
3899 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3902 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3904 83. Patch from Sieve maintainer to fix unterminated string problem in
3905 "vacation" handling.
3907 84. Some minor changes to the Linux configuration files to help with other
3908 OS variants using glibc.
3910 85. One more patch for Sieve to update vacation handling to latest spec.
3913 ----------------------------------------------------
3914 See the note above about the 4.44 and 4.50 releases.
3915 ----------------------------------------------------
3921 1. Change 4.43/35 introduced a bug that caused file counts to be
3922 incorrectly computed when quota_filecount was set in an appendfile
3925 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3926 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3929 3. Give more explanation in the error message when the command for a transport
3930 filter fails to execute.
3932 4. There are several places where Exim runs a non-Exim command in a
3933 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3934 was being done only for the command run by the queryprogram router. It is
3935 now done for all such subprocesses. The other cases are: ${run, transport
3936 filters, and the commands run by the lmtp and pipe transports.
3938 5. Some older OS have a limit of 256 on the maximum number of file
3939 descriptors. Exim was using setrlimit() to set 1000 as a large value
3940 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3941 systems. I've change it so that if it can't get 1000, it tries for 256.
3943 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3944 was an oversight, and furthermore, ever since the addition of extra
3945 controls (e.g. 4.43/32), the checks on when to allow different forms of
3946 "control" were broken. There should now be diagnostics for all cases when a
3947 control that does not make sense is encountered.
3949 7. $recipients is now available in the predata ACL (oversight).
3951 8. Tidy the search cache before the fork to do a delivery from a message
3952 received from the command line. Otherwise the child will trigger a lookup
3953 failure and thereby defer the delivery if it tries to use (for example) a
3954 cached ldap connection that the parent has called unbind on.
3956 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3957 of $address_data from the recipient verification was clobbered by the
3958 sender verification.
3960 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3961 was its contents. (It was OK if the option was not defined at all.)
3963 11. A "Completed" log line is now written for messages that are removed from
3964 the spool by the -Mrm option.
3966 12. $host_address is now set to the target address during the checking of
3967 ignore_target_hosts.
3969 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3970 being passed; this would have caused $sender_host_name to have been used if
3971 matching the list had actually called for a host name (not very likely,
3972 since this list is usually IP addresses). A host name is now passed as
3975 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3976 code that specifies a non-restarting handler (typically sigaction() in
3977 modern systems) in an attempt to fix a rare and obscure crash bug.
3979 15. Narrowed the window for a race in the daemon that could cause it to ignore
3980 SIGCHLD signals. This is not a major problem, because they are used only to
3981 wake it up if nothing else does.
3983 16. A malformed maildirsize file could cause Exim to calculate negative values
3984 for the mailbox size or file count. Odd effects could occur as a result.
3985 The maildirsize information is now recalculated if the size or filecount
3988 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3989 support for a long time. Removed HAVE_SYS_VFS_H.
3991 18. Updated exipick to current release from John Jetmore.
3993 19. Allow an empty sender to be matched against a lookup in an address list.
3994 Previously the only cases considered were a regular expression, or an
3997 20. Exim went into a mad DNS lookup loop when doing a callout where the
3998 host was specified on the transport, if the DNS lookup yielded more than
4001 21. The RFC2047 encoding function was originally intended for short strings
4002 such as real names; it was not keeping to the 75-character limit for
4003 encoded words that the RFC imposes. It now respects the limit, and
4004 generates multiple encoded words if necessary. To be on the safe side, I
4005 have increased the buffer size for the ${rfc2047: expansion operator from
4008 22. Failure to deliver a bounce message always caused it to be frozen, even if
4009 there was an errors_to setting on the router. The errors_to setting is now
4012 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4013 canonical form (fully expanded) before being placed in
4014 $sender_host_address.
4016 24. Updated eximstats to version 1.33
4018 25. Include certificate and key file names in error message when GnuTLS fails
4019 to set them up, because the GnuTLS error message doesn't include the name
4020 of the failing file when there is a problem reading it.
4022 26. Expand error message when OpenSSL has problems setting up cert/key files.
4025 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4026 (this can affect the format of dates).
4028 28. exim_tidydb, when checking for the continued existence of a message for
4029 which it has found a message-specific retry record, was not finding
4030 messages that were in split spool directories. Consequently, it was
4031 deleting retry records that should have stayed in existence.
4033 29. eximstats updated to version 1.35
4034 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4035 1.35 - bugfix such that pie charts by volume are generated correctly
4037 30. The SPA authentication driver was not abandoning authentication and moving
4038 on to the next authenticator when an expansion was forced to fail,
4039 contradicting the general specification for all authenticators. Instead it
4040 was generating a temporary error. It now behaves as specified.
4042 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4043 (the order specifies the preference for clients). The order is now AES256,
4044 AES128, 3DES, ARCFOUR128.
4046 31. Small patch to Sieve code - explicitly set From: when generating an
4049 32. Exim crashed if a remote delivery caused a very long error message to be
4050 recorded - for instance if somebody sent an entire SpamAssassin report back
4051 as a large number of 550 error lines. This bug was coincidentally fixed by
4052 increasing the size of one of Exim's internal buffers (big_buffer) that
4053 happened as part of the Exiscan merge. However, to be on the safe side, I
4054 have made the code more robust (and fixed the comments that describe what
4057 33. Some experimental protocols are using DNS PTR records for new purposes. The
4058 keys for these records are domain names, not reversed IP addresses. The
4059 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4060 leaves it alone. Component reversal etc. now happens only for IP addresses.
4063 34. The host_aton() function is supposed to be passed a string that is known
4064 to be a valid IP address. However, in the case of IPv6 addresses, it was
4065 not checking this. This is a hostage to fortune. Exim now panics and dies
4066 if the condition is not met. A case was found where this could be provoked
4067 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4068 components; fortuitously, this particular loophole had already been fixed
4069 by change 4.50/55 or 4.44/33 above.
4071 If there are any other similar loopholes, the new check in host_aton()
4072 itself should stop them being exploited. The report I received stated that
4073 data on the command line could provoke the exploit when Exim was running as
4074 exim, but did not say which command line option was involved. All I could
4075 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4076 running as the user.
4079 35. There was a buffer overflow vulnerability in the SPA authentication code
4080 (which came originally from the Samba project). I have added a test to the
4081 spa_base64_to_bits() function which I hope fixes it.
4084 36. The daemon start-up calls getloadavg() while still root for those OS that
4085 need the first call to be done as root, but it missed one case: when
4086 deliver_queue_load_max is set with deliver_drop_privilege. This is
4087 necessary for the benefit of the queue runner, because there is no re-exec
4088 when deliver_drop_privilege is set.
4090 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4091 was in use, was not putting the data itself into the right store pool;
4092 consequently, it could be overwritten for a subsequent message in the same
4093 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4096 38. Sometimes the final signoff response after QUIT could fail to get
4097 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4098 < 0 before doing a fflush(). This bug looks as though it goes back to the
4099 introduction of TLS in release 3.20, but "sometimes" must have been rare
4100 because the tests only now provoked it.
4106 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4107 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4108 Now it returns an integer. A number of calls were still expecting a T/F
4109 return. Fortuitously, in all cases, the tests worked in OK situations,
4110 which is the norm. However, things would have gone wrong on any write
4111 failures on the smtp file descriptor. This function is used when sending
4112 messages over SMTP and also when doing verify callouts.
4114 2. When Exim is called to do synchronous delivery of a locally submitted
4115 message (the -odf or -odi options), it no longer closes stderr before doing
4118 3. Implemented the mua_wrapper option.
4120 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4122 5. Implemented the functions header_remove(), header_testname(),
4123 header_add_at_position(), and receive_remove_recipient(), and exported them
4126 6. If an ACL "warn" statement specified the addition of headers, Exim already
4127 inserted X-ACL-Warn: at the start if there was no header name. However, it
4128 was not making this test for the second and subsequent header lines if
4129 there were newlines in the string. This meant that an invalid header could
4130 be inserted if Exim was badly configured.
4132 7. Allow an ACL "warn" statement to add header lines at the start or after all
4133 the Received: headers, as well as at the end.
4135 8. Added the rcpt_4xx retry error code.
4137 9. Added postmaster_mailfrom=xxx to callout verification option.
4139 10. Added mailfrom=xxxx to the callout verification option, for verify=
4142 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4143 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4145 12. Inserted some casts to stop certain compilers warning when using pointer
4146 differences as field lengths or precisions in printf-type calls (mostly
4147 affecting debugging statements).
4149 13. Added optional readline() support for -be (dynamically loaded).
4151 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4152 same clock tick as a message's arrival, so that its received time was the
4153 same as the "first fail" time on the retry record, and that message
4154 remained on the queue past the ultimate address timeout, every queue runner
4155 would try a delivery (because it was past the ultimate address timeout) but
4156 after another failure, the ultimate address timeout, which should have then
4157 bounced the address, did not kick in. This was a "< instead of <=" error;
4158 in most cases the first failure would have been in the next clock tick
4159 after the received time, and all would be well.
4161 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4162 being recognized when the domain list was tested by the match_domain
4163 condition in an expansion string.
4165 16. Added the ${str2b64: operator.
4167 17. Exim was always calling setrlimit() to set a large limit for the number of
4168 processes, without checking whether the existing limit was already
4169 adequate. (It did check for the limit on file descriptors.) Furthermore,
4170 errors from getrlimit() and setrlimit() were being ignored. Now they are
4171 logged to the main and panic logs, but Exim does carry on, to try to do its
4172 job under whatever limits there are.
4174 18. Imported PCRE 5.0.
4176 19. Trivial typo in log message " temporarily refused connection" (the leading
4179 20. If the log selector return_path_on_delivery was set and an address was
4180 redirected to /dev/null, the delivery process crashed because it assumed
4181 that a return path would always be set for a "successful" delivery. In this
4182 case, the whole delivery is bypassed as an optimization, and therefore no
4185 21. Internal re-arrangement: the function for sending a challenge and reading
4186 a response while authentication was assuming a zero-terminated challenge
4187 string. It's now changed to take a pointer and a length, to allow for
4188 binary data in such strings.
4190 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4192 23. Exim was not respecting finduser_retries when seeking the login of the
4193 uid under which it was called; it was always trying 10 times. (The default
4194 setting of finduser_retries is zero.) Also, it was sleeping after the final
4195 failure, which is pointless.
4197 24. Implemented tls_on_connect_ports.
4199 25. Implemented acl_smtp_predata.
4201 26. If the domain in control=submission is set empty, Exim assumes that the
4202 authenticated id is a complete email address when it generates From: or
4203 Sender: header lines.
4205 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4206 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4207 chown and chgrp in /bin and hostname in /usr/bin.
4209 28. Exim was keeping the "process log" file open after each use, just as it
4210 does for the main log. This opens the possibility of it remaining open for
4211 long periods when the USR1 signal hits a daemon. Occasional processlog
4212 errors were reported, that could have been caused by this. Anyway, it seems
4213 much more sensible not to leave this file open at all, so that is what now
4216 29. The long-running daemon process does not normally write to the log once it
4217 has entered its main loop, and it closes the log before doing so. This is
4218 so that log files can straightforwardly be renamed and moved. However,
4219 there are a couple of unusual error situations where the daemon does write
4220 log entries, and I had neglected to close the log afterwards.
4222 30. The text of an SMTP error response that was received during a remote
4223 delivery was being truncated at 512 bytes. This is too short for some of
4224 the long messages that one sometimes sees. I've increased the limit to
4227 31. It is now possible to make retry rules that apply only when a message has a
4228 specific sender, in particular, an empty sender.
4230 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4231 it possible to be selective about when SMTP synchronization is enforced.
4233 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4235 32. Implemented hosts_connection_nolog.
4237 33. Added an ACL for QUIT.
4239 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4242 35. Added mailbox_size and mailbox_filecount to appendfile.
4244 36. Added control = no_multiline_responses to ACLs.
4246 37. There was a bug in the logic of the code that waits for the clock to tick
4247 in the case where the clock went backwards by a substantial amount such
4248 that the microsecond fraction of "now" was more than the microsecond
4249 fraction of "then" (but the whole seconds number was less).
4251 38. Added support for the libradius Radius client library this is found on
4252 FreeBSD (previously only the radiusclient library was supported).
4258 1. When certain lookups returned multiple values in the form name=value, the
4259 quoting of the values was not always being done properly. Specifically:
4260 (a) If the value started with a double quote, but contained no whitespace,
4262 (b) If the value contained whitespace other than a space character (i.e.
4263 tabs or newlines or carriage returns) it was not quoted.
4264 This fix has been applied to the mysql and pgsql lookups by writing a
4265 separate quoting function and calling it from the lookup code. The fix
4266 should probably also be applied to nisplus, ibase and oracle lookups, but
4267 since I cannot test any of those, I have not disturbed their existing code.
4269 2. A hit in the callout cache for a specific address caused a log line with no
4270 reason for rejecting RCPT. Now it says "Previous (cached) callout
4271 verification failure".
4273 3. There was an off-by-one bug in the queryprogram router. An over-long
4274 return line was truncated at 256 instead of 255 characters, thereby
4275 overflowing its buffer with the terminating zero. As well as fixing this, I
4276 have increased the buffer size to 1024 (and made a note to document this).
4278 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4279 when Exim is waiting for an SMTP response from a remote server, Exim
4280 restarts its select() call on the socket, thereby resetting its timeout.
4281 This is not a problem when such interrupts are rare. Somebody set up a cron
4282 job to run exiwhat every 2 minutes, which is less than the normal select()
4283 timeout (5 or 10 minutes). This meant that the select() timeout never
4284 kicked in because it was always reset. I have fixed this by comparing the
4285 time when an interrupt arrives with the time at the start of the first call
4286 to select(). If more time than the timeout has elapsed, the interrupt is
4287 treated as a timeout.
4289 5. Some internal re-factoring in preparation for the addition of Sieve
4290 extensions (by MH). In particular, the "personal" test is moved to a
4291 separate function, and given an option for scanning Cc: and Bcc: (which is
4292 not set for Exim filters).
4294 6. When Exim created an email address using the login of the caller as the
4295 local part (e.g. when creating a From: or Sender: header line), it was not
4296 quoting the local part when it contained special characters such as @.
4298 7. Installed new OpenBSD configuration files.
4300 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4301 try to make them clearer.
4303 9. Callout options, other than the timeout value, were being ignored when
4304 verifying sender addresses in header lines. For example, when using
4306 verify = header_sender/callout=no_cache
4308 the cache was (incorrectly) being used.
4310 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4311 only the Cygwin environment.
4313 11. When return_path_on_delivery was set as a log selector, if different remote
4314 addresses in the same message used different return paths and parallel
4315 remote delivery occurred, the wrong values would sometimes be logged.
4316 (Whenever a remote delivery process finished, the return path value from
4317 the most recently started remote delivery process was logged.)
4319 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4320 lines when AUTH and/or TLS are in use. This is the "received protocol"
4321 field. Exim used to use "asmtp" for authenticated SMTP, without any
4322 indication (in the protocol name) for TLS use. Now it follows the RFC and
4323 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4324 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4325 names appear in log lines as well as in Received: header lines.
4327 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4328 extensions, and comparison tests, and to fix some bugs.
4330 14. Changes to the "personal" filter test:
4332 (1) The test was buggy in that it was just doing the equivalent of
4333 "contains" tests on header lines. For example, if a user's address was
4334 anne@some.where, the "personal" test would incorrectly be true for
4336 To: susanne@some.where
4338 This test is now done by extracting each address from the header in turn,
4339 and checking the entire address. Other tests that are part of "personal"
4340 are now done using regular expressions (for example, to check local parts
4341 of addresses in From: header lines).
4343 (2) The list of non-personal local parts in From: addresses has been
4344 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4345 taken from the Sieve specification recommendations.
4347 (3) If the message contains any header line starting with "List-" it is
4348 treated as non-personal.
4350 (4) The test for "circular" in the Subject: header line has been removed
4351 because it now seems ill-conceived.
4353 15. Minor typos in src/EDITME comments corrected.
4355 16. Installed latest exipick from John Jetmore.
4357 17. If headers_add on a router specified a text string that was too long for
4358 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4359 of string_sprintf() is now avoided.
4361 18. $message_body_size was not set (it was always zero) when running the DATA
4362 ACL and the local_scan() function.
4364 19. For the "mail" command in an Exim filter, no default was being set for
4365 the once_repeat time, causing a random time value to be used if "once" was
4366 specified. (If the value happened to be <= 0, no repeat happened.) The
4367 default is now 0s, meaning "never repeat". The "vacation" command was OK
4368 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4369 (I found it when inspecting the code).
4371 20. There is now an overall timeout for performing a callout verification. It
4372 defaults to 4 times the callout timeout, which applies to individual SMTP
4373 commands during the callout. The overall timeout applies when there is more
4374 than one host that can be tried. The timeout is checked before trying the
4375 next host. This prevents very long delays if there are a large number of
4376 hosts and all are timing out (e.g. when the network connections are timing
4377 out). The value of the overall timeout can be changed by specifying an
4378 additional sub-option for "callout", called "maxwait". For example:
4380 verify = sender/callout=5s,maxwait=20s
4382 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4383 to the end before writing, but this should make it even safer).
4385 22. Exim was forgetting that it had advertised PIPELINING for the second and
4386 subsequent messages on an SMTP connection. It was also not resetting its
4387 memory on STARTTLS and an internal HELO.
4389 23. When Exim logs an SMTP synchronization error within a session, it now
4390 records whether PIPELINING has been advertised or not.
4392 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4393 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4394 rather than long int.
4396 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4398 26. Added the never_mail option to autoreply.
4404 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4405 crash if the getsockname() call failed; this can happen if a connection is
4406 closed very soon after it is established. The problem was simply in the
4407 order in which certain operations were done, causing Exim to try to write
4408 to the SMTP stream before it had set up the file descriptor. The bug has
4409 been fixed by making things happen in the correct order.
4415 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4416 before the connection was closed, thus losing the rejection response.
4418 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4419 some early Solaris releases, but causes trouble in current releases where
4420 socklen_t is defined.
4422 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4425 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4428 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4429 that releasing the top of it at the end releases what was used for sub-
4430 expansions (unless the block got too big). However, discard this block if
4431 the first thing is a variable or header, so that we can use its block when
4432 it is dynamic (useful for very large $message_headers, for example).
4434 6. Lookups now cache *every* query, not just the most recent. A new, separate
4435 store pool is used for this. It can be recovered when all lookup caches are
4436 flushed. Lookups now release memory at the end of their result strings.
4437 This has involved some general refactoring of the lookup sources.
4439 7. Some code has been added to the store_xxx() functions to reduce the amount
4440 of flapping under certain conditions.
4442 8. log_incoming_interface used to affect only the <= reception log lines. Now
4443 it causes the local interface and port to be added to several more SMTP log
4444 lines, for example "SMTP connection from", and rejection lines.
4446 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4448 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4450 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4451 could be overwritten at the end of the current message (or the start of a
4452 new message if it was set in a HELO ACL). The value is now preserved for
4453 the duration of the SMTP connection.
4455 12. If a transport had a headers_rewrite setting, and a matching header line
4456 contained an unqualified address, that address was qualified, even if it
4457 did not match any rewriting rules. The underlying bug was that the values
4458 of the flags that permit the existence of unqualified sender and recipient
4459 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4460 non-local messages, and by -bnq for local messages) were not being
4461 preserved with the message after it was received.
4463 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4464 "next input=" as part of the text comprising the host identity instead of
4465 the correct text. The code was using the same buffer for two different
4466 strings. However, depending on which order the printing function evaluated
4467 its arguments, the bug did not always show up. Under Linux, for example, my
4468 test suite worked just fine.
4470 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4471 doesn't work with some older versions of Perl. It has been changed to "my",
4472 which in any case is probably the better facility to use.
4474 15. A really picky compiler found some instances of statements for creating
4475 error messages that either had too many or two few arguments for the format
4478 16. The size of the buffer for calls to the DNS resolver has been increased
4479 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4480 for addresses that have a lot of PTR records. This alleviates a problem; it
4481 does not fully solve it.
4483 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4484 buffer now truncates the list and logs the incident, which is the same
4485 action as happens when Exim is looking up a host name and its aliases.
4486 Previously in this situation something unpredictable would happen;
4487 sometimes it was "internal error: store_reset failed".
4489 18. If a server dropped the connection unexpectedly when an Exim client was
4490 using GnuTLS and trying to read a response, the client delivery process
4491 crashed while trying to generate an error log message.
4493 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4494 string, for example:
4496 warn message = H1: something\nH2: something
4498 the text was added as a single header line from Exim's point of view
4499 though it ended up OK in the delivered message. However, searching for the
4500 second and subsequent header lines using $h_h2: did not work. This has been
4501 fixed. Similarly, if a system filter added multiple headers in this way,
4502 the routers could not see them.
4504 20. Expanded the error message when iplsearch is called with an invalid key to
4505 suggest using net-iplsearch in a host list.
4507 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4508 ACLs are no longer actually imposed (and a message to that effect is
4511 22. If a "gecos" field in a passwd entry contained escaped characters, in
4512 particular, if it contained a \" sequence, Exim got it wrong when building
4513 a From: or a Sender: header from that name. A second bug also caused
4514 incorrect handling when an unquoted " was present following a character
4515 that needed quoting.
4517 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4518 was not being matched caselessly.
4520 24. Arranged for all hyphens in the exim.8 source to be escaped with
4523 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4524 properly. Recipient callout cache records were still being keyed to include
4525 the sender, even when use_sender was set false. This led to far more
4526 callouts that were necessary. The sender is no longer included in the key
4527 when use_sender is false.
4529 26. Added "control = submission" modifier to ACLs.
4531 27. Added the ${base62d: operator to decode base 62 numbers.
4533 28. dnsdb lookups can now access SRV records.
4535 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4536 the configuration file.
4538 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4539 (-v) mode. This makes the output for a verbose queue run more intelligible.
4541 31. Added a use_postmaster feature to recipient callouts.
4543 32. Added the $body_zerocount variable, containing the number of binary zero
4544 bytes in the message body.
4546 33. The time of last modification of the "new" subdirectory is now used as the
4547 "mailbox time last read" when there is a quota error for a maildir
4550 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4552 35. Added +ignore_unknown as a special item in host lists.
4554 36. Code for decoding IPv6 addresses in host lists is now included, even if
4555 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4556 address was recognized as an IP address, but was then not correctly decoded
4557 into binary, causing unexpected and incorrect effects when compared with
4564 1. Very minor rewording of debugging text in manualroute to say "list of
4565 hosts" instead of "hostlist".
4567 2. If verify=header_syntax was set, and a header line with an unqualified
4568 address (no domain) and a large number of spaces between the end of the
4569 name and the colon was received, the reception process suffered a buffer
4570 overflow, and (when I tested it) crashed. This was caused by some obsolete
4571 code that should have been removed. The fix is to remove it!
4573 3. When running in the test harness, delay a bit after writing a bounce
4574 message to get a bit more predictability in the log output.
4576 4. Added a call to search_tidyup() just before forking a reception process. In
4577 theory, someone could use a lookup in the expansion of smtp_accept_max_
4578 per_host which, without the tidyup, could leave open a database connection.
4580 5. Added the variables $recipient_data and $sender_data which get set from a
4581 lookup success in an ACL "recipients" or "senders" condition, or a router
4582 "senders" option, similar to $domain_data and $local_part_data.
4584 6. Moved the writing of debug_print from before to after the "senders" test
4587 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4588 problems for message scanning, either using a data ACL, or using
4589 local_scan() because the Received: header was not generated till after they
4590 were called (in order to set the time as the time of reception completion).
4591 I have revised the way this works. The header is now generated after the
4592 body is received, but before the ACL or local_scan() are called. After they
4593 are run, the timestamp in the header is updated.
4599 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4600 before starting a queue runner without re-exec. This happened only when
4601 deliver_drop_privilege was set or when the Exim user was set to root. The
4602 effect of the bug was that timeouts during subsequent deliveries caused
4603 crashes instead of being properly handled. The handler is now left at its
4604 default (and expected) setting.
4606 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4607 message, again when deliver_drop_privilege is set or Exim is run as root.
4608 The bug described in (1) was not present in this case, but the tidying up
4609 of the other signals was missing. I have made the two cases consistent.
4611 3. The ignore_target_hosts setting on a manualroute router was being ignored
4612 for hosts that were looked up using the /MX notation.
4614 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4617 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4618 operated on the sender address. After changing the $sender_address to <>
4619 for the sender address verify, Exim was re-instated it as the original
4620 (before rewriting) address, but remembering that it had rewritten it, so it
4621 wasn't rewriting it again. This bug also had the effect of breaking the
4622 sender address verification caching when the sender address was rewritten.
4624 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4625 This has been changed so that if the ip literal address matches
4626 ignore_target_hosts, the router declines.
4628 7. Added expansion conditions match_domain, match_address, and match_local_
4629 part (NOT match_host).
4631 8. The placeholder for the Received: header didn't have a length field set.
4633 9. Added code to Exim itself and to exim_lock to test for a specific race
4634 condition that could lead to file corruption when using MBX delivery. The
4635 issue is with the lockfile that is created in /tmp. If this file is removed
4636 after a process has opened it but before that process has acquired a lock,
4637 there is the potential for a second process to recreate the file and also
4638 acquire a lock. This could lead to two Exim processes writing to the file
4639 at the same time. The added code performs the same test as UW imapd; it
4640 checks after acquiring the lock that its file descriptor still refers to
4641 the same named file.
4643 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4644 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4647 11. Added the smtp_active_hostname option. If used, this will typically be made
4648 to depend on the incoming interface address. Because $interface_address is
4649 not set up until the daemon has forked a reception process, error responses
4650 that can happen earlier (such as "too many connections") no longer contain
4653 12. If an expansion in a condition on a "warn" statement fails because a lookup
4654 defers, the "warn" statement is abandoned, and the next ACL statement is
4655 processed. Previously this caused the whole ACL to be aborted.
4657 13. Added the iplsearch lookup type.
4659 14. Added ident_timeout as a log selector.
4661 15. Added tls_certificate_verified as a log selector.
4663 16. Added a global option tls_require_ciphers (compare the smtp transport
4664 option of the same name). This controls incoming TLS connections.
4666 17. I finally figured out how to make tls_require_ciphers do a similar thing
4667 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4668 before starting the TLS session.
4670 18. Tabs are now shown as \t in -bP output.
4672 19. If the log selector return_path_on_delivery was set, Exim crashed when
4673 bouncing a message because it had too many Received: header lines.
4675 20. If two routers both had headers_remove settings, and the first one included
4676 a superfluous trailing colon, the final name in the first list and the
4677 first name in the second list were incorrectly joined into one item (with a
4678 colon in the middle).
4684 1. Added -C and -D options to the exinext utility, mainly to make it easier
4685 to include in the automated testing, but these could be helpful when
4686 multiple configurations are in use.
4688 2. The exinext utility was not formatting the output nicely when there was
4689 an alternate port involved in the retry record key, nor when there was a
4690 message id as well (for retries that were specific to a specific message
4691 and a specific host). It was also confused by IPv6 addresses, because of
4692 the additional colons they contain. I have fixed the IPv4 problem, and
4693 patched it up to do a reasonable job for IPv6.
4695 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4696 delivery, the log line now contains "pipelined" if PIPELINING was used.
4698 4. An SMTP transport process used to panic and die if the bind() call to set
4699 an explicit outgoing interface failed. This has been changed; it is now
4700 treated in the same way as a connect() failure.
4702 5. A reference to $sender_host_name in the part of a conditional expansion
4703 that was being skipped was still causing a DNS lookup. This no longer
4706 6. The def: expansion condition was not recognizing references to header lines
4707 that used bh_ and bheader_.
4709 7. Added the _cache feature to named lists.
4711 8. The code for checking quota_filecount in the appendfile transport was
4712 allowing one more file than it should have been.
4714 9. For compatibility with Sendmail, the command line option
4722 and sets the incoming protocol and host name (for trusted callers). The
4723 host name and its colon can be omitted when only the protocol is to be set.
4724 Note the Exim already has two private options, -pd and -ps, that refer to
4725 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4726 "s", but I don't think that's a major issue.
4728 10. A number of refactoring changes to the code, none of which should affect
4731 (a) The number of logging options was getting close to filling up the
4732 32-bit word that was used as a bit map. I have split them into two classes:
4733 those that are passed in the argument to log_write(), and those that are
4734 only ever tested independently outside of that function. These are now in
4735 separate 32-bit words, so there is plenty of room for expansion again.
4736 There is no change in the user interface or the logging behaviour.
4738 (b) When building, for example, log lines, the code previously used a
4739 macro that called string_cat() twice, in order to add two strings. This is
4740 not really sufficiently general. Furthermore, there was one instance where
4741 it was actually wrong because one of the argument was used twice, and in
4742 one call a function was used. (As it happened, calling the function twice
4743 did not affect the overall behaviour.) The macro has been replaced by a
4744 function that can join an arbitrary number of extra strings onto a growing
4747 (c) The code for expansion conditions now uses a table and a binary chop
4748 instead of a serial search (which was left over from when there were very
4749 few conditions). Also, it now recognizes conditions like "pam" even when
4750 the relevant support is not compiled in: a suitably worded error message is
4751 given if an attempt is made to use such a condition.
4753 11. Added ${time_interval:xxxxx}.
4755 12. A bug was causing one of the ddress fields not to be passed back correctly
4756 from remote delivery subprocesses. The field in question was not being
4757 subsequently used, so this caused to problems in practice.
4759 13. Added new log selectors queue_time and deliver_time.
4761 14. Might have fixed a bug in maildirsizefile handling that threw up
4762 "unexpected character" debug warnings, and recalculated the data
4763 unnecessarily. In any case, I expanded the warning message to give more
4766 15. Added the message "Restricted characters in address" to the statements in
4767 the default ACL that block characters like @ and % in local parts.
4769 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4770 Three changes have been made:
4772 (a) There was a serious bug; a negative response to MAIL caused the whole
4773 recipient domain to be cached as invalid, thereby blocking all messages
4774 to all local parts at the same domain, from all senders. This bug has
4775 been fixed. The domain is no longer cached after a negative response to
4776 MAIL if the sender used is not empty.
4778 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4781 (c) A new callout option, "use_sender" has been added for people who want
4782 the modified behaviour.
4788 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4791 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4792 indeed breaks things for older releases.
4794 3. Added additional logging to the case where there is a problem reading data
4795 from a filter that is running in a subprocess using a pipe, in order to
4796 try to track down a specific problem.
4798 4. Testing facility fudge: when running in the test harness and attempting
4799 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4800 getting "No route to host". Convert this to a timeout.
4802 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4805 6. Some OS don't have socklen_t but use size_t instead. This affects the
4806 fifth argument of getsockopt() amongst other things. This is now
4807 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4808 can be set for individual OS. I have set it for SunOS5, OSF1, and
4809 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4810 some earlier ones do not.
4812 7. Change 4.30/15 was not doing the test caselessly.
4814 8. The standard form for an IPv6 address literal was being rejected by address
4815 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4816 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4817 this, as well as the form without the "IPv6" on the front (but only when
4818 address literals are enabled, of course).
4820 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4822 10. Exim crashed if a message with an empty sender address specified by -f
4823 encountered a router with an errors_to setting. This could be provoked only
4824 by a command such as
4828 where an empty string was supplied; "<>" did not hit this bug.
4830 11. Installed PCRE release 4.5.
4832 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4833 remained set. It is now erased.
4835 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4836 times from message ids (which are base 36 rather than the normal 62).
4838 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4839 were being counted as actual protocol errors, and logged if the log
4840 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4841 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4842 and DATA following a set of rejected RCPTs do not count as protocol errors.
4843 In other words, Exim assumes they were pipelined, though this may not
4844 actually be the case. Of course, in all cases the client gets an
4845 appropriate error code.
4847 15. If a lookup fails in an ACL condition, a message about the failure may
4848 be available; it is used if testing the ACL cannot continue, because most
4849 such messages specify what the cause of the deferral is. However, some
4850 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4851 that caused an old message to be retained and used if a later statement
4852 caused a defer, replacing the real cause of the deferral.
4854 16. If an IP address had so many PTR records that the DNS lookup buffer
4855 was not large enough to hold them, Exim could crash while trying to process
4856 the truncated data. It now detects and logs this case.
4858 17. Further to 4.21/58, another change has been made: if (and only if) the
4859 first line of a message (the first header line) ends with CRLF, a bare LF
4860 in a subsequent header line has a space inserted after it, so as not to
4861 terminate the header.
4863 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4864 unnecessarily, used atoi() instead of strtol(), and didn't check the
4865 termination when getting file sizes from file names by regex.
4867 19. Completely re-implemented the support for maildirsize files, in the light
4868 of a number of problems with the previous contributed implementation
4869 (4.30/29). In particular:
4871 . If the quota is zero, the maildirsize file is maintained, but no quota is
4874 . If the maildir directory does not exist, it is created before any attempt
4875 to write a maildirsize file.
4877 . The quota value in the file is just a cache; if the quota is changed in
4878 the transport, the new value overrides.
4880 . A regular expression is available for excluding directories from the
4883 20. The autoreply transport checks the characters in options that define the
4884 message's headers; it allows continued headers, but it was checking with
4885 isspace() after an embedded newline instead of explicitly looking for a
4888 21. If all the "regular" hosts to which an address was routed had passed their
4889 expiry times, and had not reached their retry times, the address was
4890 bounced, even if fallback hosts were defined. Now Exim should go on to try
4893 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4894 equivalent code in the SMTP transport. Some hosts send humungous responses
4895 to HELO/EHLO, more than 1024 it seems.
4897 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4898 gives compiler warnings in some environments. I've now done it "properly",
4901 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4902 (because of problems with the built-in one) was declared to return uschar *
4903 instead of char *, causing compiler failure.
4905 25. Fixed a file descriptor leak when processing alias/forward files.
4907 26. Fixed a minor format string issue in dbfn.c.
4909 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4911 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4912 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4915 29. When a sender address is verified, it is cached, to save repeating the test
4916 when there is more than one recipient in a message. However, when the
4917 verification involves a callout, it is possible for different callout
4918 options to be set for different recipients. It is too complicated to keep
4919 track of this in the cache, so now Exim always runs a verification when a
4920 callout is required, relying on the callout cache for the optimization.
4921 The overhead is duplication of the address routing, but this should not be
4924 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4925 to be verified with callout=postmaster, and the main callout worked but the
4926 postmaster check failed, the verification correctly failed. However, if a
4927 subsequent RCPT command asked for sender verification *without* the
4928 postmaster check, incorrect caching caused this verification also to fail,
4931 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4932 it was not caching the DNS options (qualify_single, search_parents) that
4933 were used when the lookup failed. A subsequent lookup with different
4934 options therefore always gave the same answer, though there were cases
4935 where it should not have. (Example: a "domains = !$mx_any" option on a
4936 dnslookup router: the "domains" option is always processed without any
4937 widening, but the router might have qualify_single set.) Now Exim uses the
4938 cached value only when the same options are set.
4940 32. Added John Jetmore's "exipick" utility to the distribution.
4942 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4943 than a timeout (e.g. a certificate is required, and is not provided), an
4944 Exim server now closes the connection immediately. Previously it waited for
4945 the client to close - but if the client is SSL, it seems that they each
4946 wait for each other, leading to a delay before one of them times out.
4948 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4949 maintained 0.8.x compatibility because I don't think many are using it, and
4950 it is clearly obsolete.
4952 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4955 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4956 client certificate was expired. A simple patch fixes this, though I don't
4957 understand the full logic of why the verify callback is called multiple
4960 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4961 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4962 which causes problems with some clients (such as the Certicom SSL Plus
4963 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4964 disables the coutermeasure allowing Eudora to connect."
4966 38. Exim was not checking that a write() to a log file succeeded. This could
4967 lead to Bad Things if a log got too big, in particular if it hit a file
4968 size limit. Exim now panics and dies if it cannot write to a log file, just
4969 as it does if it cannot open a log file.
4971 39. Modified OS/Makefile-Linux so that it now contains
4973 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4975 The two -D definitions ensure that Exim is compiled with large file
4976 support, which makes it possible to handle log files that are bigger than
4979 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4980 instance) a domain was checked against a named list that involved a lookup,
4981 causing $domain_data to be set, then another domain was checked against the
4982 same list, then the first domain was re-checked, the value of $domain_data
4983 after the final check could be wrong. In particular, if the second check
4984 failed, it could be set empty. This bug probably also applied to
4987 41. The strip_trailing_dot option was not being applied to the address given
4988 with the -f command-line option.
4990 42. The code for reading a message's header from the spool was incrementing
4991 $received_count, but never initializing it. This meant that the value was
4992 incorrect (doubled) while delivering a message in the same process in which
4993 it was received. In the most common configuration of Exim, this never
4994 happens - a fresh exec is done - but it can happen when
4995 deliver_drop_privilege is set.
4997 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4998 it now includes up to 150 characters of the unexpected data in the log
5001 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5002 and building data strings. The size of both of these buffers was 10 000
5003 bytes - far larger than anybody would *ever* want, thought I. Needless to
5004 say, somebody hit the limit. I have increased the maximum line length to
5005 20 000 and the maximum data length of concatenated lines to 100 000. I have
5006 also fixed two bugs, because there was no checking on these buffers. Tsk,
5007 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5008 buffer is too small.
5010 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5011 lsearch lookups. Now it does.
5013 46. When parsing a route_list item in a manualroute router, a fixed-length
5014 buffer was used for the list of hosts. I made this 1024 bytes long,
5015 thinking that nobody would ever have a list of hosts that long. Wrong.
5016 Somebody had a whole pile of complicated expansion conditions, and the
5017 string was silently truncated, leading to an expansion error. It turns out
5018 that it is easier to change to an unlimited length (owing to other changes
5019 that have happened since this code was originally written) than to build
5020 structure for giving a limitation error. The length of the item that
5021 expands into the list of hosts is now unlimited.
5023 47. The lsearch lookup could not handle data where the length of text line was
5024 more than 4095 characters. Such lines were truncated, leading to shortened
5025 data being returned. It should now handle lines of any length.
5027 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5028 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5031 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5032 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5033 no longer happens in comment lines. A list of replacements is now placed
5034 at the head of all of the source files, except those whose only change is
5035 to replace PERL_COMMAND in the very first #! line.
5037 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5038 messages on the queue, with a bottom-up merge sort, using code contributed
5039 by Michael Haardt. This should make operations like -bp somewhat faster on
5040 large queues. It won't affect queue runners, except when queue_run_in_order
5043 51. Installed eximstats 1.31 in the distribution.
5045 52. Added support for SRV lookups to the dnslookup router.
5047 53. If an ACL referred to $message_body or $message_body_end, the value was not
5048 reset for any messages that followed in the same SMTP session.
5050 54. The store-handling optimization for building very long strings was not
5051 differentiating between the different store pools. I don't think this
5052 actually made any difference in practice, but I've tidied it.
5054 55. While running the routers to verify a sender address, $sender_address
5055 was still set to the sender address. This is wrong, because when routing to
5056 send a bounce to the sender, it would be empty. Therefore, I have changed
5057 it so that, while verifying a sender address, $sender_address is set to <>.
5058 (There is no change to what happens when verifying a recipient address.)
5060 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5061 target A or AAAA records (if not already returned) without resetting the
5062 qualify_single or search_parents options of the DNS resolver. These are
5063 inappropriate in this case because the targets of MX and SRV records must
5064 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5065 target that, when qualified, matched something in the local domain. These
5066 two options are now turned off when doing these lookups.
5068 57. It seems that at least some releases of Reiserfs (which does not have the
5069 concept of a fixed number of inodes) returns zero and not -1 for the
5070 number of available inodes. This interacted badly with check_spool_inodes,
5071 which assumed that -1 was the "no such thing" setting. What I have done is
5072 to check that the total number of inodes is greater than zero before doing
5073 the test of how many are available.
5075 58. When a "warn" ACL statement has a log_message modifier, the message is
5076 remembered, and not repeated. This is to avoid a lot of repetition when a
5077 message has many recipients that cause the same warning to be written.
5078 Howewer, Exim was preserving the list of already written lines for an
5079 entire SMTP session, which doesn't seem right. The memory is now reset if a
5080 new message is started.
5082 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5083 debugging output unless log_rewrite was also set.
5085 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5086 of (int)(handle) when we know that handle contains (void *)(-1).
5088 61. The Exim daemon panic-logs an error return when it closes the incoming
5089 connection. However "connection reset by peer" seems to be common, and
5090 isn't really an error worthy of noting specially, so that particular error
5093 62. When Exim is trying to find all the local interfaces, it used to panic and
5094 die if the ioctl to get the interface flags failed. However, it seems that
5095 on at least one OS (Solaris 9) it is possible to have an interface that is
5096 included in the list of interfaces, but for which you get a failure error
5097 for this call. This happens when the interface is not "plumbed" into a
5098 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5099 failure of the "get flags" call assumes that the interface is down.
5101 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5102 makes life easier for people who are doing arithmetic on fields extracted
5103 from dates, where you often get leading zeros that should not be
5104 interpreted as octal.
5106 64. Added qualify_domain to the redirect router, to override the global
5109 65. If a pathologically long header line contained very many addresses (the
5110 report of this problem mentioned 10 000) and each of them was rewritten,
5111 Exim could use up a very large amount of memory. (It kept on making new
5112 copies of the header line as it rewrote, and never released the old ones.)
5113 At the expense of a bit more processing, the header rewriting function has
5114 been changed so that it no longer eats memory in this way.
5116 66. The generation of the Received: header has been moved from the time that a
5117 message starts to be received, to the time that it finishes. The timestamp
5118 in the Received: header should now be very close to that of the <= log
5119 line. There are two side-effects of this change:
5121 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5122 logged header lines no longer include the local Received: line, because
5123 it has not yet been created. The same applies to a copy of the message
5124 that is returned to a non-SMTP sender when a message is rejected.
5126 (b) When a filter file is tested using -bf, no additional Received: header
5127 is added to the test message. After some thought, I decided that this
5130 This change does not affect the value of $received_for. It is still set
5131 after address rewriting, but before local_scan() is called.
5133 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5135 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5136 gave an unhelpful panic error message, and a defer error. I have managed to
5137 change this behaviour so that it now rejects any supplied certificate,
5138 which seems right, as the list of acceptable certificates is empty.
5140 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5141 gave an unhelpful defer error. I have not managed to make this reject any
5142 supplied certificates, but the error message it gives is "no certificate
5143 supplied", which is not helpful.
5145 70. exigrep's output now also includes lines that are not associated with any
5146 message, but which match the given pattern. Implemented by a patch from
5147 Martin Sluka, which also tidied up the Perl a bit.
5149 71. Recipient callout verification, like sender verification, was using <> in
5150 the MAIL FROM command. This isn't really the right thing, since the actual
5151 sender may affect whether the remote host accepts the recipient or not. I
5152 have changed it to use the actual sender in the callout; this means that
5153 the cache record is now keyed on a recipient/sender pair, not just the
5154 recipient address. There doesn't seem to be a real danger of callout loops,
5155 since a callout by the remote host to check the sender would use <>.
5156 [SEE ABOVE: changed after hitting problems.]
5158 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5159 temporary errors. However, in the case of such a code being given after
5160 the end of a data transmission (i.e. after ".") Exim was failing to write
5161 a retry record for the message. (Yes, there was some broken host that was
5162 actually sending 8xx at this point.)
5164 73. An unknown lookup type in a host list could cause Exim to panic-die when
5165 the list was checked. (An example that provoked this was putting <; in the
5166 middle of a list instead of at the start.) If this happened during a DATA
5167 ACL check, a -D file could be left lying around. This kind of configuration
5168 error no longer causes Exim to die; instead it causes a defer errror. The
5169 incident is still logged to the main and panic logs.
5171 74. Buglet left over from Exim 3 conversion. The message "too many messages
5172 in one connection" was written to the rejectlog but not the mainlog, except
5173 when address rewriting (yes!) was being logged.
5175 75. Added write_rejectlog option.
5177 76. When a system filter was run not as root (that is, when system_filter_user
5178 was set), the values of the $n variables were not being returned to the
5179 main process; thus, they were not subsequently available in the $sn
5182 77. Added +return_path_on_delivery log selector.
5184 78. A connection timeout was being treated differently from recipients deferred
5185 when testing hosts_max_try with a message that was older than the host's
5186 retry timeout. (The host should not be counted, thus allowing all hosts to
5187 be tried at least once before bouncing.) This may have been the cause of an
5188 occasionally reported bug whereby a message would remain on the queue
5189 longer than the retry timeout, but would be bounced if a delivery was
5190 forced. I say "may" because I never totally pinned down the problem;
5191 setting up timeout/retry tests is difficult. See also the next item.
5193 79. The ultimate address timeout was not being applied to errors that involved
5194 a combination of host plus message (for example, a timeout on a MAIL
5195 command). When an address resolved to a number of possible hosts, and they
5196 were not all tried for each delivery (e.g. because of hosts_max_try), a
5197 message could remain on the queue longer than the retry timeout.
5199 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5202 81. Fixed an obscure SMTP outgoing bug which required at least the following
5203 conditions: (a) there was another message waiting for the same server;
5204 (b) the server returned 5xx to all RCPT commands in the first message so
5205 that the message was not completed; (c) the server dropped the connection
5206 or gave a negative response to the RSET that Exim sends to abort the
5207 transaction. The observed case was a dropped connection after DATA that had
5208 been sent in pipelining mode. That is, the server had advertised PIPELINING
5209 but was not implementing it correctly. The effect of the bug was incorrect
5210 behaviour, such as trying another host, and this could lead to a crash.
5216 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5217 and daemon.c were passed as pointers to ints; they should have been
5218 pointers to socklen_t variables (which are typically unsigned ints).
5220 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5223 3. Fixed a really odd bug that affected only the testing scheme; patching a
5224 certain fixed string in the binary changed the value of another string that
5225 happened to be identical to the end of the original first string.
5227 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5228 name", it returns that address as the IP address. On some operating
5229 systems (e.g. Solaris), it also passes back the IP address string as the
5230 "host name". However, on others (e.g. Linux), it passes back an empty
5231 string. Exim wasn't checking for this, and was changing the host name to an
5232 empty string, assuming it had been canonicized.
5234 5. Although rare, it is permitted to have more than one PTR record for a given
5235 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5236 all the names associated with an address, because they do in Solaris.
5237 However, it seems that they do not in Linux for data that comes from the
5238 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5239 I found this out when I moved to a new Linux workstation and tried to run
5240 the Exim test suite.
5242 To get round this problem I have changed the code so that it now does its
5243 own call to the DNS to look up PTR records when searching for a host name.
5244 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5245 addresses that are only in /etc/hosts are still found.
5247 This behaviour is, however, controlled by an option called host_lookup_
5248 order, which defaults to "bydns:byaddr". If people want to use the other
5249 order, or indeed, just use one or the other means of lookup, they can
5250 specify it in this variable.
5252 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5253 some operating systems, this comes back from gethostbyaddr() as an empty
5254 string, and this is what Exim used to test for. However, it seems that in
5255 other systems, "." is yielded. Exim now tests for this case too.
5257 7. The values of check_spool_space and check_log_space are now held internally
5258 as a number of kilobytes instead of an absolute number of bytes. If a
5259 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5260 kilobyte. This means that much larger values can be stored.
5262 8. Exim monitor: an attempt to get the action menu when not actually pointing
5263 at a message produces an empty menu entitled "No message selected". This
5264 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5265 no entries in it ("Shell widget menu has zero width and/or height"). So I
5266 have added a single, blank menu entry in this case.
5268 9. Added ${quote_local_part.
5270 10. MIME decoding is now applied to the contents of Subject: header lines when
5273 11. Now that a reference to $sender_host_address automatically causes a reverse
5274 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5275 host lookup before query-style lookups in lists that might use this
5276 variable. This has therefore been abolished, and the "net-" prefix is no
5277 longer necessary for query-style lookups.
5279 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5280 have been a typo for LFLAGS, so it has been changed.
5282 13. The install script calls Exim with "-C /dev/null" in order to find the
5283 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5284 to be output. Howeve, since Exim outputs its version number before the
5285 error, it didn't break the script. It just looked ugly. I fixed this by
5286 always allowing "-C /dev/null" if the caller is root.
5288 14. Ignore overlarge ACL variable number when reading spool file - insurance
5289 against a later release with more variables having written the file.
5291 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5292 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5293 this, as well as the form without the "IPv6" on the front.
5295 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5296 OS/Makefile-Darwin file.
5298 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5299 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5301 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5302 variable. However, some OS use ints for inodes. I've added cast to long int
5303 to get rid of the compiler warning.
5305 19. I had forgotten to lock out "/../" in configuration file names when
5306 ALT_CONFIG_PREFIX was set.
5308 20. Routers used for verification do not need to specify transports. However,
5309 if such a router generated a host list, and callout was configured, Exim
5310 crashed, because it could not find a port number from the (non-existent)
5311 transport. It now assumes port 25 in this circumstance.
5313 21. Added the -t option to exigrep.
5315 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5316 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5317 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5318 them is set, LOOKUP_LSEARCH is forced.
5320 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5322 24. Added sender and host information to the "rejected by local_scan()" log
5323 line; previously there was no indication of these.
5325 25. Added .include_if_exists.
5327 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5328 Linux. It turns out that not all file systems support this. Apparently some
5329 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5330 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5331 not supported on the file descriptor, is now ignored when Exim is trying to
5332 sync a directory. This applies only to Linux.
5334 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5336 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5337 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5340 29. Added support for maildirsize files from supplied patch (modified a bit).
5342 30. The use of :fail: followed by an empty string could lead Exim to respond to
5343 sender verification failures with (e.g.):
5345 550 Verification failed for <xxx>
5346 550 Sender verify failed
5348 where the first response line was missing the '-' that indicates it is not
5349 the final line of the response.
5351 31. The loop for finding the name of the user that called Exim had a hardwired
5352 limit of 10; it now uses the value of finduser_retries, which is used for
5353 all other user lookups.
5355 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5358 33. Exim was neglecting to zero errno before one call of strtol() when
5359 expanding a string and expecting an integer value. On some systems this
5360 resulted in spurious "integer overflow" errors. Also, it was casting the
5361 result into an int without checking.
5363 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5364 did not work. The code looks as if it has *never* worked, though it appears
5365 to have been documented since at least releast 1.62. I have made it work.
5367 35. The "timeout_DNS" error in retry rules, also documented since at least
5368 1.62, also never worked. As it isn't clear exactly what this means, and
5369 clearly it isn't a major issue, I have abolished the feature by treating it
5370 as "timeout", and writing a warning to the main and panic logs.
5372 36. The display of retry rules for -brt wasn't always showing the error code
5375 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5376 timeout_connect_A, timeout_connect_MX.
5378 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5379 to the empty sender.
5381 39. The daemon was not analysing the content of -oX till after it had closed
5382 stderr and disconnected from the controlling terminal. This meant that any
5383 syntax errors were only noted on the panic log, and the return code from
5384 the command was 0. By re-arranging the code a little, I've made the
5385 decoding happen first, so such errors now appear on stderr, and the return
5386 code is 1. However, the actual setting up of the sockets still happens in
5387 the disconnected process, so errors there are still only recorded on the
5390 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5391 connections (as happens on some IP stacks) was logged at start up time as
5392 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5393 it from "IPv6 and IPv4", which means that two separate sockets are being
5396 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5397 says whether AF_INET or AF_INET6 was passed as an argument.
5399 42. Exiwhat output was messed up when time zones were included in log
5402 43. Exiwhat now gives more information about the daemon's listening ports,
5403 and whether -tls-on-connect was used.
5405 44. The "port" option of the smtp transport is now expanded.
5407 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5408 silently ignored. Now an error message is written to the main and panic
5411 46. There's a new ACL modifier called "logwrite" which writes to a log file
5412 as soon as it is encountered.
5414 47. Added $local_user_uid and $local_user_gid at routing time.
5416 48. Exim crashed when trying to verify a sender address that was being
5419 49. Exim was recognizing only a space character after ".include". It now also
5420 recognizes a tab character.
5422 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5423 extracting the relevant information from the specification. The man page no
5424 longer contains scrambled data for the -d option, and I've added a section
5425 at the front about calling Exim under different names.
5427 51. Added "extra_headers" argument to the "mail" command in filter files.
5429 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5432 53. Installed eximstats 1.29.
5434 54. Added transport_filter_timeout as a generic transport option.
5436 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5437 Cc: header lines. This was required by RFC 822, but it not required by RFC
5440 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5441 incoming messages that did not have them. Now it does so only if the
5442 message originates locally, that is, if there is no associated remote host
5443 address. When Resent- header lines are present, this applies to the Resent-
5444 lines rather than the non-Resent- lines.
5446 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5447 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5449 58. Messages for configuration errors now include the name of the main
5450 configuration file - useful now that there may be more than one file in a
5451 list (.included file names were always shown).
5453 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5454 for those rare installations that do not start the daemon as root or run it
5455 setuid root. I've cut out the call to initgroups() if the daemon is not
5458 60. The Exim user and group can now be bound into the binary as text strings
5459 that are looked up at the start of Exim's processing.
5461 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5463 62. Added $mailstore_basename variable.
5465 63. Installed patch to sieve.c from Michael Haardt.
5467 64. When Exim failed to open the panic log after failing to open the main log,
5468 the original message it was trying to log was written to stderr and debug
5469 output, but if they were not available (the usual case in production), it
5470 was lost. Now it is written to syslog before the two lines that record the
5471 failures to open the logs.
5473 65. Users' Exim filters run in subprocesses under the user's uid. It is
5474 possible for a "deliver" command or an alias in a "personal" command to
5475 provoke an address rewrite. If logging of address rewriting is configured,
5476 this fails because the process is not running as root or exim. There may be
5477 a better way of dealing with this, but for the moment (because 4.30 needs
5478 to be released), I have disabled address rewrite logging when running a
5479 filter in a non-root, non-exim process.
5485 1. The buildconfig auxiliary program wasn't quoting the value set for
5486 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5487 not defined. This bug was masked in 4.22 by the effect that was fixed in
5490 2. Some messages that were rejected after a message id was allocated were
5491 shown as "incomplete" by exigrep. It no longer does this for messages that
5492 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5494 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5495 have allow_domain_literals set, the ID did not get logged in the <= line.
5496 Domain literals are now always recognized in Message-ID: header lines.
5498 4. The first argument for a ${extract expansion item is the key name or field
5499 number. Leading and trailing spaces in this item were not being ignored,
5500 causing some misleading effects.
5502 5. When deliver_drop_privilege was set, single queue runner processes started
5503 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5504 same command in the process it spins off) were not dropping privilege.
5506 6. When the daemon running as "exim" started a queue runner, it always
5507 re-executed Exim in the spun-off process. This is a waste of effort when
5508 deliver_drop_privilege is set. The new process now just calls the
5509 queue-runner function directly.
5515 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5518 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5519 ignored. Though the use of -oP was forcing the writing of a pid file, it
5520 was always written to the default place.
5522 3. If the message "no IP address found for host xxxx" is generated during
5523 incoming verification, it is now followed by identification of the incoming
5524 connection (so you can more easily find what provoked it).
5526 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5528 5. Added some features to "harden" Exim a bit more against certain attacks:
5530 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5531 be put in Local/Makefile. This is like the never_users runtime option,
5532 but it cannot be overridden. The default setting is "root".
5534 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5535 prefix string with which any file named in a -C command line option
5538 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5539 is retained for -C and -D only if the caller of Exim is root. Without
5540 it, the exim user may also use -C and -D and retain privilege.
5542 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5543 command line option is disabled.
5545 6. Macro names set by the -D option must start with an upper case letter, just
5546 like macro names defined in the configuration file.
5548 7. Added "dereference=" facility to LDAP.
5550 8. Two instances of the typo "uknown" in the source files are fixed.
5552 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5553 the Configure-Makefile script screwed up while processing it.
5555 10. Incorporated PCRE 4.4.
5557 11. The SMTP synchronization check was not operating right at the start of an
5558 SMTP session. For example, it could not catch a HELO sent before the client
5559 waited for the greeting. There is now a check for outstanding input at the
5560 point when the greeting is written. Because of the duplex, asynchronous
5561 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5562 way, but not yet received, when the check is performed.
5564 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5565 on TCP/IP sockets, because this apparently causes some broken clients to
5568 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5569 unchanged) from the Cygwin maintainer.
5571 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5572 was supported instead of testing for mbx. Effectively a typo.
5574 15. The spa authenticator server code was not checking that the input it
5575 received was valid base64.
5577 16. The debug output line for the "set" modifier in ACLs was not showing the
5578 name of the variable that was being set.
5580 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5582 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5583 lookup on its own. Something else was needed to trigger the lookup. For
5584 example, a match in host_lookup or the need for a host name in a host list.
5585 Now, if $sender_host_name is referenced and the host name has not yet been
5586 looked up, a lookup is performed. If the lookup fails, the variable remains
5587 empty, and $host_lookup_failed is set to "1".
5589 19. Added "eqi" as a case-independent comparison operator.
5591 20. The saslauthd authentication condition could segfault if neither service
5592 nor realm was specified.
5594 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5595 error message that was logged was misleading, and incoming SMTP
5596 connections were dropped. The message is now more accurate, and temporary
5597 errors are given to SMTP connections.
5599 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5600 (except RSET) with a 421 error, until QUIT is received. However, it was
5601 failing to send a response to QUIT.
5603 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5604 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5605 ACL is now run after the helo_try_verify_hosts code.
5607 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5608 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5609 case-independent, so other case variants are also recognized). Apparently
5610 some systems use these upper case variants.
5612 25. If more than two messages were waiting for the same host, and a transport
5613 filter was specified for the transport, Exim sent two messages over the
5614 same TCP/IP connection, and then failed with "socket operation on non-
5615 socket" when it tried to send the third.
5617 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5619 27. The extern definition of crypt16() in expand.c was not being excluded when
5620 the OS had its own crypt16() function.
5622 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5623 as a preferred synonym for return_size_limit, both as an option and as an
5626 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5628 30. Changed the default configuration ACL to relax the local part checking rule
5629 for addresses that are not in any local domains. For these addresses,
5630 slashes and pipe symbols are allowed within local parts, but the sequence
5631 /../ is explicitly forbidden.
5633 31. SPA server authentication was not clearing the challenge buffer before
5636 32. log_message in a "warn" ACL statement was writing to the reject log as
5637 well as to the main log, which contradicts the documentation and doesn't
5638 seem right (because no rejection is happening). So I have stopped it.
5640 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5641 However, I am unable to do any testing of this.
5643 34. Fixed an infelicity in the appendfile transport. When checking directories
5644 for a mailbox, to see if any needed to be created, it was accidentally
5645 using path names with one or more superfluous leading slashes; tracing
5646 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5648 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5649 contents are added to the log line that is written for every discarded
5650 recipient. (Previously a log_message setting was ignored.)
5652 36. The ${quote: operator now quotes the string if it is empty.
5654 37. The install script runs exim in order to find its version number. If for
5655 some reason other than non-existence or emptiness, which it checks, it
5656 could not run './exim', it was installing it with an empty version number,
5657 i.e. as "exim-". This error state is now caught, and the installation is
5660 38. An argument was missing from the function that creates an error message
5661 when Exim fails to connect to the socket for saslauthd authentication.
5662 This could cause Exim to crash, or give a corrupted message.
5664 39. Added isip, isip4, and isip6 to ${if conditions.
5666 40. The ACL variables $acl_xx are now saved with the message, and can be
5667 accessed later in routers, transports, and filters.
5669 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5670 strings in the file are not string-expanded.
5672 42. If a MAIL command specified a SIZE value that was too large to fit into an
5673 int variable, the check against message_size_limit failed. Such values are
5674 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5675 day this will have to be increased, but I don't think I want to be around
5676 when emails are that large.
5683 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5684 iconv() is not standard in FreeBSD.
5686 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5687 IPv6 enabled. The observed symptom was a segmentation fault on return from
5688 the function os_common_find_running_interfaces() in src/os.c.
5690 3. In the check_special_case() function in daemon.c I had used "errno" as an
5691 argument name, which causes warnings on some systems. This was basically a
5692 typo, since it was named "eno" in the comments!
5694 4. The code that waits for the clock to tick (at a resolution of some fraction
5695 of a second) so as to ensure message-id uniqueness was always waiting for
5696 at least one whole tick, when it could have waited for less. [This is
5697 almost certainly not relevant at current processor speeds, where it is
5698 unlikely to ever wait at all. But we try to future-proof.]
5700 5. The function that sleeps for a time interval that includes fractions of a
5701 second contained a race. It did not block SIGALRM between setting the
5702 timer, and suspending (a couple of lines later). If the interval was short
5703 and the sigsuspend() was delayed until after it had expired, the suspension
5704 never ended. On busy systems this could lead to processes getting stuck for
5707 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5708 process, before it forks any delivery processes. The open lookup caching
5709 mechanism meant that the open file or database connection was passed into
5710 the delivery process. The problem was that delivery processes always tidy
5711 up cached lookup data. This could cause a problem for the next delivery
5712 process started by the queue runner, because the external queue runner
5713 process does not know about the closure. So the next delivery process
5714 still has data in the lookup cache. In the case of a file lookup, there was
5715 no problem because closing a file descriptor in a subprocess doesn't affect
5716 the parent. However, if the lookup was caching a connection to a database,
5717 the connection was closed, and the second delivery process was likely to
5718 see errors such as "PGSQL: query failed: server closed the connection
5719 unexpectedly". The problem has been fixed by closing all cached lookups
5720 in a queue runner before running a delivery process.
5722 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5723 seem to have the "const" qualifier which it has on other OS. I've
5726 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5727 *of the same type* (client or server) with the same public name that an
5728 error should be diagnosed.
5730 9. When Exim looked up a host name for an IP address, but failed to find the
5731 original IP address when looking up the host name (a safety check), it
5732 output the message "<ip address> does not match any IP for NULL", which was
5733 confusing, to say the least. The bug was that the host name should have
5734 appeared instead of "NULL".
5736 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5737 user that is built into the binary, and the -C or -D options is used, root
5738 privilege is dropped before the configuration file is read. In addition,
5739 logging is switched to stderr instead of the normal log files. If the
5740 configuration then re-defines the Exim user, the unprivileged environment
5741 is probably not what is expected, so Exim logs a panic warning message (but
5744 However, if deliver_drop_privilege is set, the unprivileged state may well
5745 be exactly what is intended, so the warning has been cut out in that case,
5746 and Exim is allowed to try to write to its normal log files.
5752 1. smtp_return_error_details was not giving details for temporary sender
5753 or receiver verification errors.
5755 2. Diagnose a configuration error if two authenticators have the same public
5758 3. Exim used not to create the message log file for a message until the first
5759 delivery attempt. This could be confusing when incoming messages were held
5760 for policy or load reasons. The message log file is now created at the time
5761 the message is received, and an initial "Received" line is written to it.
5763 4. The automatically generated man page for command line options had a minor
5764 bug that caused no ill effects; however, a more serious problem was that
5765 the procedure for building the man page automatically didn't always
5766 operate. Consequently, release 4.20 contains an out-of-date version. This
5767 shouldn't happen again.
5769 5. When building Exim with embedded Perl support, the script that builds the
5770 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5771 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5773 6. The freeze_tell option was not being used for messages that were frozen on
5774 arrival, either by an ACL or by local_scan().
5776 7. Added the smtp_incomplete_transaction log selector.
5778 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5779 was accepting AUTH without a new EHLO.
5781 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5782 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5785 10. Logging of TCP/IP connections (when configured) now happens in the main
5786 daemon process instead of the child process, so that the TCP/IP connection
5787 count is more accurate (but it can never be perfect).
5789 11. The use of "drop" in a nested ACL was not being handled correctly in the
5790 outer ACL. Now, if condition failure induced by the nested "drop" causes
5791 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5792 or "require"), the connection is dropped.
5794 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5795 that yield "discard" can now be used with an "accept" or a "discard" verb,
5796 but an error is generated for any others (because I can't see a useful way
5797 to define what should happen).
5799 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5800 are now processed in the same way as lines in the Exim configuration file.
5801 In particular, continuation lines are supported.
5803 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5805 15. Added -ti meaning -t -i.
5807 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5808 domains, and warn by logging if others are found.
5810 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5811 returned from GIFCONF when Exim is trying to find the list of interfaces on
5812 a host. The code in os.c has been modified to copy each ifreq to an aligned
5813 structure in all cases.
5815 Also, in some cases, the returned ifreq's were being copied to a 'struct
5816 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5817 means the last couple of bytes of an IPv6 address could be chopped if the
5818 ifreq contained only a normal sockaddr (14 bytes storage).
5820 18. Named domain lists were not supported in the hosts_treat_as_local option.
5821 An entry such as +xxxx was not recognized, and was treated as a literal
5824 19. Ensure that header lines added by a DATA ACL are included in the reject log
5825 if the ACL subsequently rejects the message.
5827 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5828 MD5 (which is deprecated).
5830 21. When testing a filter file using -bf, Exim was writing a message when it
5831 took the sender from a "From " line in the message, but it was not doing so
5832 when it took $return_path from a Return-Path: header line. It now does.
5834 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5835 with a valid header line field name (a series of printing characters
5836 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5838 23. Changed "disc" in the source to "disk" to conform to the documentation and
5839 the book and for uniformity.
5841 24. Ignore Sendmail's -Ooption=value command line item.
5843 25. When execve() failed while trying to run a command in a pipe transport,
5844 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5845 could be confused with a return value of 69 from the command itself. This
5846 has been changed to 127, the value the shell returns if it is asked to run
5847 a non-existent command. The wording for the related log line suggests a
5848 non-existent command as the problem.
5850 26. If received_header_text expands to an empty string, do not add a Received:
5851 header line to the message. (Well, it adds a token one on the spool, but
5852 marks it "old" so that it doesn't get used or transmitted.)
5854 27. Installed eximstats 1.28 (addition of -nt option).
5856 28. There was no check for failure on the call to getsockname() in the daemon
5857 code. This can fail if there is a shortage of resources on the system, with
5858 ENOMEM, for example. A temporary error is now given on failure.
5860 29. Contrary to the C standard, it seems that in some environments, the
5861 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5862 program. Exim now does this explicitly; it affects the formatting of
5863 timestamps using strftime().
5865 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5866 complaints. I've now initialized all the variables, to avoid this.
5868 32. Header lines added by a system filter were not being "seen" during
5869 transport-time rewrites.
5871 33. The info_callback() function passed to OpenSSL is set up with type void
5872 (*)(SSL *, int, int), as described somewhere. However, when calling the
5873 function (actually a macro) that sets it up, the type void(*)() is
5874 expected. I've put in a cast to prevent warnings from picky compilers.
5876 34. If a DNS black list lookup found a CNAME record, but there were no A
5877 records associated with the domain it pointed at, Exim crashed.
5879 35. If a DNS black list lookup returned more than one A record, Exim ignored
5880 all but the first. It now scans all returned addresses if a particular IP
5881 value is being sought. In this situation, the contents of the
5882 $dnslist_value variable are a list of all the addresses, separated by a
5885 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5886 to accept a host name and all its aliases if the forward lookup for any of
5887 them yielded the IP address of the incoming connection. Now it accepts only
5888 those names whose forward lookup yields the correct IP address. Any other
5889 names are discarded. This closes a loophole whereby a rogue DNS
5890 administrator could create reverse DNS records to break through a
5891 wildcarded host restriction in an ACL.
5893 37. If a user filter or a system filter that ran in a subprocess used any of
5894 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5895 the wrong values were passed to the pipe command ($thisaddress had the
5896 value of $0, $0 had the value of $1, etc). This bug was introduced by
5897 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5899 38. Improved the line breaking for long SMTP error messages from ACLs.
5900 Previously, if there was no break point between 40 and 75 characters, Exim
5901 left the rest of the message alone. Two changes have been made: (a) I've
5902 reduced the minimum length to 35 characters; (b) if it can't find a break
5903 point between 35 and 75 characters, it looks ahead and uses the first one
5904 that it finds. This may give the occasional overlong line, but at least the
5905 remaining text gets split now.
5907 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5908 file descriptors might be low, and that setting 1000 would always raise it.
5909 It turns out that in some environments, the limit is already over 1000 and
5910 that lowering it causes trouble. So now Exim takes care not to decrease it.
5912 40. When delivering a message, the value of $return_path is set to $sender_
5913 address at the start of routing (routers may change the value). By an
5914 oversight, this default was not being set up when an address was tested by
5915 -bt or -bv, which affected the outcome if any router or filter referred to
5918 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5919 log only when "message" or "log_message" are set. However, if one of the
5920 conditions was an address verification, or a call to a nested ACL, the
5921 messages generated by the underlying test were being passed through. This
5922 no longer happens. The underlying message is available in $acl_verify_
5923 message for both "message" and "log_message" expansions, so it can be
5924 passed through if needed.
5926 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5927 new expansion $bh_ to give the encoded byte string without charset
5928 translation. Translation happens only if iconv() is available; HAVE_ICONV
5929 indicates this at build time. HEADERS_CHARSET gives the charset to
5930 translate to; headers_charset can change it in the configuration, and
5931 "headers charset" can change it in an individual filter file.
5933 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5934 that creates RFC 2047 encoded "words" labels them as that charset instead
5935 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5936 expansion operator; (ii) when Exim creates a From: line for a local
5937 message; (iii) when a header line is rewritten to include a "phrase" part.
5939 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5940 buggy, causing it to skip the first lines of messages whose message ID
5941 ended in 'D'. This would not have bitten before Exim release 4.14, because
5942 message IDs were unlikely to end in 'D' before then. The effect was to have
5943 incorrect size information for certain domains.
5945 45. #include "config.h" was missing at the start of the crypt16.c module. This
5946 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5949 46. If there was a timeout during a "random" callout check, Exim treated it as
5950 a failure of the random address, and carried on sending RSET and the real
5951 address. If the delay was just some slowness somewhere, the response to the
5952 original RCPT would be taken as a response to RSET and so on, causing
5953 mayhem of various kinds.
5955 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5956 when I implemented it. It didn't allow for the fact that some option values
5957 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5958 the right test for positive values.
5960 48. Domain names in DNS records are case-independent. Exim always looks them up
5961 in lower case. Some resolvers return domain names in exactly the case they
5962 appear in the zone file, that is, they may contain uppercase letters. Not
5963 all resolvers do this - some return always lower case. Exim was treating a
5964 change of case by a resolver as a change of domain, similar to a widening
5965 of a domain abbreviation. This triggered its re-routing code and so it was
5966 trying to route what was effectively the same domain again. This normally
5967 caused routing to fail (because the router wouldn't handle the domain
5968 twice). Now Exim checks for this case specially, and just changes the
5969 casing of the domain that it ultimately uses when it transmits the message
5972 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5975 50. If a filter generated a file delivery with a non-absolute name (possible if
5976 no home directory exists for the router), the forbid_file option was not
5979 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5980 the existing equality matching.
5982 52. Exim was using ints instead of ino_t variables in some places where it was
5983 dealing with inode numbers.
5985 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5986 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5987 called TMPDIR, and if it finds it is different, it changes its value.
5989 54. The smtp_printf() function is now made available to local_scan() so
5990 additional output lines can be written before returning. There is also an
5991 smtp_fflush() function to enable the detection of a dropped connection.
5992 The variables smtp_input and smtp_batched_input are exported to
5995 55. Changed the default runtime configuration: the message "Unknown user"
5996 has been removed from the ACL, and instead placed on the localuser router,
5997 using the cannot_route_message feature. This means that any verification
5998 failures that generate their own messages won't get overridden. Similarly,
5999 the "Unrouteable address" message that was in the ACL for unverifiable
6000 relay addresses has also been removed.
6002 56. Added hosts_avoid_esmtp to the smtp transport.
6004 57. The exicyclog script was not checking for the esoteric option
6005 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6006 will work only if exicyclog is run under the appropriate euid.
6008 58. Following a discussion on the list, the rules by which Exim recognises line
6009 endings on incoming messages have been changed. The -dropcr and drop_cr
6010 options are now no-ops, retained only for backwards compatibility. The
6011 following line terminators are recognized: LF CRLF CR. However, special
6012 processing applies to CR:
6014 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6015 nor a local message in the state where . is a terminator.
6017 (ii) If a bare CR is encountered in a header line, an extra space is added
6018 after the line terminator so as not to end the header. The reasoning
6019 behind this is that bare CRs in header lines are most likely either
6020 to be mistakes, or people trying to play silly games.
6022 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6023 was being incorrectly given as 18 bytes larger than it should have been.
6024 This is a VOB (very old bug).
6026 60. This may never have affected anything current, but just in case it has:
6027 When the local host is found other than at the start of a list of hosts,
6028 the local host, those with the same MX, and any that follow, are discarded.
6029 When the list in question was part of a longer list of hosts, the following
6030 hosts (not currently being processed) were also being discarded. This no
6031 longer happens. I'm not sure if this situation could ever has previously
6034 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6037 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6038 header. This is something that is recommended in a new Internet Draft, and
6039 is something that is documented as being done by Sendmail. There are two
6040 possible values. For messages generated by the autoreply transport, Exim
6043 Auto-Submitted: auto-replied
6045 whereas for all other generated messages (e.g. bounces) it adds
6047 Auto-Submitted: auto-generated
6049 63. The "personal" condition in filters now includes a test for the
6050 Auto-Submitted: header. If it contains the string "auto-" the message it
6051 not considered personal.
6053 64. Added rcpt_include_affixes as a generic transport option.
6055 65. Added queue_only_override (default true).
6057 66. Added the syslog_duplication option.
6059 67. If what should have been the first header line of a message consisted of
6060 a space followed by a colon, Exim was mis-interpreting it as a header line.
6061 It isn't of course - it is syntactically invalid and should therefore be
6062 treated as the start of the message body. The misbehaviour could have
6063 caused a number of strange effects, including loss of data in subsequent
6064 header lines, and spool format errors.
6066 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6067 client host had authenticated. This control can now be exercised by an ACL
6068 for more flexibility.
6070 69. By default, callouts do not happen when testing with -bh. There is now a
6071 variant, -bhc, which does actually run the callout code, including
6072 consulting and updating the callout cache.
6074 70. Added support for saslauthd authentication, courtesy of Alexander
6077 71. If statvfs() failed on the spool or log directories while checking their
6078 size for availability, Exim confusingly gave the error "space shortage".
6079 Furthermore, in debugging mode it crashed with a floating point exception.
6080 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6081 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6082 serious problem, Exim now writes to the main and panic logs when this
6083 happens, with details of the failure. It then refuses to accept the
6084 incoming message, giving the message "spool directory problem" or "log
6085 directory problem" with a 421 code for SMTP messages.
6087 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6088 0, 1, and 2 exist, because some OS complain for execs without them (see
6089 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6090 descriptors. However, the code omitted to check that the open succeeded,
6091 causing mysterious errors if for some reason the permissions on /dev/null
6092 got screwed. Now Exim writes a message to the main and panic logs, and
6093 bombs out if it can't open /dev/null.
6095 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6096 interact so that it is all more flexible. It is supposed to remain
6097 backwards compatible. Also added extra_local_interfaces.
6099 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6100 to bomb out with an assertion failure - to the client this appears as a
6101 connection drop. This problem occurs in the part of the code that was taken
6102 from the Samba project. Fortunately, the assertion is in a very simple
6103 function, so I have fixed this by reproducing the function inline in the
6104 one place where it is called, and arranging for authentication to fail
6105 instead of killing the process with assert().
6107 75. The SPA client code was not working when the server requested OEM rather
6108 than Unicode encoding.
6110 76. Added code to make require_files with a specific uid setting more usable in
6111 the case where statting the file as root fails - usually a non-root-mounted
6112 NFS file system. When this happens and the failure is EACCES, Exim now
6113 forks a subprocess and does the per-uid checking as the relevant uid.
6115 77. Added process_log_path.
6117 78. If log_file_path was not explicitly set, a setting of check_log_space or
6118 check_log_inodes was ignored.
6120 79. If a space check for the spool or log partitions fails, the incident is now
6121 logged. Of course, in the latter case the data may get lost...
6123 80. Added the %p formatting code to string_format() so that it can be used to
6124 print addresses in debug_print(). Adjusted all the address printing in the
6125 debugging in store.c to use %p rather than %d.
6127 81. There was a concern that a line of code in smtp_in.c could overflow a
6128 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6129 initially expressed, the concern was not well-founded, because trailing
6130 spaces are removed early. However, if the trailing spaces were followed by
6131 a NULL, they did not get removed, so the overflow was possible. Two fixes
6134 (a) I re-wrote the offending code in a cleaner fashion.
6135 (b) If an incoming SMTP command contains a NULL character, it is rejected
6138 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6139 runs initgroups(), so that if the Exim user is in any additional groups,
6140 they will be used during message reception.
6146 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob