1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way. Bug 68.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and priority
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
61 and remote servers. Remote servers can be IPv6 addresses, and
64 JH/17 Bug 68: The spamd_address main option now supports an optional
65 timeout value per server.
67 JH/18 Bug 1581: Router and transport options headers_add/remove can
68 now have the list separator specified.
70 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
73 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
76 JH/21 Support for the A6 type of dns record is withdrawn.
78 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
79 rather than the verbs used.
81 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
82 from 255 to 1024 chars.
84 JH/24 Verification callouts now attempt to use TLS by default.
86 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
87 are generic router options now. The defaults didn't change.
89 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
90 Original patch from Alexander Shikoff, worked over by JH.
92 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
93 files if ZCAT_COMMAND is not executable.
95 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
97 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
99 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
100 Normally benign, it bites when the pair was led to by a CNAME;
101 modern usage is to not canoicalize the domain to a CNAME target
102 (and we were inconsistent anyway for A-only vs AAAA+A).
104 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
109 TL/01 When running the test suite, the README says that variables such as
110 no_msglog_check are global and can be placed anywhere in a specific
111 test's script, however it was observed that placement needed to be near
112 the beginning for it to behave that way. Changed the runtest perl
113 script to read through the entire script once to detect and set these
114 variables, reset to the beginning of the script, and then run through
115 the script parsing/test process like normal.
117 TL/02 The BSD's have an arc4random API. One of the functions to induce
118 adding randomness was arc4random_stir(), but it has been removed in
119 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
120 function when detected.
122 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
123 cause callback expansion.
125 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
126 syntax errors in an expansion can be treated as a string instead of
127 logging or causing an error, due to the internal use of bool_lax
128 instead of bool when processing it.
130 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
131 server certificates when making smtp deliveries.
133 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
135 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
137 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
139 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
140 Merged patch from Sebastian Wiedenroth.
142 JH/05 Fix results-pipe from transport process. Several recipients, combined
143 with certificate use, exposed issues where response data items split
144 over buffer boundaries were not parsed properly. This eventually
145 resulted in duplicates being sent. This issue only became common enough
146 to notice due to the introduction of conection certificate information,
147 the item size being so much larger. Found and fixed by Wolfgang Breyha.
149 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
150 size buffer was used, resulting in syntax errors when an expansion
153 JH/07 Add support for directories of certificates when compiled with a GnuTLS
154 version 3.3.6 or later.
156 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
157 is EXPERIMENTAL_EVENT, the main-configuration and transport options
158 both become "event_action", the variables become $event_name, $event_data
159 and $event_defer_errno. There is a new variable $verify_mode, usable in
160 routers, transports and related events. The tls:cert event is now also
161 raised for inbound connections, if the main configuration event_action
164 TL/06 In test suite, disable OCSP for old versions of openssl which contained
165 early OCSP support, but no stapling (appears to be less than 1.0.0).
167 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
168 server certificate names available under the smtp transport option
169 "tls_verify_cert_hostname" now do not permit multi-component wildcard
172 JH/10 Time-related extraction expansions from certificates now use the main
173 option "timezone" setting for output formatting, and are consistent
174 between OpenSSL and GnuTLS compilations. Bug 1541.
176 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
177 encoded parameter in the incoming message. Bug 1558.
179 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
180 include certificate info, eximon was claiming there were spoolfile
183 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
185 JH/14 Log delivery-related information more consistently, using the sequence
186 "H=<name> [<ip>]" wherever possible.
188 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
189 are problematic for Debian distribution, omit them from the release
192 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
194 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
197 JH/17 Fix a null-indirection in certextract expansions when a nondefault
198 output list separator was used.
203 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
204 checkers that were complaining about end of non-void function with no
207 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
208 This was a regression intruduced in 4.83 by another bugfix.
210 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
212 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
213 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
219 TF/01 Correctly close the server side of TLS when forking for delivery.
221 When a message was received over SMTP with TLS, Exim failed to clear up
222 the incoming connection properly after forking off the child process to
223 deliver the message. In some situations the subsequent outgoing
224 delivery connection happened to have the same fd number as the incoming
225 connection previously had. Exim would try to use TLS and fail, logging
226 a "Bad file descriptor" error.
228 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
229 utilities have not been installed.
231 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
232 temporary space as the ACL may create new global variables.
234 TL/01 LDAP support uses per connection or global context settings, depending
235 upon the detected version of the libraries at build time.
237 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
238 to extract and use the src ip:port in logging and expansions as if it
239 were a direct connection from the outside internet. PPv2 support was
240 updated based on HAProxy spec change in May 2014.
242 JH/02 Add ${listextract {number}{list}{success}{fail}}.
244 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
245 Properly escape header and check for NULL return.
247 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
250 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
252 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
253 characters in header names, implemented as a verify condition.
254 Contributed by Michael Fischer v. Mollard.
256 TL/05 Rename SPF condition results err_perm and err_temp to standardized
257 results permerror and temperror. Previous values are deprecated but
258 still accepted. In a future release, err_perm and err_temp will be
259 completely removed, which will be a backward incompatibility if the
260 ACL tests for either of these two old results. Patch contributed by
261 user bes-internal on the mailing list.
263 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
265 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
266 selectors, in both main and reject logs.
268 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
271 JH/07 Add malware type "sock" for talking to simple daemon.
273 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
275 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
276 routers/transports under cutthrough routing.
278 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
279 numbers. Touch up "bool" conditional to keep the same definition.
281 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
283 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
285 JH/12 Expand items in router/transport headers_add or headers_remove lists
286 individually rather than the list as a whole. Bug 1452.
288 Required for reasonable handling of multiple headers_ options when
289 they may be empty; requires that headers_remove items with embedded
290 colons must have them doubled (or the list-separator changed).
292 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
293 view the policy declared in the DMARC record. Currently, $dmarc_status
294 is a combined value of both the record presence and the result of the
297 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
299 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
300 dnslookup router and the smtp transport (applying to the forward
303 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
304 of ldap servers used for a specific lookup. Patch provided by Heiko
307 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
308 New variable $lookup_dnssec_authenticated for observability.
310 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
311 Patch submitted by Lars Timman.
313 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
315 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
316 Requires trusted mode and valid format message id, aborts otherwise.
317 Patch contributed by Heiko Schlichting.
319 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
320 certextract with support for various fields. Bug 1358.
322 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
323 is requested by default, modifiable by smtp transport option
326 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
327 operate on certificate variables to give certificate fingerprints
328 Also new ${sha256:cert_variable}.
330 JH/23 The PRDR feature is moved from being Experimental into the mainline.
332 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
333 Christian Aistleitner.
335 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
337 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
338 file. Patch from Wolfgang Breyha.
340 JH/25 Expand the coverage of the delivery $host and $host_address to
341 client authenticators run in verify callout. Bug 1476.
343 JH/26 Port service names are now accepted for tls_on_connect_ports, to
344 align with daemon_smtp_ports. Bug 72.
346 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
347 support and error reporting did not work properly.
349 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
350 and is readable. Patch from Andrew Colin Kissa.
352 TL/14 Enhance documentation of ${run expansion and how it parses the
353 commandline after expansion, particularly in the case when an
354 unquoted variable expansion results in an empty value.
356 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
358 PP/02 Fix internal collision of T_APL on systems which support RFC3123
359 by renaming away from it. Addresses GH issue 15, reported by
362 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
364 TL/15 SECURITY: prevent double expansion in math comparison functions
365 (can expand unsanitized data). Not remotely exploitable.
372 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
374 PP/02 Make -n do something, by making it not do something.
375 When combined with -bP, the name of an option is not output.
377 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
380 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
381 $sender_host_name and config options to manage this, and basic check
384 PP/05 DSCP support for outbound connections and control modifier for inbound.
386 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
387 (Only plugin which currently uses this is kerberos4, which nobody should
388 be using, but we should make it available and other future plugins might
389 conceivably use it, even though it would break NAT; stuff *should* be
390 using channel bindings instead).
392 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
393 name; added for Sendmail compatibility; requires admin caller.
394 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
395 just ignore it); requires trusted caller.
396 Also parse but ignore: -Ac -Am -X<logfile>
399 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
401 TL/02 Add +smtp_confirmation as a default logging option.
403 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
404 Patch by Magnus Holmgren from 2007-02-20.
406 TL/04 Bugzilla 1281 - Spec typo.
407 Bugzilla 1283 - Spec typo.
408 Bugzilla 1290 - Spec grammar fixes.
410 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
412 TL/06 Add Experimental DMARC support using libopendmarc libraries.
414 TL/07 Fix an out of order global option causing a segfault. Reported to dev
415 mailing list by by Dmitry Isaikin.
417 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
419 JH/02 Support "G" suffix to numbers in ${if comparisons.
421 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
423 NM/01 Bugzilla 1197 - Spec typo
424 Bugzilla 1196 - Spec examples corrections
426 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
428 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
429 gnutls_enable_pkcs11, but renamed to more accurately indicate its
432 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
433 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
435 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
436 "acl {{name}{arg}...}", and optional args on acl condition
439 JH/05 Permit multiple router/transport headers_add/remove lines.
441 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
443 JH/07 Avoid using a waiting database for a single-message-only transport.
444 Performance patch from Paul Fisher. Bugzilla 1262.
446 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
449 JH/09 Add $headers_added variable, with content from use of ACL modifier
450 add_header (but not yet added to the message). Bugzilla 199.
452 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
453 Pulled from Bugzilla 817 by Wolfgang Breyha.
455 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
457 (nb: this is the same fix as in Exim 4.80.1)
459 JH/11 Add A= logging on delivery lines, and a client_set_id option on
462 JH/12 Add optional authenticated_sender logging to A= and a log_selector
465 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
467 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
468 advertise SMTP AUTH mechanism to us, instead of a generic
469 protocol violation error. Also, make Exim more robust to bad
470 data from the Dovecot auth socket.
472 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
474 When a queue runner is handling a message, Exim first routes the
475 recipient addresses, during which it prunes them based on the retry
476 hints database. After that it attempts to deliver the message to
477 any remaining recipients. It then updates the hints database using
480 So if a recipient address works intermittently, it can get repeatedly
481 deferred at routing time. The retry hints record remains fresh so the
482 address never reaches the final cutoff time.
484 This is a fairly common occurrence when a user is bumping up against
485 their storage quota. Exim had some logic in its local delivery code
486 to deal with this. However it did not apply to per-recipient defers
487 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
489 This change adds a proper retry rule check during routing so that the
490 final cutoff time is checked against the message's age. We only do
491 this check if there is an address retry record and there is not a
492 domain retry record; this implies that previous attempts to handle
493 the address had the retry_use_local_parts option turned on. We use
494 this as an approximation for the destination being like a local
495 delivery, as in LMTP.
497 I suspect this new check makes the old local delivery cutoff check
498 redundant, but I have not verified this so I left the code in place.
500 TF/02 Correct gecos expansion when From: is a prefix of the username.
502 Test 0254 submits a message to Exim with the header
506 When I ran the test suite under the user fanf2, Exim expanded
507 the header to contain my full name, whereas it should have added
508 a Resent-Sender: header. It erroneously treats any prefix of the
509 username as equal to the username.
511 This change corrects that bug.
513 GF/01 DCC debug and logging tidyup
514 Error conditions log to paniclog rather than rejectlog.
515 Debug lines prefixed by "DCC: " to remove any ambiguity.
517 TF/03 Avoid unnecessary rebuilds of lookup-related code.
519 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
520 Bug spotted by Jeremy Harris; was flawed since initial commit.
521 Would have resulted in OCSP responses post-SNI triggering an Exim
522 NULL dereference and crash.
524 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
526 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
527 Bug detection, analysis and fix by Samuel Thibault.
528 Bugzilla 1331, Debian bug #698092.
530 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
532 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
533 Server implementation by Todd Lyons, client by JH.
534 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
535 config variable "prdr_enable" controls whether the server
536 advertises the facility. If the client requests PRDR a new
537 acl_data_smtp_prdr ACL is called once for each recipient, after
538 the body content is received and before the acl_smtp_data ACL.
539 The client is controlled by bolth of: a hosts_try_prdr option
540 on the smtp transport, and the server advertisement.
541 Default client logging of deliveries and rejections involving
542 PRDR are flagged with the string "PRDR".
544 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
545 fclose(). Diagnosis by Todd Lyons.
547 PP/17 Update configure.default to handle IPv6 localhost better.
548 Patch by Alain Williams (plus minor tweaks).
551 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
552 This is now consistent with GnuTLS, and is now documented: the
553 previous undocumented portable approach to treating the option as
554 unset was to force an expansion failure. That still works, and
555 an empty string is now equivalent.
557 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
558 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
559 not performing validation itself.
561 PP/20 Added force_command boolean option to pipe transport.
562 Patch from Nick Koston, of cPanel Inc.
564 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
567 TF/04 Added udpsend ACL modifer and hexquote expansion operator
569 PP/21 Fix eximon continuous updating with timestamped log-files.
570 Broken in a format-string cleanup in 4.80, missed when I repaired the
571 other false fix of the same issue.
572 Report and fix from Heiko Schlichting.
575 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
576 Report from Prashanth Katuri.
578 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
579 It's SecureTransport, so affects any MacOS clients which use the
580 system-integrated TLS libraries, including email clients.
582 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
583 using a MIME ACL for non-SMTP local injection.
584 Report and assistance in diagnosis by Warren Baker.
586 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
588 JH/16 Fix comparisons for 64b. Bugzilla 1385.
590 TL/09 Add expansion variable $authenticated_fail_id to keep track of
591 last id that failed so it may be referenced in subsequent ACL's.
593 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
596 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
597 ldap library initialization, allowing self-signed CA's to be
598 used. Also properly sets require_cert option later in code by
599 using NULL (global ldap config) instead of ldap handle (per
600 session). Bug diagnosis and testing by alxgomz.
602 TL/12 Enhanced documentation in the ratelimit.pl script provided in
603 the src/util/ subdirectory.
605 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
606 renamed to Transport Post Delivery Action by Jeremy Harris, as
609 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
610 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
611 redis_servers = needs to be configured which will be used by the redis
612 lookup. Patch from Warren Baker, of The Packet Hub.
614 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
616 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
617 hostname or reverse DNS when processing a host list. Used suggestions
618 from multiple comments on this bug.
620 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
622 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
623 Missed a few lines, added it to make the runtest require no keyboard
626 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
627 contains upper case chars. Make router use caseful_local_part.
629 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
630 support when GnuTLS has been built with p11-kit.
636 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
638 This, or similar/improved, will also be change PP/11 of 4.82.
644 PP/01 Handle short writes when writing local log-files.
645 In practice, only affects FreeBSD (8 onwards).
646 Bugzilla 1053, with thanks to Dmitry Isaikin.
648 NM/01 Bugzilla 949 - Documentation tweak
650 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
653 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
655 PP/02 Implemented gsasl authenticator.
657 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
659 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
660 `pkg-config foo` for cflags/libs.
662 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
663 with rest of GSASL and with heimdal_gssapi.
665 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
666 `pkg-config foo` for cflags/libs for the TLS implementation.
668 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
669 properties get this fed in as external SSF. A number of robustness
670 and debugging improvements to the cyrus_sasl authenticator.
672 PP/08 cyrus_sasl server now expands the server_realm option.
674 PP/09 Bugzilla 1214 - Log authentication information in reject log.
675 Patch by Jeremy Harris.
677 PP/10 Added dbmjz lookup type.
679 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
681 PP/12 MAIL args handles TAB as well as SP, for better interop with
682 non-compliant senders.
683 Analysis and variant patch by Todd Lyons.
685 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
686 Bug report from Lars Müller <lars@samba.org> (via SUSE),
687 Patch from Dirk Mueller <dmueller@suse.com>
689 PP/13 tls_peerdn now print-escaped for spool files.
690 Observed some $tls_peerdn in wild which contained \n, which resulted
691 in spool file corruption.
693 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
694 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
695 or write after TLS renegotiation, which otherwise led to messages
698 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
699 as a tracking header (ie: a signed header comes before the signature).
700 Patch from Wolfgang Breyha.
702 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
703 comma-sep list; embedded commas doubled.
705 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
707 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
709 Report and patch from Dmitry Banschikov.
711 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
712 Removed SSL_clear() after SSL_new() which led to protocol negotiation
713 failures. We appear to now support TLS1.1+ with Exim.
715 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
716 lets Exim select keys and certificates based upon TLS SNI from client.
717 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
718 before an outbound SMTP session. New log_selector, +tls_sni.
720 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
721 NULL dereference. Report and patch from Alun Jones.
723 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
724 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
725 Not seeing resolver debug output on NetBSD, but suspect this is a
726 resolver implementation change.
728 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
729 Left warnings. Added "eximon gdb" invocation mode.
731 PP/21 Defaulting "accept_8bitmime" to true, not false.
733 PP/22 Added -bw for inetd wait mode support.
735 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
736 locate the relevant includes and libraries. Made this the default.
738 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
739 Bugzilla 1246, report and most of solution from Tomasz Kusy.
741 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
742 This may cause build issues on older platforms.
744 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
745 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
746 gnutls_require_mac & gnutls_require_protocols (no longer supported).
747 Added SNI support via GnuTLS too.
748 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
750 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
752 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
753 Applied second patch from Janne, implementing suggestion to default
754 multiple-strings-in-record handling to match SPF spec.
756 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
758 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
759 read-only, out of scope).
760 Patch from Wolfgang Breyha, report from Stuart Northfield.
762 PP/29 Fix three issues highlighted by clang analyser static analysis.
763 Only crash-plausible issue would require the Cambridge-specific
764 iplookup router and a misconfiguration.
765 Report from Marcin Mirosław.
767 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
769 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
770 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
771 As part of this, removing so much warning spew let me fix some minor
772 real issues in debug logging.
774 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
775 assignment on my part. Fixed.
777 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
778 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
779 Janne Snabb (who went above and beyond: thank you).
781 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
782 string otherwise requires a connection and a bunch more work and it's
783 relatively easy to get wrong. Should also expose TLS library linkage
786 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
787 64-bit ${eval} (JH/03).
789 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
790 GNU libc to support some of the 64-bit stuff, should not lead to
791 conflicts. Defined before os.h is pulled in, so if a given platform
792 needs to override this, it can.
794 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
795 protection layer was required, which is not implemented.
796 Bugzilla 1254, patch from Wolfgang Breyha.
798 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
799 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
800 tls_dhparam take prime identifiers. Also unbreak combination of
801 OpenSSL+DH_params+TLSSNI.
803 PP/39 Disable SSLv2 by default in OpenSSL support.
809 PP/01 Solaris build fix for Oracle's LDAP libraries.
810 Bugzilla 1109, patch from Stephen Usher.
812 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
814 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
817 TF/02 Fix a couple more cases where we did not log the error message
818 when unlink() failed. See also change 4.74-TF/03.
820 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
821 lock up or crash if it happened to be inside a call to libc when it
822 got a SIGUSR1 from exiwhat.
824 The SIGUSR1 handler appends the current process status to the process
825 log which is later printed by exiwhat. It used to use the general
826 purpose logging code to do this, but several functions it calls are
827 not safe for signals.
829 The new output code in the SIGUSR1 handler is specific to the process
830 log, and simple enough that it's easy to inspect for signal safety.
831 Removing some special cases also simplifies the general logging code.
832 Removing the spurious timestamps from the process log simplifies
835 TF/04 Improved ratelimit ACL condition.
837 The /noupdate option has been deprecated in favour of /readonly which
838 has clearer semantics. The /leaky, /strict, and /readonly update modes
839 are mutually exclusive. The update mode is no longer included in the
840 database key; it just determines when the database is updated. (This
841 means that when you upgrde Exim will forget old rate measurements.)
843 Exim now checks that the per_* options are used with an update mode that
844 makes sense for the current ACL. For example, when Exim is processing a
845 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
846 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
847 must specify per_mail/readonly. If you omit the update mode it defaults to
848 /leaky where that makes sense (as before) or /readonly where required.
850 The /noupdate option is now undocumented but still supported for
851 backwards compatibility. It is equivalent to /readonly except that in
852 ACLs where /readonly is required you may specify /leaky/noupdate or
853 /strict/noupdate which are treated the same as /readonly.
855 A useful new feature is the /count= option. This is a generalization
856 of the per_byte option, so that you can measure the throughput of other
857 aggregate values. For example, the per_byte option is now equivalent
858 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
860 The per_rcpt option has been generalized using the /count= mechanism
861 (though it's more complicated than the per_byte equivalence). When it is
862 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
863 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
864 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
865 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
866 Note that using per_rcpt with a non-readonly update mode in more than
867 one ACL will cause the recipients to be double-counted. (The per_mail
868 and per_byte options don't have this problem.)
870 The handling of very low rates has changed slightly. If the computed rate
871 is less than the event's count (usually one) then this event is the first
872 after a long gap. In this case the rate is set to the same as this event's
873 count, so that the first message of a spam run is counted properly.
875 The major new feature is a mechanism for counting the rate of unique
876 events. The new per_addr option counts the number of different
877 recipients that someone has sent messages to in the last time period. It
878 behaves like per_rcpt if all the recipient addresses are different, but
879 duplicate recipient addresses do not increase the measured rate. Like
880 the /count= option this is a general mechanism, so the per_addr option
881 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
882 example, measure the rate that a client uses different sender addresses
883 with the options per_mail/unique=$sender_address. There are further
884 details in the main documentation.
886 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
888 TF/06 Removed a few PCRE remnants.
890 TF/07 Automatically extract Exim's version number from tags in the git
891 repository when doing development or release builds.
893 PP/02 Raise smtp_cmd_buffer_size to 16kB.
894 Bugzilla 879. Patch from Paul Fisher.
896 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
897 Heavily based on revision 40f9a89a from Simon Arlott's tree.
900 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
902 PP/05 Variable $av_failed, true if the AV scanner deferred.
903 Bugzilla 1078. Patch from John Horne.
905 PP/06 Stop make process more reliably on build failure.
906 Bugzilla 1087. Patch from Heiko Schlittermann.
908 PP/07 Make maildir_use_size_file an _expandable_ boolean.
909 Bugzilla 1089. Patch from Heiko Schlittermann.
911 PP/08 Handle ${run} returning more data than OS pipe buffer size.
912 Bugzilla 1131. Patch from Holger Weiß.
914 PP/09 Handle IPv6 addresses with SPF.
915 Bugzilla 860. Patch from Wolfgang Breyha.
917 PP/10 GnuTLS: support TLS 1.2 & 1.1.
919 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
922 PP/11 match_* no longer expand right-hand-side by default.
923 New compile-time build option, EXPAND_LISTMATCH_RHS.
924 New expansion conditions, "inlist", "inlisti".
926 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
928 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
930 PP/14 fix log_write() format string regression from TF/03.
931 Bugzilla 1152. Patch from Dmitry Isaikin.
937 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
939 PP/02 Harmonised TLS library version reporting; only show if debugging.
940 Layout now matches that introduced for other libraries in 4.74 PP/03.
942 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
944 PP/04 New "dns_use_edns0" global option.
946 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
949 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
950 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
952 TK/01 Updated PolarSSL code to 0.14.2.
953 Bugzilla 1097. Patch from Andreas Metzler.
955 PP/07 Catch divide-by-zero in ${eval:...}.
958 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
961 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
962 format-string attack -- SECURITY: remote arbitrary code execution.
964 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
965 time unintentionally subject to list matching rules, letting the header
966 cause arbitrary Exim lookups (of items which can occur in lists, *not*
967 arbitrary string expansion). This allowed for information disclosure.
969 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
970 INT_MIN/-1 -- value coerced to INT_MAX.
976 NM/01 Workround for PCRE version dependency in version reporting
979 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
980 This fixes portability to compilers other than gcc, notably
981 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
983 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
984 makefiles for portability to HP-UX and POSIX correctness.
986 PP/01 Permit LOOKUP_foo enabling on the make command-line.
987 Also via indented variable definition in the Makefile.
988 (Debugging by Oliver Heesakkers).
990 PP/02 Restore caching of spamd results with expanded spamd_address.
991 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
993 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
994 Improves build reliability. Fix from: Frank Elsner
996 NM/02 Fix wide character breakage in the rfc2047 coding
997 Fixes bug 1064. Patch from Andrey N. Oktyabrski
999 NM/03 Allow underscore in dnslist lookups
1000 Fixes bug 1026. Patch from Graeme Fowler
1002 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1003 Code patches from Adam Ciarcinski of NetBSD.
1005 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1008 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1009 is logged, to avoid truncation. Patch from John Horne.
1011 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1012 Patch from Jakob Hirsch.
1014 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1015 SQL string expansion failure details.
1016 Patch from Andrey Oktyabrski.
1018 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1019 Patch from Simon Arlott.
1021 PP/09 New lookups functionality failed to compile on old gcc which rejects
1022 extern declarations in function scope.
1023 Patch from Oliver Fleischmann
1025 PP/10 Use sig_atomic_t for flags set from signal handlers.
1026 Check getgroups() return and improve debugging.
1027 Fixed developed for diagnosis in bug 927 (which turned out to be
1030 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1031 Patch from Mark Zealey.
1033 PP/12 Bugzilla 1056: Improved spamd server selection.
1034 Patch from Mark Zealey.
1036 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1037 Based on patch from Heiko Schlittermann.
1039 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1040 Patch from Uwe Doering, sign-off by Michael Haardt.
1042 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1043 variable declaration deep within a block. Bug and patch from
1046 PP/15 lookups-Makefile IRIX compatibilty coercion.
1048 PP/16 Make DISABLE_DKIM build knob functional.
1050 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1051 Patch by Simon Arlott
1053 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1054 variable argument macros. Our copy now differs from upstream.
1060 TF/01 Failure to get a lock on a hints database can have serious
1061 consequences so log it to the panic log.
1063 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1064 controlled using the smtp_confirmation log selector.
1066 TF/03 Include the error message when we fail to unlink a spool file.
1068 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1069 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1070 for maintaining out-of-tree patches for some time.
1072 PP/01 Bugzilla 139: Documentation and portability issues.
1073 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1074 Handle per-OS dynamic-module compilation flags.
1076 PP/02 Let /dev/null have normal permissions.
1077 The 4.73 fixes were a little too stringent and complained about the
1078 permissions on /dev/null. Exempt it from some checks.
1079 Reported by Andreas M. Kirchwitz.
1081 PP/03 Report version information for many libraries, including
1082 Exim version information for dynamically loaded libraries. Created
1083 version.h, now support a version extension string for distributors
1084 who patch heavily. Dynamic module ABI change.
1086 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1087 privilege escalation vulnerability whereby the Exim run-time user
1088 can cause root to append content of the attacker's choosing to
1091 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1094 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1095 If dropping privileges for untrusted macros, we disabled normal logging
1096 on the basis that it would fail; for the Exim run-time user, this is not
1097 the case, and it resulted in successful deliveries going unlogged.
1098 Fixed. Reported by Andreas Metzler.
1104 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1105 only prepend for the Resent-* case. Fixes regression introduced in
1106 Exim 4.70 by NM/22 for Bugzilla 607.
1108 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1109 increasing numbers of administrators be bitten by this.
1111 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1113 PP/03 Bugzilla 994: added openssl_options main configuration option.
1115 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1117 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1119 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1121 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1122 without a peer certificate, leading to a segfault because of an
1123 assumption that peers always have certificates. Be a little more
1124 paranoid. Problem reported by Martin Tscholak.
1126 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1127 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1128 NB: ClamAV planning to remove STREAM in "middle of 2010".
1129 CL also introduces -bmalware, various -d+acl logging additions and
1130 more caution in buffer sizes.
1132 PP/09 Implemented reverse_ip expansion operator.
1134 PP/10 Bugzilla 937: provide a "debug" ACL control.
1136 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1138 PP/12 Bugzilla 973: Implement --version.
1140 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1142 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1144 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1146 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1147 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1148 ignore trailing whitespace.
1150 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1152 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1155 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1156 Notification from Dr Andrew Aitchison.
1158 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1159 ExtendedDetectionInfo response format.
1160 Notification from John Horne.
1162 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1165 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1166 XSL and documented dependency on system catalogs, with examples of how
1169 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1172 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1173 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1174 configuration file which is writeable by the Exim user or group.
1176 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1177 of configuration files to cover files specified with the -C option if
1178 they are going to be used with root privileges, not just the default
1181 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1182 option (effectively making it always true).
1184 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1185 files to be used while preserving root privileges.
1187 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1188 that rogue child processes cannot use them.
1190 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1191 run-time user, instead of root.
1193 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1194 Exim run-time user without dropping privileges.
1196 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1197 result string, instead of calling string_vformat() twice with the same
1200 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1201 for other users. Others should always drop root privileges if they use
1202 -C on the command line, even for a whitelisted configure file.
1204 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1206 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1212 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1213 $data_path, and $header_path variables; fixed documentation bugs and
1216 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1217 exipick to access non-standard spools, including the "frozen" queue
1220 NM/01 Bugzilla 965: Support mysql stored procedures.
1221 Patch from Alain Williams
1223 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1225 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1226 Patch from Andreas Metzler
1228 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1229 Patch from Kirill Miazine
1231 NM/05 Bugzilla 671: Added umask to procmail example.
1233 JJ/03 installed exipick 20100323.0, fixing doc bug
1235 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1236 directory. Notification and patch from Dan Rosenberg.
1238 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1240 TK/02 Improve log output when DKIM signing operation fails.
1242 MH/01 Treat the transport option dkim_domain as a colon separated
1243 list, not as a single string, and sign the message with each element,
1244 omitting multiple occurences of the same signer.
1246 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1247 Bugzilla 985, 986. Patch by Simon Arlott
1249 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1250 Patch by Simon Arlott
1252 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1253 MBX locking. Notification from Dan Rosenberg.
1259 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1261 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1263 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1265 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1267 NM/04 Bugzilla 915: Flag broken perl installation during build.
1273 TK/01 Added patch by Johannes Berg that expands the main option
1274 "spamd_address" if it starts with a dollar sign.
1276 TK/02 Write list of recipients to X-Envelope-Sender header when building
1277 the mbox-format spool file for content scanning (suggested by Jakob
1280 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1281 (http://www.dcc-servers.net/) support via dccifd. Activated by
1282 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1284 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1285 by Mark Daniel Reidel <mr@df.eu>.
1287 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1288 When building exim an external PCRE library is now needed -
1289 PCRE is a system library on the majority of modern systems.
1290 See entry on PCRE_LIBS in EDITME file.
1292 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1293 conversation. Added nologin parameter to request.
1294 Patch contributed by Kirill Miazine.
1296 TF/01 Do not log submission mode rewrites if they do not change the address.
1298 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1300 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1301 log files in place. Contributed by Roberto Lima.
1303 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1305 TF/03 Bugzilla 615: When checking the local_parts router precondition
1306 after a local_part_suffix or local_part_prefix option, Exim now
1307 does not use the address's named list lookup cache, since this
1308 contains cached lookups for the whole local part.
1310 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1311 Robert Millan. Documentation is in experimental-spec.txt.
1313 TF/04 Bugzilla 668: Fix parallel build (make -j).
1315 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1317 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1318 Patch provided by Jan Srzednicki.
1320 TF/05 Leading white space used to be stripped from $spam_report which
1321 wrecked the formatting. Now it is preserved.
1323 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1324 that they are available at delivery time.
1326 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1328 TF/08 TLS error reporting now respects the incoming_interface and
1329 incoming_port log selectors.
1331 TF/09 Produce a more useful error message if an SMTP transport's hosts
1332 setting expands to an empty string.
1334 NM/06 Bugzilla 744: EXPN did not work under TLS.
1335 Patch provided by Phil Pennock.
1337 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1338 Patch provided by Richard Godbee.
1340 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1341 acl_smtp_notquit, added index entry.
1343 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1344 Patch provided by Eugene Bujak.
1346 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1347 accept(). Patch provided by Maxim Dounin.
1349 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1350 Patch provided by Phil Pennock.
1352 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1354 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1355 Patch provided by Brad "anomie" Jorsch.
1357 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1358 Patch provided by Dean Brooks.
1360 TK/05 Add native DKIM support (does not depend on external libraries).
1362 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1363 Patch provided by Graeme Fowler.
1365 NM/16 Bugzilla 851: Documentation example syntax fix.
1367 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1369 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1372 NM/19 Bugzilla 745: TLS version reporting.
1373 Patch provided by Phil Pennock.
1375 NM/20 Bugzilla 167: bool: condition support.
1376 Patch provided by Phil Pennock.
1378 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1379 clients. Patch provided by Phil Pennock.
1381 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1382 Patch provided by Brad "anomie" Jorsch.
1384 NM/23 Bugzilla 687: Fix misparses in eximstats.
1385 Patch provided by Heiko Schlittermann.
1387 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1388 Patch provided by Heiko Schlittermann.
1390 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1391 plus update to original patch.
1393 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1395 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1396 Patch provided by David Brownlee.
1398 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1400 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1402 NM/30 Bugzilla 888: TLS documentation bugfixes.
1404 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1406 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1407 Unlike the original bugzilla I have changed all shell scripts in src tree.
1409 NM/33 Bugzilla 898: Transport filter timeout fix.
1410 Patch by Todd Rinaldo.
1412 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1413 Patch by Serge Demonchaux.
1415 NM/35 Bugzilla 39: Base64 decode bug fixes.
1416 Patch by Jakob Hirsch.
1418 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1420 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1422 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1424 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1430 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1431 ALT-N's libdkim that I have put here:
1432 http://duncanthrax.net/exim-experimental/
1434 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1435 were called 'true' and it seems that C99 defines that as a reserved
1436 keyword to be used with 'bool' variable types. That means you could
1437 not include C99-style headers which use bools without triggering
1438 build errors in sieve.c.
1440 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1441 as mailq or other aliases. Changed the --help handling significantly
1442 to do whats expected. exim_usage() emits usage/help information.
1444 SC/01 Added the -bylocaldomain option to eximstats.
1446 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1448 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1450 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1456 PH/01 Another patch from the Sieve maintainer.
1458 PH/02 When an IPv6 address is converted to a string for single-key lookup
1459 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1460 dots are used instead of colons so that keys in lsearch files need not
1461 contain colons. This was done some time before quoting was made available
1462 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1463 (notated using the quote facility) so as to distinguish them from IPv4
1464 keys. This meant that lookups for IP addresses in host lists did not work
1465 for iplsearch lookups.
1467 This has been fixed by arranging for IPv6 addresses to be expressed with
1468 colons if the lookup type is iplsearch. This is not incompatible, because
1469 previously such lookups could never work.
1471 The situation is now rather anomolous, since one *can* have colons in
1472 ordinary lsearch keys. However, making the change in all cases is
1473 incompatible and would probably break a number of configurations.
1475 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1478 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1479 conversion specification without a maximum field width, thereby enabling
1480 a rogue spamd server to cause a buffer overflow. While nobody in their
1481 right mind would setup Exim to query an untrusted spamd server, an
1482 attacker that gains access to a server running spamd could potentially
1483 exploit this vulnerability to run arbitrary code as the Exim user.
1485 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1486 $primary_hostname instead of what libspf2 thinks the hosts name is.
1488 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1489 a directory entry by the name of the lookup key. Previously, if a
1490 symlink pointed to a non-existing file or a file in a directory that
1491 Exim lacked permissions to read, a lookup for a key matching that
1492 symlink would fail. Now it is enough that a matching directory entry
1493 exists, symlink or not. (Bugzilla 503.)
1495 PH/03 The body_linecount and body_zerocount variables are now exported in the
1498 PH/04 Added the $dnslist_matched variable.
1500 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1501 This means they are set thereafter only if the connection becomes
1504 PH/06 Added the client_condition to authenticators so that some can be skipped
1505 by clients under certain conditions.
1507 PH/07 The error message for a badly-placed control=no_multiline_responses left
1508 "_responses" off the end of the name.
1510 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1512 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1513 (without spaces) instead of just copying the configuration text.
1515 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1517 PH/11 Added $max_received_linelength.
1519 PH/12 Added +ignore_defer and +include_defer to host lists.
1521 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1522 way in which PCRE > 7.0 is built.
1524 PH/14 Implemented queue_only_load_latch.
1526 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1527 MAIL command. The effect was to mangle the value on 64-bit systems.
1529 PH/16 Another patch from the Sieve maintainer.
1531 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1533 PH/18 If a system quota error occurred while trying to create the file for
1534 a maildir delivery, the message "Mailbox is full" was not appended to the
1535 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1536 only to a quota excession during the actual writing of the file.
1538 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1539 characters?) which causes problems in log lines. The DN values are now
1540 passed through string_printing() before being added to log lines.
1542 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1543 and InterBase are left for another time.)
1545 PH/21 Added message_body_newlines option.
1547 PH/22 Guard against possible overflow in moan_check_errorcopy().
1549 PH/23 POSIX allows open() to be a macro; guard against that.
1551 PH/24 If the recipient of an error message contained an @ in the local part
1552 (suitably quoted, of course), incorrect values were put in $domain and
1553 $local_part during the evaluation of errors_copy.
1559 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1560 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1563 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1564 issue a MAIL command.
1566 PH/02 In an ACL statement such as
1568 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1570 if a client was not listed at all, or was listed with a value other than
1571 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1572 the condition was not true (as it should be), so access was not denied.
1573 The bug was that the ! inversion was incorrectly passed on to the second
1574 item. This has been fixed.
1576 PH/03 Added additional dnslists conditions == and =& which are different from
1577 = and & when the dns lookup returns more than one IP address.
1579 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1580 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1582 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1583 FSYNC, which compiles an option called disable_fsync that allows for
1584 bypassing fsync(). The documentation is heavily laced with warnings.
1586 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1588 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1589 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1590 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1591 including adding "make clean"; (3) Added -fPIC when compiling the test
1592 dynamically loaded module, to get rid of a warning.
1594 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1595 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1596 = 0s. The bug is otherwise harmless.
1598 PH/07 There was a bug in the dovecot authenticator such that the value of
1599 $auth1 could be overwritten, and so not correctly preserved, after a
1600 successful authentication. This usually meant that the value preserved by
1601 the server_setid option was incorrect.
1603 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1605 PH/09 Installed PCRE release 7.0.
1607 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1608 run for batched SMTP input. It is now run at the start of every message
1609 in the batch. While fixing this I discovered that the process information
1610 (output by running exiwhat) was not always getting set for -bs and -bS
1611 input. This is fixed, and it now also says "batched" for BSMTP.
1613 PH/11 Added control=no_pipelining.
1615 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1616 patch, slightly modified), and move the expansion of helo_data till after
1617 the connection is made in the smtp transport (so it can use these
1620 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1622 PH/14 Added log_selector = +pid.
1624 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1626 PH/16 Add ${if forany and ${if forall.
1628 PH/17 Added dsn_from option to vary the From: line in DSNs.
1630 PH/18 Flush SMTP output before performing a callout, unless control =
1631 no_callout_flush is set.
1633 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1634 was true (the default) a successful delivery failed to delete the retry
1635 item, thus causing premature timeout of the address. The bug is now
1638 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1640 PH/21 Long custom messages for fakedefer and fakereject are now split up
1641 into multiline reponses in the same way that messages for "deny" and
1642 other ACL rejections are.
1644 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1645 with slight modification.
1647 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1648 draft, changing the syntax and factoring some duplicate code".
1650 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1651 for deliveries of the second and subsequent messages over the same SMTP
1654 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1655 ${reduce, with only minor "tidies".
1657 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1659 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1660 expansion side effects.
1662 PH/27 When a message times out after an over-quota error from an Exim-imposed
1663 quota, the bounce message says "mailbox is full". This message was not
1664 being given when it was a system quota that was exceeded. It now should
1667 MH/03 Made $recipients available in local_scan(). local_scan() already has
1668 better access to the recipient list through recipients_list[], but
1669 $recipients can be useful in postmaster-provided expansion strings.
1671 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1672 in the case of a MAIL command with additional options following the
1673 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1674 were accidentally chopped off.
1676 PH/29 SMTP synchronization checks are implemented when a command is read -
1677 there is a check that no more input is waiting when there shouldn't be
1678 any. However, for some commands, a delay in an ACL can mean that it is
1679 some time before the response is written. In this time, more input might
1680 arrive, invalidly. So now there are extra checks after an ACL has run for
1681 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1682 pipelining has not been advertised.
1684 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1686 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1687 returned in the SMTP response when smtp_return_error_details was false.
1688 This has been fixed.
1690 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1691 instead of the C I/O that was originally supplied, because problems were
1692 reported on Solaris.
1694 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1695 Exim which did not show up earlier: it was assuming that a call to
1696 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1697 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1698 was a macro that became an assignment, so it seemed to work. This has
1699 changed to a proper function call with a void return, hence the compile
1700 error. Exim's code has been fixed.
1702 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1705 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1707 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1709 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1710 because it turns out to be too big or there is a timeout) there is no
1711 "Completed" line in the log. When some messages of this type were
1712 selected by exigrep, they were listed as "not completed". Others were
1713 picked up by some special patterns. I have improved the selection
1714 criteria to be more general.
1716 PH/38 The host_find_failed option in the manualroute router can now be set
1717 to "ignore", to completely ignore a host whose IP address cannot be
1718 found. If all hosts are ignored, the behaviour is controlled by the new
1719 host_all_ignored option.
1721 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1722 homing or because of multiple MX records with /mx) generated more than
1723 one IP address, and the following item turned out to be the local host,
1724 all the secondary addresses of the first item were incorrectly removed
1725 from the list, along with the local host and any following hosts (which
1726 is what is supposed to happen).
1728 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1729 whoever called Exim into the -H file. In the case of the daemon it was
1730 behaving confusingly. When first started, it used values for whoever
1731 started the daemon, but after a SIGHUP it used the Exim user (because it
1732 calls itself on a restart). I have changed the code so that it now always
1735 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1736 message are rejected with the same error (e.g. no authentication or bad
1737 sender address), and a DATA command is nevertheless sent (as can happen
1738 with PIPELINING or a stupid MUA), the error message that was given to the
1739 RCPT commands is included in the rejection of the DATA command. This is
1740 intended to be helpful for MUAs that show only the final error to their
1743 PH/42 Another patch from the Sieve maintainer.
1745 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1746 Eximstats - Fixed some broken HTML links and added missing column headers
1748 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1749 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1751 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1753 PH/43 Yet another patch from the Sieve maintainer.
1755 PH/44 I found a way to check for a TCP/IP connection going away before sending
1756 the response to the final '.' that terminates a message, but only in the
1757 case where the client has not sent further data following the '.'
1758 (unfortunately, this is allowed). However, in many cases there won't be
1759 any further data because there won't be any more messages to send. A call
1760 to select() can be used: if it shows that the input is "ready", there is
1761 either input waiting, or the socket has been closed. An attempt to read
1762 the next input character can distinguish the two cases. Previously, Exim
1763 would have sent an OK response which the client would never have see.
1764 This could lead to message repetition. This fix should cure that, at
1765 least in a lot of common cases.
1767 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1768 advertised in response to EHLO.
1774 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1775 fixed by 4.65/MH/01 (is this a record?) are fixed:
1777 (i) An empty string was always treated as zero by the numeric comparison
1778 operators. This behaviour has been restored.
1780 (ii) It is documented that the numeric comparison operators always treat
1781 their arguments as decimal numbers. This was broken in that numbers
1782 starting with 0 were being interpreted as octal.
1784 While fixing these problems I realized that there was another issue that
1785 hadn't been noticed. Values of message_size_limit (both the global option
1786 and the transport option) were treated as octal if they started with 0.
1787 The documentation was vague. These values are now always treated as
1788 decimal, and I will make that clear in the documentation.
1794 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1795 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1798 MH/01 Don't check that the operands of numeric comparison operators are
1799 integers when their expansion is in "skipping" mode (fixes bug
1800 introduced by 4.64-PH/07).
1802 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1803 child addresses, Exim now panics and dies. Previously, because the count
1804 is held in a short int, deliveries were likely to be lost. As such a
1805 large number of recipients for a single message is ridiculous
1806 (performance will be very, very poor), I have chosen to impose a limit
1807 rather than extend the field.
1813 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1814 leftover -K file (the existence of which was triggered by #402).
1815 While we were at it, introduced process PID as part of the -K
1816 filename. This should rule out race conditions when creating
1819 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1820 processing considerably. Previous code took too long for large mails,
1821 triggering a timeout which in turn triggers #401.
1823 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1824 in the DK code in transports.c. sendfile() is not really portable,
1825 hence the _LINUX specificness.
1827 TF/01 In the add_headers option to the mail command in an Exim filter,
1828 there was a bug that Exim would claim a syntax error in any
1829 header after the first one which had an odd number of characters
1832 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1833 callout verification, Exim cached a "reject" for the entire domain. This
1834 is correct for most verifications, but it is not correct for a recipient
1835 verification with use_sender or use_postmaster set, because in that case
1836 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1837 case of MAIL FROM:<> rejection from other early rejections (e.g.
1838 rejection of HELO). When verifying a recipient using a non-null MAIL
1839 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1840 Whatever the result of the callout, the value of the domain cache is
1841 left unchanged (for any other kind of callout, getting as far as trying
1842 RCPT means that the domain itself is ok).
1844 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1847 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1848 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1849 session, the daemon ignores ECONNECT errors and logs others; it now
1850 ignores EPIPE as well.
1852 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1853 (quoted-printable decoding).
1855 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1856 later the small subsequent patch to fix an introduced bug.
1858 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1860 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1862 PH/08 An error is now given if message_size_limit is specified negative.
1864 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1865 to be given (somewhat) arbitrary names.
1867 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1870 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1871 miscellaneous code fixes
1873 PH/10 Added the log_reject_target ACL modifier to specify where to log
1876 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1877 hostname. This is wrong, because it relates to the incoming message (and
1878 probably the interface on which it is arriving) and not to the outgoing
1879 callout (which could be using a different interface). This has been
1880 changed to use the value of the helo_data option from the smtp transport
1881 instead - this is what is used when a message is actually being sent. If
1882 there is no remote transport (possible with a router that sets up host
1883 addresses), $smtp_active_hostname is used.
1885 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1886 tweaks were necessary in order to get it to work (see also 21 below):
1887 (a) The code assumed that strncpy() returns a negative number on buffer
1888 overflow, which isn't the case. Replaced with Exim's string_format()
1890 (b) There were several signed/unsigned issues. I just did the minimum
1891 hacking in of casts. There is scope for a larger refactoring.
1892 (c) The code used strcasecmp() which is not a standard C function.
1893 Replaced with Exim's strcmpic() function.
1894 (d) The code set only $1; it now sets $auth1 as well.
1895 (e) A simple test gave the error "authentication client didn't specify
1896 service in request". It would seem that Dovecot has changed its
1897 interface. Fortunately there's a specification; I followed it and
1898 changed what the client sends and it appears to be working now.
1900 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1903 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1904 address A is aliased to B and C, where B exists and C does not. Without
1905 -v the output is "A verified" because verification stops after a
1906 successful redirection if more than one address is generated. However,
1907 with -v the child addresses are also verified. Exim was outputting "A
1908 failed to verify" and then showing the successful verification for C,
1909 with its parentage. It now outputs "B failed to verify", showing B's
1910 parentage before showing the successful verification of C.
1912 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1913 look up a TXT record in a specific list after matching in a combined
1916 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1917 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1918 they consult the DNS. I had assumed they would set it the way they
1919 wanted; and indeed my experiments on Linux seem to show that in some
1920 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1921 To be on the safe side, however, I have now made the interface to
1922 host_find_byname() similar to host_find_bydns(), with an argument
1923 containing the DNS resolver options. The host_find_byname() function now
1924 sets these options at its start, just as host_find_bydns() does. The smtp
1925 transport options dns_qualify_single and dns_search_parents are passed to
1926 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1927 of host_find_byname() use the default settings of RES_DEFNAMES
1928 (qualify_single) but not RES_DNSRCH (search_parents).
1930 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1931 spool_read_header() do less string testing, by means of a preliminary
1932 switch on the second character of optional "-foo" lines. (This is
1933 overdue, caused by the large number of possibilities that now exist.
1934 Originally there were few.) While I was there, I also converted the
1935 str(n)cmp tests so they don't re-test the leading "-" and the first
1936 character, in the hope this might squeeze out yet more improvement.
1938 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1939 flag allowing group syntax was set by the header_syntax check but not
1940 turned off, possible causing trouble later; (2) The flag was not being
1941 set at all for the header_verify test, causing "group"-style headers to
1942 be rejected. I have now set it in this case, and also caused header_
1943 verify to ignore an empty address taken from a group. While doing this, I
1944 came across some other cases where the code for allowing group syntax
1945 while scanning a header line wasn't quite right (mostly, not resetting
1946 the flag correctly in the right place). These bugs could have caused
1947 trouble for malformed header lines. I hope it is now all correct.
1949 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1950 with the "reply" argument non-NULL. The code, however (which originally
1951 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1952 but it didn't always do it. This confused somebody who was copying the
1953 code for some other use. I have removed all the tests.
1955 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1956 feature that was used to support insecure browsers during the U.S. crypto
1957 embargo. It requires special client support, and Exim is probably the
1958 only MTA that supported it -- and would never use it because real RSA is
1959 always available. This code has been removed, because it had the bad
1960 effect of slowing Exim down by computing (never used) parameters for the
1961 RSA_EXPORT functionality.
1963 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1964 authenticator to fail if there's a tab character in the incoming data
1965 (there should never be unless someone is messing about, as it's supposed
1966 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1967 if the connection is using TLS or if the remote IP is the same as the
1968 local IP, and the "valid-client-cert option" if a client certificate has
1971 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1972 authenticators. This can be used for authorization after authentication
1973 succeeds. (In the case of plaintext, it servers for both authentication
1976 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1977 if any retry times were supplied.
1979 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1980 connection, where there is no client IP address to check. In this
1981 situation, the verify now always succeeds.
1983 PH/25 Applied John Jetmore's -Mset patch.
1985 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1987 PH/27 In a string expansion for a processed (not raw) header when multiple
1988 headers of the same name were present, leading whitespace was being
1989 removed from all of them, but trailing whitespace was being removed only
1990 from the last one. Now trailing whitespace is removed from each header
1991 before concatenation. Completely empty headers in a concatenation (as
1992 before) are ignored.
1994 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1995 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1997 PH/29 [Removed. This was a change that I later backed out, and forgot to
1998 correct the ChangeLog entry (that I had efficiently created) before
1999 committing the later change.]
2001 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2002 address errors (4xx response to RCPT) over the same connection as other
2003 messages routed to the same hosts. Such deliveries are always "forced",
2004 so retry times are not inspected. This resulted in far too many retries
2005 for the affected addresses. The effect occurred only when there were more
2006 hosts than the hosts_max_try setting in the smtp transport when it had
2007 the 4xx errors. Those hosts that it had tried were not added to the list
2008 of hosts for which the message was waiting, so if all were tried, there
2009 was no problem. Two fixes have been applied:
2011 (i) If there are any address or message errors in an SMTP delivery, none
2012 of the hosts (tried or untried) are now added to the list of hosts
2013 for which the message is waiting, so the message should not be a
2014 candidate for sending over the same connection that was used for a
2015 successful delivery of some other message. This seems entirely
2016 reasonable: after all the message is NOT "waiting for some host".
2017 This is so "obvious" that I'm not sure why it wasn't done
2018 previously. Hope I haven't missed anything, but it can't do any
2019 harm, as the worst effect is to miss an optimization.
2021 (ii) If, despite (i), such a delivery is accidentally attempted, the
2022 routing retry time is respected, so at least it doesn't keep
2023 hammering the server.
2025 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2026 in ${readsocket because some servers need this prod.
2028 PH/32 Added some extra debug output when updating a wait-xxx database.
2030 PH/33 The hint "could be header name not terminated by colon", which has been
2031 given for certain expansion errors for a long time, was not being given
2032 for the ${if def:h_colon_omitted{... case.
2034 PH/34 The spec says: "With one important exception, whenever a domain list is
2035 being scanned, $domain contains the subject domain." There was at least
2036 one case where this was not true.
2038 PH/35 The error "getsockname() failed: connection reset by peer" was being
2039 written to the panic log as well as the main log, but it isn't really
2040 panic-worthy as it just means the connection died rather early on. I have
2041 removed the panic log writing for the ECONNRESET error when getsockname()
2044 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2045 runs only) independently of the message's sender address. This meant
2046 that, if the 4xx error was in fact related to the sender, a different
2047 message to the same recipient with a different sender could confuse
2048 things. In particualar, this can happen when sending to a greylisting
2049 server, but other circumstances could also provoke similar problems.
2050 I have changed the default so that the retry time for these errors is now
2051 based a combination of the sender and recipient addresses. This change
2052 can be overridden by setting address_retry_include_sender=false in the
2055 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2056 remote server are returned as part of bounce messages. This was not
2057 happening for LMTP over a pipe (the lmtp transport), but now it is the
2058 same for both kinds of LMTP.
2060 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2061 in header lines that were in fact CNAMEs. This is no longer the case.
2063 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2064 and queue runs started by the daemon processed all messages. This has
2065 been fixed so that -R and -S can now usefully be given with -q<time>.
2067 PH/40 Import PCRE release 6.7 (fixes some bugs).
2069 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2071 PH/42 Give an error if -q is specified more than once.
2073 PH/43 Renamed the variables $interface_address and $interface_port as
2074 $received_ip_address and $received_port, to make it clear that these
2075 values apply to message reception, and not to the outgoing interface when
2076 a message is delivered. (The old names remain recognized, of course.)
2078 PH/44 There was no timeout on the connect() call when using a Unix domain
2079 socket in the ${readsocket expansion. There now is.
2081 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2082 be meaningful with "accept".
2084 SC/01 Eximstats V1.43
2085 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2087 SC/02 Eximstats V1.44
2088 Use a glob alias rather than an array ref in the generated
2089 parser. This improves both readability and performance.
2091 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2092 Collect SpamAssassin and rejection statistics.
2093 Don't display local sender or destination tables unless
2094 there is data to show.
2095 Added average volumes into the top table text output.
2097 SC/04 Eximstats V1.46
2098 Collect data on the number of addresses (recipients)
2099 as well as the number of messages.
2101 SC/05 Eximstats V1.47
2102 Added 'Message too big' to the list of mail rejection
2103 reasons (thanks to Marco Gaiarin).
2105 SC/06 Eximstats V1.48
2106 Mainlog lines which have GMT offsets and are too short to
2107 have a flag are now skipped.
2109 SC/07 Eximstats V1.49 (Alain Williams)
2110 Added the -emptyok flag.
2112 SC/08 Eximstats V1.50
2113 Fixes for obtaining the IP address from reject messages.
2115 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2116 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2117 whitesspace changes from 4.64-PH/27
2119 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2122 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2123 are found, allow negative numbers in numeric criteria)
2125 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2127 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2128 to match changes made in 4.64-PH/43
2130 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2132 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2133 "domain" is actually an IP address, and force a failure. This locks out
2134 those revolvers/nameservers that support "A-for-A" lookups, in
2135 contravention of the specifications.
2137 PH/48 When a host name was looked up from an IP address, and the subsequent
2138 forward lookup of the name timed out, the host name was left in
2139 $sender_host_name, contrary to the specification.
2141 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2142 restricted to single-key lookups, Exim was not diagnosing an error if
2143 * or *@ was used with a query-style lookup.
2145 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2147 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2148 long ago, but noone interested enough thought of it. Let's just say that
2149 the "1.1" means that there are some new functions that weren't there at
2150 some point in the past.
2152 PH/51 Error processing for expansion failure of helo_data from an smtp
2153 transport during callout processing was broken.
2155 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2156 tested/used via the -bh/-bhc/-bs options.
2158 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2159 bug, fixed in subsequent PCRE releases).
2161 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2162 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2164 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2170 SC/01 Use a glob alias rather than an array ref in eximstats generated
2171 parser. This improves both readability and performance.
2173 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2174 Don't display local sender or destination tables in eximstats unless
2175 there is data to show.
2176 Added average volumes into the eximstats top table text output.
2178 SC/03 Collect data on the number of addresses (recipients) as well
2179 as the number of messages in eximstats.
2181 TF/01 Correct an error in the documentation for the redirect router. Exim
2182 does (usually) call initgroups() when daemonizing.
2184 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2185 with consistent privilege compared to when running as a daemon.
2187 TF/03 Note in the spec that $authenticated_id is not set for local
2188 submissions from trusted users.
2190 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2191 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2193 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2194 by adding some example configuration directives to the default
2195 configuration file. A little bit of work is required to uncomment the
2196 directives and define how usernames and passwords are checked, but
2197 there is now a framework to start from.
2199 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2200 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2201 without this. I don't know how relevant this is to other LDAP libraries.
2203 PH/02 Add the verb name to the "unknown ACL verb" error.
2205 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2207 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2209 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2210 directory not expanded when it should be if an expanded home directory
2211 was set for the address (which is overridden by the transport).
2213 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2216 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2217 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2218 because it is too late at that time, and has no effect.
2220 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2221 security issue with \' (bugzilla #107). I could not use the
2222 PQescapeStringConn() function, because it needs a PGconn value as one of
2225 PH/08 When testing addresses using -bt, indicate those final addresses that
2226 are duplicates that would not cause an additional delivery. At least one
2227 person was confused, thinking that -bt output corresponded to deliveries.
2228 (Suppressing duplicates isn't a good idea as you lose the information
2229 about possibly different redirections that led to the duplicates.)
2231 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2232 systems where poll() doesn't work, in particular OS X.
2234 PH/10 Added more information to debugging output for retry time not reached.
2236 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2237 operations in malware.c.
2239 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2242 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2243 syslog_duplication set false, log lines that would normally be written
2244 both the the main log and to the reject log were not written to syslog at
2247 PH/14 In the default configuration, change the use of "message" in ACL warn
2248 statements to "add_header".
2250 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2251 not followed by a command (e.g. "seen endif").
2253 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2254 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2257 PH/17 Added extra conditions to the default value of delay_warning_condition
2261 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2262 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2263 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2266 The Auto-Submitted: and various List- headers are standardised, whereas I
2267 don't think Precedence: ever was.
2269 PH/18 Refactored debugging code in route_finduser() to show more information,
2270 in particular, the error code if getpwnam() issues one.
2272 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2273 This is apparently needed in addition to the PH/07 change above to avoid
2274 any possible encoding problems.
2276 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2277 but not after initializing Perl.
2279 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2280 output them only if debugging. By default they are written stderr,
2281 apparently, which is not desirable.
2283 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2286 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2289 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2291 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2292 authenticated or an ident call has been made. Suppress the default
2293 values for $authenticated_id and $authenticated_sender (but permit -oMai
2294 and -oMas) when testing with -bh.
2296 PH/24 Re-jigged the order of the tests in the default configuration so that the
2297 tests for valid domains and recipients precede the DNS black list and CSA
2298 tests, on the grounds that those ones are more expensive.
2300 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2301 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2302 This bug exists in every version of Exim that I still have, right back to
2305 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2306 However, an attempt to turn on case-sensitivity in a regex key by
2307 including (?-i) didn't work because the subject string was already
2308 lowercased, and the effects were non-intuitive. It turns out that a
2309 one-line patch can be used to allow (?-i) to work as expected.
2315 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2316 other effects) broke the use of negated acl sub-conditions.
2318 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2321 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2322 "Deny" causes Exim to reject the incoming connection with a 554 error.
2323 Unfortunately, if there is a major crisis, such as a disk failure,
2324 tcp-wrappers gives "deny", whereas what one would like would be some
2325 kind of temporary error. A kludge has been added to help with this.
2326 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2327 554 error is used if errno is still zero or contains ENOENT (which occurs
2328 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2331 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2333 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2334 errors. Otherwise a message that provokes a temporary error (when other
2335 messages do not) can cause a whole host to time out.
2337 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2338 addresses were routed directly to files or pipes from a redirect router.
2339 File deliveries just didn't batch; pipe deliveries might have suffered
2342 PH/06 A failure to get a lock for a hints database would erroneously always say
2343 "Failed to get write lock", even when it was really a read lock.
2345 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2346 of 0600. This has been changed to use the value of the lockfile_mode
2347 option (which defaults to 0600).
2349 PH/08 Applied small patch from the Sieve maintainer.
2351 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2352 folder from quota calculations, a direct delivery into this folder messed
2353 up the contents of the maildirsize file. This was because the regex was
2354 used only to exclude .Trash (or whatever) when the size of the mailbox
2355 was calculated. There was no check that a delivery was happening into an
2356 excluded directory. This bug has been fixed by ignoring all quota
2357 processing for deliveries into excluded directories.
2359 PH/10 Added the maildirfolder_create_regex option to appendfile.
2365 PH/01 The code for finding all the local interface addresses on a FreeBSD
2366 system running IPv6 was broken. This may well have applied to all BSD
2367 systems, as well as to others that have similar system calls. The broken
2368 code found IPv4 interfaces correctly, but gave incorrect values for the
2369 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2370 that it would not match correctly against @[] and not recognize the IPv6
2373 PH/02 The ipliteral router was not recognizing addresses of the form user@
2374 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2376 PH/03 Added disable_ipv6.
2378 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2379 decoded form, because it is most often used to construct To: headers
2380 lines in autoreplies, and the decoded form may well be syntactically
2381 invalid. However, $reply_address has leading white space removed, and all
2382 newlines turned into spaces so that the autoreply transport does not
2385 PH/05 If group was specified without a user on a router, and no group or user
2386 was specified on a transport, the group from the router was ignored.
2388 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2389 for visible compile-time settings that can be used to change these
2390 numbers, for those that want even more. Backwards compatibility with old
2391 spool files has been maintained. However, going back to a previous Exim
2392 release will lost any variables that are in spool files.
2394 PH/07 Two small changes when running in the test harness: increase delay when
2395 passing a TCP/IP connection to a new process, in case the original
2396 process has to generate a bounce, and remove special handling of
2397 127.0.0.2 (sic), which is no longer necessary.
2399 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2400 be the same on different OS.
2402 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2405 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2406 whether --show-vars was specified or not
2408 JJ/02 exipick: Added support for new ACL variable spool format introduced
2411 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2412 syntactically invalid From: or Reply-to: line, and a filter used this to
2413 generate an autoreply, and therefore failed to obtain an address for the
2414 autoreply, Exim could try to deliver to a non-existent relative file
2415 name, causing unrelated and misleading errors. What now happens is that
2416 it logs this as a hard delivery error, but does not attempt to create a
2419 PH/11 The exinext utility has a -C option for testing purposes, but although
2420 the given file was scanned by exinext itself; it wasn't being passed on
2421 when Exim was called.
2423 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2424 an end-of-file indication when reading a command response.
2426 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2427 compiled. In many other places in Exim, IPv6 addresses are always
2428 recognized, so I have changed this. It also means that IPv4 domain
2429 literals of the form [IPV4:n.n.n.n] are now always recognized.
2431 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2432 used if the router is not running as root, for example, when verifying at
2433 ACL time, or when using -bh. The debugging output from this situation was
2434 non-existent - all you got was a failure to exec. I have made two
2437 (a) Failures to set uid/gid, the current directory, or a process leader
2438 in a subprocess such as that created by queryprogram now generate
2439 suitable debugging ouput when -d is set.
2441 (b) The queryprogram router detects when it is not running as root,
2442 outputs suitable debugging information if -d is set, and then runs
2443 the subprocess without attempting to change uid/gid.
2445 PH/15 Minor change to Makefile for building test_host (undocumented testing
2448 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2449 additional section of a DNS packet that returns MX or SRV records.
2450 Instead, it always explicitly searches for A/AAAA records. This avoids
2451 major problems that occur when a DNS server includes only records of one
2452 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2453 fixed another bug: if SRV records were looked up and the corresponding
2454 address records were *not* found in the additional section, the port
2455 values from the SRV records were lost.
2457 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2458 using the correct key (the original address) when searching the retry
2459 rules in order to find which one to use for generating the retry hint.
2461 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2462 adding the default one. Similarly, if it contains a Reply-To: header, the
2463 errors_reply_to option, if set, is not used.
2465 PH/19 When calculating a retry time, Exim used to measure the "time since
2466 failure" by looking at the "first failed" field in the retry record. Now
2467 it does not use this if it is later than than the arrival time of the
2468 message. Instead it uses the arrival time. This makes for better
2469 behaviour in cases where some deliveries succeed, thus re-setting the
2470 "first failed" field. An example is a quota failure for a huge message
2471 when small messages continue to be delivered. Without this change, the
2472 "time since failure" will always be short, possible causing more frequent
2473 delivery attempts for the huge message than are intended.
2474 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2476 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2477 $1, $2, $3) because the numerical variables can be reset during some
2478 expansion items (e.g. "match"), thereby losing the authentication data.
2480 PH/21 Make -bV show the size of off_t variables so that the test suite can
2481 decide whether to run tests for quotas > 2G.
2483 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2484 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2485 filecount value is greater than 2G or if a quota value is greater than 2G
2486 on a system where the size of off_t is not greater than 4, a panic error
2489 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2490 never match. The debug and -bh output now contains an explicit error
2491 message indicating a malformed IPv4 address or mask.
2493 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2494 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2495 PH/23 above applies.
2497 PH/25 Do not write to syslog when running in the test harness. The only
2498 occasion when this arises is a failure to open the main or panic logs
2499 (for which there is an explicit test).
2501 PH/26 Added the /no_tell option to "control=freeze".
2503 PH/27 If a host name lookup failed very early in a connection, for example, if
2504 the IP address matched host_lookup and the reverse lookup yielded a name
2505 that did not have a forward lookup, an error message of the form "no IP
2506 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2507 could be logged. Now it outputs the IP address instead of "NULL".
2509 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2510 allows the sender and the authenticated sender to be set when
2511 submitting a message from within Exim. Since child_open_exim() is
2512 documented for local_scan(), the new function should be too.
2514 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2515 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2516 results in an empty string is now treated as unset.
2518 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2520 PH/31 Added sender_verify_fail logging option.
2522 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2523 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2524 tidied the source and removed it altogether.
2526 PH/33 When a queue run was abandoned because the load average was too high, a
2527 log line was always written; now it is written only if the queue_run log
2528 selector is set. In addition, the log line for abandonment now contains
2529 information about the queue run such as the pid. This is always present
2530 in "start" and "stop" lines but was omitted from the "abandon" line.
2532 PH/34 Omit spaces between a header name and the colon in the error message that
2533 is given when verify = headers_syntax fails (if there are lots of them,
2534 the message gets confusing).
2536 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2537 names, as there are now some PTR records that contain slashes. This check
2538 is only to protect against broken name servers that fall over on strange
2539 characters, so the fact that it applies to all lookups doesn't matter.
2541 PH/36 Now that the new test suite is complete, we can remove some of the
2542 special code in Exim that was needed for the old test suite. For example,
2543 sorting DNS records because real resolvers return them in an arbitrary
2544 order. The new test suite's fake resolver always returns records in the
2547 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2548 bounces) except when queue_only is set, to avoid logging races between
2549 the different processes.
2551 PH/38 Panic-die if .include specifies a non-absolute path.
2553 PH/39 A tweak to the "H" retry rule from its user.
2555 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2556 a label. They prevented compilation on older perls.
2558 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2559 a warning to be raised on newish perls.
2561 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2562 on queue. Changes to match documented behaviour of showing count of
2563 messages matching specified criteria.
2565 PH/40 Changed the default ident timeout from 30s to 5s.
2567 PH/41 Added support for the use of login_cap features, on those BSD systems
2568 that have them, for controlling the resources used by pipe deliveries.
2570 PH/42 The content-scanning code uses fopen() to create files in which to put
2571 message data. Previously it was not paying any attention to the mode of
2572 the files. Exim runs with umask(0) because the rest of the code creates
2573 files with open(), and sets the required mode explicitly. Thus, these
2574 files were ending up world-writeable. This was not a big issue, because,
2575 being within the spool directory, they were not world-accessible. I have
2576 created a function called modefopen, which takes an additional mode
2577 argument. It sets umask(777), creates the file, chmods it to the required
2578 mode, then resets the umask. All the relevant calls to fopen() in the
2579 content scanning code have been changed to use this function.
2581 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2582 to 24 hours. This avoids potential overflow problems when processing G
2583 and H retry rules. I suspect nobody ever tinkers with this value.
2585 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2587 PH/45 When the plaintext authenticator is running as a client, the server's
2588 challenges are checked to ensure they are valid base64 strings. By
2589 default, the authentication attempt is cancelled if an invalid string is
2590 received. Setting client_ignore_invalid_base64 true ignores these errors.
2591 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2592 they are received. Thus, the responses can be made to depend on the
2593 challenges. If an invalid string is ignored, an empty string is placed in
2596 PH/46 Messages that are created by the autoreply transport now contains a
2597 References: header, in accordance with RFCs 2822 and 3834.
2599 PH/47 Added authenticated_sender_force to the smtp transport.
2601 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2603 PH/49 Installed latest patch from the Sieve maintainer.
2605 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2606 was also set, the appendfile transport was unnecessarily scanning a
2607 directory of message files (e.g. for maildir delivery) to find the count
2608 of files (along with the size), even though it did not need this
2609 information. It now does the scan only if it needs to find either the
2610 size of the count of files.
2612 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2614 PH/52 Two bugs concerned with error handling when the smtp transport is
2617 (i) Exim was not creating retry information for temporary errors given
2618 for individual recipients after the DATA command when the smtp transport
2619 was used in LMTP mode. This meant that they could be retried too
2620 frequently, and not timed out correctly.
2622 (ii) Exim was setting the flag that allows error details to be returned
2623 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2624 recipients that were returned after the DATA command.
2626 PH/53 This is related to PH/52, but is more general: for any failing address,
2627 when detailed error information was permitted to be returned to the
2628 sender, but the error was temporary, then after the final timeout, only
2629 "retry timeout exceeded" was returned. Now it returns the full error as
2630 well as "retry timeout exceeded".
2632 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2633 do this, and (what is worse) MTAs that accept it.
2635 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2636 will now be deprecated.
2638 PH/56 New os.c-cygwin from the Cygwin maintainer.
2640 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2641 formats (previously only available in exim formats via -bpr, -bpru,
2642 and -bpra. Now also available in native and exiqgrep formats)
2644 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2645 with very large, slow to parse queues
2647 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2649 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2651 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2652 responses to authentication challenges, though it was showing the
2653 challenges; (ii) I've removed the CR characters from the debug output for
2656 PH/58 Allow for the insertion of a newline as well as a space when a string
2657 is turned into more than one encoded-word during RFC 2047 encoding. The
2658 Sieve code now uses this.
2660 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2661 data_4xx, lost_connection, tls_required.
2663 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2664 message was being sent as an SMTP response.
2666 PH/61 Add -l and -k options to exicyclog.
2668 PH/62 When verifying, if an address was redirected to one new address, so that
2669 verification continued, and the new address failed or deferred after
2670 having set something in $address_data, the value of $address_data was not
2671 passed back to the ACL. This was different to the case when no
2672 redirection occurred. The value is now passed back in both cases.
2674 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2675 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2676 use login_cap.h, so on its own it isn't the distinguishing feature. The
2677 new name refers directly to the setclassresources() function.
2679 PH/65 Added configuration files for NetBSD3.
2681 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2683 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2684 is preferred over IPv4.
2686 PH/68 The bounce_return_message and bounce_return_body options were not being
2687 honoured for bounces generated during the reception of non-SMTP messages.
2688 In particular, this applied to messages rejected by the ACL. This bug has
2689 been fixed. However, if bounce_return_message is true and bounce_return_
2690 body is false, the headers that are returned for a non-SMTP message
2691 include only those that have been read before the error was detected.
2692 (In the case of an ACL rejection, they have all been read.)
2694 PH/69 The HTML version of the specification is now built in a directory called
2695 spec_html instead of spec.html, because the latter looks like a path with
2696 a MIME-type, and this confuses some software.
2698 PH/70 Catch two compiler warnings in sieve.c.
2700 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2701 function verify_get_ident() calls ip_connect() to connect a socket, but
2702 if the "connect()" function timed out, ip_connect() used to close the
2703 socket. However, verify_get_ident() also closes the socket later, and in
2704 between Exim writes to the log, which may get opened at this point. When
2705 the socket was closed in ip_connect(), the log could get the same file
2706 descriptor number as the socket. This naturally causes chaos. The fix is
2707 not to close the socket in ip_connect(); the socket should be closed by
2708 the function that creates it. There was only one place in the code where
2709 this was missing, in the iplookup router, which I don't think anybody now
2710 uses, but I've fixed it anyway.
2712 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2713 well as to direct DNS lookups. Otherwise the handling of names in host
2714 lists is inconsistent and therefore confusing.
2720 PH/01 Two changes to the default runtime configuration:
2722 (1) Move the checks for relay_from_hosts and authenticated clients from
2723 after to before the (commented out) DNS black list checks.
2725 (2) Add control=submission to the relay_from_hosts and authenticated
2726 clients checks, on the grounds that messages accepted by these
2727 statements are most likely to be submissions.
2729 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2731 (1) Generate an error if the third argument for the ${prvs expansion is
2734 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2737 (3) Reset the variables that are obtained from the first argument of
2738 ${prvscheck and used in the second argument before leaving the code,
2739 because their memory is reclaimed, so using them afterwards may do
2742 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2743 one (it's much easier than Tom thought :-).
2745 (5) Because of (4), we can now allow for the use of $prvscheck_result
2746 inside the third argument.
2748 PH/03 For some reason, the default setting of PATH when running a command from
2749 a pipe transport was just "/usr/bin". I have changed it to
2752 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2753 anything to be listed in the output from -bV.
2755 PH/05 When a filter generated an autoreply, the entire To: header line was
2756 quoted in the delivery log line, like this:
2758 => >A.N.Other <ano@some.domain> <original@ddress> ...
2760 This has been changed so that it extracts the operative address. There
2761 may be more than one such address. If so, they are comma-separated, like
2764 => >ano@some.domain,ona@other.domain <original@ddress> ...
2766 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2767 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2768 not being looked up in the rDNS to get a host name, Exim was showing the
2769 IP address twice in Received: lines, even though the IP addresses were
2770 identical. For example:
2772 Received: from [1.2.3.4] (helo=[1.2.3.4])
2774 However, if the real host name was known, it was omitting the HELO data
2775 if it matched the actual IP address. This has been tidied up so that it
2776 doesn't show the same IP address twice.
2778 PH/07 When both +timestamp and +memory debugging was on, the value given by
2779 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2780 called by the debug printing, thereby overwriting the timestamp buffer.
2781 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2783 PH/08 When the original message was included in an autoreply transport, it
2784 always said "this is a copy of the message, including all the headers",
2785 even if body_only or headers_only was set. It now gives an appropriate
2788 PH/09 Applied a patch from the Sieve maintainer which:
2790 o fixes some comments
2791 o adds the (disabled) notify extension core
2792 o adds some debug output for the result of if/elsif tests
2793 o points to the current vacation draft in the documentation
2794 and documents the missing references header update
2798 o fixes a bug in processing the envelope test (when testing
2799 multiple envelope elements, the last element determinted the
2802 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2803 Electronic Mail") by including:
2805 Auto-submitted: auto-generated
2807 in the messages that it generates (bounce messages and others, such as
2808 warnings). In the case of bounce messages for non-SMTP mesages, there was
2809 also a typo: it was using "Auto_submitted" (underscore instead of
2810 hyphen). Since every message generated by Exim is necessarily in response
2811 to another message, thes have all been changed to:
2813 Auto-Submitted: auto-replied
2815 in accordance with these statements in the RFC:
2817 The auto-replied keyword:
2819 - SHOULD be used on messages sent in direct response to another
2820 message by an automatic process,
2822 - MUST NOT be used on manually-generated messages,
2824 - MAY be used on Delivery Status Notifications (DSNs) and Message
2825 Disposition Notifications (MDNs),
2827 - MUST NOT be used on messages generated by automatic or periodic
2828 processes, except for messages which are automatic responses to
2831 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2832 to the default Received: header definition.
2834 PH/12 Added log selector acl_warn_skipped (default on).
2836 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2837 variables because (a) they are in the wrong storage pool and (b) even if
2838 they were copied, it wouldn't work properly because of the caching.
2840 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2841 checking when decoding. Apparently there are clients that generate
2842 overlong encoded strings. Why am I not surprised?
2844 PH/15 If the first argument of "${if match_address" was not empty, but did not
2845 contain an "@" character, Exim crashed. Now it writes a panic log message
2846 and treats the condition as false.
2848 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2850 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2851 extension "envelope-auth". The code is finished and in agreement with
2852 other implementations, but there is no documentation so far and in fact,
2853 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2854 not changing the active code.
2856 Print executed "if" and "elsif" statements when debugging is used. This
2857 helps a great deal to understand what a filter does.
2859 Document more things not specified clearly in RFC3028. I had all this
2860 sorted out, when out of a sudden new issues came to my mind. Oops."
2862 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2865 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2866 canonical form (as documented). However, after a host name lookup from
2867 the IP address, check_host() was doing a simple string comparison with
2868 addresses acquired from the DNS when checking that the found name did
2869 have the original IP as one of its addresses. Since any found IPv6
2870 addresses are likely to be in abbreviated form, the comparison could
2871 fail. Luckily, there already exists a function for doing the comparison
2872 by converting both addresses to binary, so now that is used instead of
2873 the text comparison.
2875 PH/20 There was another similar case to PH/19, when a complete host name was
2876 given in a host list; looking up its IP address could give an abbreviated
2877 form, whereas the current host's name might or might not be abbreviated.
2878 The same fix has been applied.
2884 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2885 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2888 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2889 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2891 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2893 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2894 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2895 Also, for "Auto-Submitted", treat anything other than "no" as
2896 non-personal, in accordance with RFC 3834. (Previously it treated
2897 anything starting "auto-" as non-personal.)
2899 TF/01 The control=submission/name=... option had a problem with syntax
2900 errors if the name included a slash character. The /name= option
2901 now slurps the rest of the string, so it can include any characters
2902 but it must come last in the list of options (after /sender_retain
2905 PH/05 Some modifications to the interface to the fake nameserver for the new
2913 TK/01 Added the "success_on_redirect" address verification option. See
2914 NewStuff for rationale and an example.
2916 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2918 PH/02 Patch to exigrep to allow it to work on syslog lines.
2920 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2921 fread() to skip over the body file's header line, because in Cygwin the
2922 header line is locked and is inaccessible.
2924 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2925 co-exist for some time) to make it clear that it is the Exim ID that is
2926 referenced, not the Message-ID: header line.
2928 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2929 string_format() function, because snprintf() does not exist on all
2932 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2933 ${stat: expansion item.
2935 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2936 protocol synchronization error", to keep the pedants happy.
2938 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2939 well as for IRIX systems, when gcc is being used. See the host.c source
2942 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2944 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2947 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2948 transport and to the smtp transport in LMTP mode.
2950 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2952 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2953 run a filter in a subprocess. This could lead to confusion in subsequent
2954 lookups in the parent process. There should also be a search_tidyup() at
2955 the end of the subprocess.
2957 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2958 only if the host matched helo_try_verify_hosts, which caused the
2959 verification to occur when the EHLO/HELO command was issued. The ACL just
2960 tested the remembered result. Now, if a previous verification attempt has
2961 not happened, "verify = helo" does it there and then.
2963 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2965 TK/03 Fix log output including CR from clamd.
2967 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2968 exist provoked a memory error which could cause a segfault.
2970 PH/15 Installed PCRE 6.2
2972 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2974 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2975 of the problem. Specifically, suggested +O2 rather than +O1 for the
2978 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2980 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2981 record, the port was not correctly logged when the outgoing_port log
2982 selector was set (it logged the transort's default port).
2984 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2985 fallback_hosts, and "hosts" in the smtp transport.
2987 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2988 host errors such as "Connection refused".
2990 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2991 authentication with radiusclient 0.4.9:
2993 - Error returned from rc_read_config was caught wrongly
2994 - Username/password not passed on to radius server due to wrong length.
2996 The presumption is that some radiusclient API changes for 4.51/PH/17
2997 were not taken care of correctly. The code is still untested by me (my
2998 Linux distribution still has 0.3.2 of radiusclient), but it was
2999 contributed by a Radius user.
3001 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3002 expanding the "port" option of the smtp transport.
3004 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3005 while reading a MIME header. Thanks to Tom Hughes for a patch.
3007 PH/24 Include config.h inside local_scan.h so that configuration settings are
3010 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3011 that in an ACL for RCPT (for example), you can examine exactly what was
3014 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3015 commands, but it was not correctly comparing the address with the actual
3016 client host address. Thus, it would show the EHLO address in Received:
3017 header lines when this was not necessary.
3019 PH/27 Added the % operator to ${eval:}.
3021 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3022 it should be ignoring failures (because with -C, for example, it has lost
3023 privilege). It wasn't ignoring creation failures other than "already
3026 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3029 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3030 because an input line was too long, either on its own, or by virtue of
3031 too many continuations, the temporary file was not being removed, and the
3032 return code was incorrect.
3034 PH/31 Missing "BOOL" in function definition in filtertest.c.
3036 PH/32 Applied Sieve patches from the maintainer.
3038 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3040 PH/33 Added "verify = not_blind".
3042 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3043 Local/Makefile (with some defaults set). These are used in built scripts
3044 such as exicyclog, but they have never been used in the exim_install
3045 script (though there are many overriding facilities there). I have
3046 arranged that the exim_install script now takes note of these two
3049 PH/35 Installed configuration files for Dragonfly.
3051 PH/36 When a locally submitted message by a trusted user did not contain a
3052 From: header, and the sender address was obtained from -f or from an SMTP
3053 MAIL command, and the trusted user did not use -F to supply a sender
3054 name, $originator_name was incorrectly used when constructing a From:
3055 header. Furthermore, $originator_name was used for submission mode
3056 messages from external hosts without From: headers in a similar way,
3057 which is clearly wrong.
3059 PH/37 Added control=suppress_local_fixups.
3061 PH/38 When log_selector = +received_sender was set, and the addition of the
3062 sender made the log line's construction buffer exactly full, or one byte
3063 less than full, an overflow happened when the terminating "\n" was
3066 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3067 when the result of a list match is failure because a DNS lookup failed.
3069 PH/40 RM_COMMAND is now used in the building process.
3071 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3072 the "build-* directories that it finds.
3074 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3075 address in a domain literal was a prefix of an interface address.
3077 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3078 when verifying a sender address, unless rewrite_headers is false.
3080 PH/44 Wrote a long comment about why errors_to addresses are verified as
3081 recipients, not senders.
3083 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3084 the ratelimit ACL was added.
3086 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3088 PH/46 Added extra information about PostgreSQL errors to the error string.
3090 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3091 avoiding the need to install special zones in a real server. This is
3092 backwards compatible; if it can't find the fake resolver, it drops back.
3093 Thus, both old and new test suites can be run.
3095 TF/02 Added util/ratelimit.pl
3097 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3098 clock is set back in time.
3100 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3101 Candler <B.Candler@pobox.com>.
3103 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3104 OK for addresses that are the result of redirections.
3106 PH/48 A number of further additions for the benefit of the new test suite,
3107 including a fake gethostbyname() that interfaces to the fake DNS resolver
3110 TF/06 The fix for widen_domains has also been applied to qualify_single and
3111 search_parents which are the other dnslookup options that can cause
3114 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3117 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3119 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3120 service" if it encountered an MX record with an empty target hostname.
3121 The message is now "an MX or SRV record indicated no SMTP service".
3123 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3124 if the DNS of the sending site is misconfigured. This is quite a
3125 common situation. This change restores the behaviour of treating a
3126 helo verification defer as a failure.
3128 PH/51 If self=fail was set on a router, the bounce message did not include the
3129 actual error message.
3135 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3137 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3138 command itself ended in error, the underlying message about the transport
3139 filter timeout was being overwritten with the pipe command error. Now the
3140 underlying error message should be appended to the second error message.
3142 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3144 PH/02 Reduce the amount of output that "make" produces by default. Full output
3145 can still be requested.
3147 PH/03 The warning log line about a condition test deferring for a "warn" verb
3148 was being output only once per connection, rather than after each
3149 occurrence (because it was using the same function as for successful
3150 "warn" verbs). This seems wrong, so I have changed it.
3152 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3153 it should not have, which might have caused a crash in the right
3154 circumstances, but probably never did.
3156 PH/04 Installed a modified version of Tony Finch's patch to make submission
3157 mode fix the return path as well as the Sender: header line, and to
3158 add a /name= option so that you can make the user's friendly name appear
3161 TF/03 Added the control = fakedefer ACL modifier.
3163 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3164 Mark Lowes for thorough testing.
3166 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3168 TK/03 Merged latest SRS patch from Miles Wilton.
3170 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3171 with the definition in sysexits.h (which is #included earlier).
3172 Fortunately, Exim does not actually use EX_OK. The code used to try to
3173 preserve the sysexits.h value, by assumimg that macro definitions were
3174 scanned for macro replacements. I have been disabused of this notion,
3175 so now the code just undefines EX_OK before #including unistd.h.
3177 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3178 in the smtp transport. When a block could not be written in a single
3179 write() function, the timeout was being re-applied to each part-write.
3180 This seems wrong - if the receiver was accepting one byte at a time it
3181 would take for ever. The timeout is now adjusted when this happens. It
3182 doesn't have to be particularly precise.
3184 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3185 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3187 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3188 without a domain if the check to <postmaster@domain> fails.
3190 SC/01 Eximstats: added -xls and the ability to specify output files
3191 (patch written by Frank Heydlauf).
3193 SC/02 Eximstats: use FileHandles for outputing results.
3195 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3197 SC/04 Eximstats: fixed display of large numbers with -nvr option
3199 SC/05 Eximstats: fixed merging of reports with empty tables.
3201 SC/06 Eximstats: added the -include_original_destination flag
3203 SC/07 Eximstats: removed tabs and trailing whitespace.
3205 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3207 TK/06 MBOX spool code: Add real "From " MBOX separator line
3208 so the .eml file is really in mbox format (even though
3209 most programs do not really care). Patch from Alex Miller.
3211 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3212 The latter is generated from $received_to and is only set if the
3213 message has one envelope recipient. SA can use these headers,
3214 obviously out-of-the-box. Patch from Alex Miller.
3216 PH/08 The ${def test on a variable was returning false if the variable's
3217 value was "0", contrary to what the specification has always said!
3218 The result should be true unless the variable is empty.
3220 PH/09 The syntax error of a character other than { following "${if
3221 def:variable_name" (after optional whitespace) was not being diagnosed.
3222 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3223 accidental colon was present, for example, could give incorrect results.
3225 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3226 result is used (not including appendfile, where other changes are about
3229 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3230 This involved changing a lot of size variables from int to off_t. It
3231 should work with maildirs and everything.
3233 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3234 spamd dying while we are connected to it.
3236 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3239 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3240 function for BDB 4.3.
3242 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3244 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3245 the sources for PCRE have been re-arranged and more files are now
3248 PH/15 The code I had for printing potentially long long variables in PH/11
3249 above was not the best (it lost precision). The length of off_t variables
3250 is now inspected at build time, and an appropriate printing format (%ld
3251 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3252 to be "long long int" or "long int". This is needed for the internal
3253 formatting function string_vformat().
3255 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3256 the configuration file to be ":syslog", then the script "guesses" where
3257 the logs files are, rather than using the compiled in default. In our
3258 case the guess is not the same as the compiled default, so the script
3259 suddenly stopped working when I started to use syslog. The patch checks
3260 to see if log_file_path is "". If so, it attempts to read it from exim
3261 with no configuration file to get the compiled in version, before it
3262 falls back to the previous guessing code."
3264 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3265 implementing BATV in an Exim configuration. See NewStuff for the gory
3268 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3269 Makefile that are specific to HP-UX.
3271 PH/18 If the "use_postmaster" option was set for a recipient callout together
3272 with the "random" option, the postmaster address was used as the MAIL
3273 FROM address for the random test, but not for the subsequent recipient
3274 test. It is now used for both.
3276 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3277 patch removes a few documentation additions to RFC 3028, because the
3278 latest draft now contains them. It adds the new en;ascii-case comparator
3279 and a new error check for 8bit text in MIME parts. Comparator and
3280 require names are now matched exactly. I enabled the subaddress
3281 extension, but it is not well tested yet (read: it works for me)."
3283 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3284 rework some of the code of TK/09 above to avoid the hardwired use of
3285 "%lld" and "long long". Replaced the call to snprintf() with a call to
3288 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3289 records point to non-existent hosts", "retry timeout exceeded", and
3290 "retry time not reached for any host after a long failure period".
3292 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3293 experimental DomainKeys support:
3295 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3296 (2) On an error such as an illegally used "control", the wrong name for
3297 the control was given.
3299 These problems did NOT occur unless DomainKeys support was compiled.
3301 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3303 PH/24 Added ${if match_ip condition.
3305 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3306 databases so that it will be absolutely obvious if a crash occurs in the
3307 DB library. This is a regular occurrence (often caused by mis-matched
3310 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3311 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3312 picked up on a user's system that detects such things. There doesn't seem
3313 to be a gcc warning option for this - only an attribute that has to be
3314 put on the function's prototype. It seems that in Fedora Core 4 they have
3315 set this on a number of new functions. No doubt there will be more in due
3318 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3319 specify a transport. However, if an address that was verified by such a
3320 router was the subject of a callout, Exim crashed because it tried to
3321 read the rcpt_include_affixes from the non-existent transport. Now it
3322 just assumes that the setting of that option is false. This bug was
3323 introduced by 4.51/PH/31.
3325 PH/28 Changed -d+all to exclude +memory, because that information is very
3326 rarely of interest, but it makes the output a lot bigger. People tend to
3327 do -d+all out of habit.
3329 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3330 code in os-type was giving problems when libc.so lives in lib64, like on
3333 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3334 aren't the modern standard, and it seems that some systems' include files
3335 don't always have them. Exim was already checking for some of the newer
3336 ones like T_AAAA, and defining it itself. I've added checks for all the
3337 record types that Exim uses.
3339 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3340 not automatically generating a new one, as it is supposed to. This
3341 prevented TLS from working. If the file did exist, but contained invalid
3342 data, a new version was generated, as expected. It was only the case of a
3343 non-existent file that was broken.
3345 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3346 with a change in libdomainkeys > 0.64.
3348 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3349 from DNS. If the selector record carries the flag, it now has
3350 precedence over the domain-wide flag.
3352 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3354 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3355 the use of a port name that isn't defined in /etc/services) occurred, the
3356 message was deferred as in a normal delivery, and thus remained on the
3357 spool, instead of being failed because of the mua_wrapper setting. This
3358 is now fixed, and I tidied up some of the mua_wrapper messages at the
3361 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3362 the messages in a hash of arrays rather than using individual hashes.
3363 This is a bit cleaner and results in dramatic memory savings, albeit
3364 at a slight CPU cost.
3366 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3367 as requested by Marc Sherman.
3369 SC/10 Eximstats: added histograms for user specified patterns as requested
3372 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3374 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3375 fopen() in the content-scanning modules that did not already have it.
3381 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3382 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3384 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3386 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3388 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3389 to test Sieve filters that use "vacation".
3391 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3392 that changes the way the GnuTLS parameters are stored in the cache file.
3393 The new format can be generated externally. For backward compatibility,
3394 if the data in the cache doesn't make sense, Exim assumes it has read an
3395 old-format file, and it generates new data and writes a new file. This
3396 means that you can't go back to an older release without removing the
3399 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3400 work if there are any delivery delays because "one_time" forces the
3401 parent to be marked "delivered", so its unseen clone is never tried
3402 again. For this reason, Exim now forbids the simultaneous setting of
3405 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3406 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3407 entry if you want to know the details. The fix, however, neglected to
3408 consider the case when local delivery batching is involved. The test for
3409 "previously delivered" was not happening when checking to see if an
3410 address could be batched with a previous (undelivered) one; under
3411 certain circumstances this could lead to multiple deliveries to the same
3414 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3415 in its include files, and this causes problems building Exim.
3417 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3418 header_syntax) but Exim was just ignoring anything given after a slash.
3419 In particular, this caused confusion with an attempt to use "verify =
3420 reverse_host_lookup/defer_ok". An error is now given when options are
3421 supplied for verify items that do not have them. (Maybe reverse_host_
3422 lookup should have a defer_ok option, but that's a different point.)
3424 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3425 defined by RFC 821) to 2048, because there were problems with some AUTH
3426 commands, and RFC 1869 says the size should be increased for extended
3427 SMTP commands that take arguments.
3429 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3432 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3433 "unknown" error; now it says that the functionality isn't in the binary.
3435 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3436 an address' error message when a string expansion fails (syntax or
3437 whatever). Otherwise the password may appear in the log. Following change
3438 PH/42 below, there is no longer a chance of it appearing in a bounce
3441 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3443 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3444 of its addresses was ever tried. (Bugzilla bug #2.)
3446 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3447 the result incorrectly in the debug output. (It correctly added a newline
3448 to what was transported.)
3450 TF/01 Added $received_time.
3452 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3453 commented out examples of how to interface to a virus scanner and to
3454 SpamAssassin. Also added commented examples of av_scanner and
3455 spamd_address settings.
3457 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3458 and controls are allowed in which ACLs. There were a couple of minor
3459 errors. Some of the entries in the conditions table (which is a table of
3460 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3461 negation of where the condition IS allowed.
3463 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3465 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3466 header file does not have a version number, so I've had to invent a new
3467 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3468 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3469 radiusclient), but it was contributed by a Radius user.
3471 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3472 files or directories, for OpenSSL.
3474 PH/19 When an Exim process that is running as root has to create an Exim log
3475 file, it does so in a subprocess that runs as exim:exim so as to get the
3476 ownership right at creation (otherwise, other Exim processes might see
3477 the file with the wrong ownership). There was no test for failure of this
3478 fork() call, which would lead to the process getting stuck as it waited
3479 for a non-existent subprocess. Forks do occasionally fail when resources
3480 run out. I reviewed all the other calls to fork(); they all seem to check
3483 PH/20 When checking for unexpected SMTP input at connect time (before writing
3484 the banner), Exim was not dealing correctly with a non-positive return
3485 from the read() function. If the client had disconnected by this time,
3486 the result was a log entry for a synchronization error with an empty
3487 string after "input=" when read() returned zero. If read() returned -1
3488 (an event I could not check), uninitialized data bytes were printed.
3489 There were reports of junk text (parts of files, etc) appearing after
3492 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3494 PH/22 Added support for macro redefinition, and (re)definition in between
3495 driver and ACL definitions.
3497 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3498 forgetting to use the resulting value; it was using the unexpanded value.
3500 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3501 hadn't been configured. The fix is from Juergen Kreileder, who
3502 understands it better than I do:
3504 "Here's what I see happening with three configured cyrus_sasl
3505 authenticators configured (plain, login, cram-md5):
3507 On startup auth_cyrus_sasl_init() gets called for each of these.
3508 This means three calls to sasl_listmech() without a specified mech_list.
3509 => SASL tests which mechs of all available mechs actually work
3510 => three warnings about OTP not working
3511 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3513 With the patch, sasl_listmech() also gets called three times. But now
3514 SASL's mech_list option is set to the server_mech specified in the the
3515 authenticator. Or in other words, the answer from sasl_listmech()
3516 gets limited to just the mech you're testing for (which is different
3518 => the return list contains just 'plain' or 'login', 'cram-md5' or
3519 nothing depending on the value of ob->server_mech.
3521 I've just tested the patch: Authentication still works fine,
3522 unavailable mechs specified in the exim configuration are still
3523 caught, and the auth.log warnings about OTP are gone."
3525 PH/25 When debugging is enabled, the contents of the command line are added
3526 to the debugging output, even when log_selector=+arguments is not
3529 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3530 answer is "GNU", and only if the return is "GNU/something" is the answer
3533 PH/27 $acl_verify_message is now set immediately after the failure of a
3534 verification in an ACL, and so is available in subsequent modifiers. In
3535 particular, the message can be preserved by coding like this:
3537 warn !verify = sender
3538 set acl_m0 = $acl_verify_message
3540 Previously, $acl_verify_message was set only while expanding "message"
3541 and "log_message" when a very denied access.
3543 PH/28 Modified OS/os.c-Linux with
3545 -#ifndef OS_LOAD_AVERAGE
3546 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3548 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3549 nomenclature these days.)
3551 PH/29 Installed patch from the Sieve maintainer that adds the options
3552 sieve_useraddress and sieve_subaddress to the redirect router.
3554 PH/30 In these circumstances:
3555 . Two addresses routed to the same list of hosts;
3556 . First host does not offer TLS;
3557 . First host accepts first address;
3558 . First host gives temporary error to second address;
3559 . Second host offers TLS and a TLS session is established;
3560 . Second host accepts second address.
3561 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3562 and peerdn, if requested) that were in fact used only for the second
3565 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3566 attention to any local part prefix or suffix that was matched by the
3567 router that accepted the address. It now behaves in the same way as it
3568 does for delivery: the affixes are removed from the local part unless
3569 rcpt_include_affixes is set on the transport.
3571 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3572 timeout during the DATA phase of an incoming message.
3574 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3575 applied a patch sanctioned by the Sieve maintainer.
3577 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3578 the uid or gid is negative. A case of a negative gid caused this to be
3579 noticed. The fix allows for either to be negative.
3581 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3582 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3585 PH/36 The change PH/12 above was broken. Fixed it.
3587 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3588 the grounds that routing the same address twice would always produce the
3589 same answer. This might have been true once, but it is certainly no
3590 longer true now. Routing a child address may depend on the previous
3591 routing that produced that child. Some complicated redirection strategies
3592 went wrong when messages had multiple recipients, and made Exim's
3593 behaviour dependent on the order in which the addresses were given.
3595 I have moved the duplicate checking until after the routing is complete.
3596 Exim scans the addresses that are assigned to local and remote
3597 transports, and removes any duplicates. This means that more work will be
3598 done, as duplicates will always all be routed, but duplicates are
3599 presumably rare, so I don't expect this is of any significance.
3601 For deliveries to pipes, files, and autoreplies, the duplicate checking
3602 still happens during the routing process, since they are not going to be
3605 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3606 It corrects a timeout issue with spamd. This is Ian's comment: "The
3607 background is that sometimes spamd either never reads data from a
3608 connection it has accepted, or it never writes response data. The exiscan
3609 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3610 blindly assumes that writes won't block so it may never time out."
3612 PH/39 Allow G after quota size as well as K and M.
3614 PH/40 The value set for $authenticated_id in an authenticator may not contain
3615 binary zeroes or newlines because the value is written to log lines and
3616 to spool files. There was no check on this. Now the value is run through
3617 the string_printing() function so that such characters are converted to
3618 printable escape sequences.
3620 PH/41 $message_linecount is a new variable that contains the total number of
3621 lines in the message. Compare $body_linecount, which is the count for the
3624 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3625 bounce and delay warning messages, except in certain special cases, which
3628 (a) An SMTP error message from a remote host;
3629 (b) A message specified in a :fail: redirection;
3630 (c) A message specified in a "fail" command in a system filter;
3631 (d) A message specified in a FAIL return from the queryprogram router;
3632 (e) A message specified by the cannot_route_message router option.
3634 In these cases only, Exim does include the error details in bounce and
3635 warning messages. There are also a few cases where bland messages such
3636 as "unrouteable address" or "local delivery error" are given.
3638 PH/43 $value is now also set for the "else" part of a ${run expansion.
3640 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3641 being worked on, but at least Exim now implements the latest version to
3644 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3645 process to complete was treated as a delivery failure, a timeout while
3646 writing the message to the pipe was logged, but erroneously treated as a
3647 successful delivery. Such timeouts include transport filter timeouts. For
3648 consistency with the overall process timeout, these timeouts are now
3649 treated as errors, giving rise to delivery failures by default. However,
3650 there is now a new Boolean option for the pipe transport called
3651 timeout_defer, which, if set TRUE, converts the failures into defers for
3652 both kinds of timeout. A transport filter timeout is now identified in
3655 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3656 systems where "make" and "gmake" are different, calling "gmake" at top
3657 level broke things. I've arranged for the value of $(MAKE) to be passed
3658 from the Makefile to this script so that it can call the same version of
3662 A note about Exim versions 4.44 and 4.50
3663 ----------------------------------------
3665 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3666 changes of various kinds. As a consequence, a big documentation update was
3667 needed. This delayed the release for rather longer than seemed good, especially
3668 in the light of a couple of (minor) security issues. Therefore, the changes
3669 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3670 release. So 4.44 and 4.50 are in effect two different branches that both start
3673 I have left the 4.50 change log unchanged; it contains all the changes since
3674 4.43. The change log for 4.44 is below; many of its items are identical to
3675 those for 4.50. This seems to be the most sensible way to preserve the
3676 historical information.
3682 1. Minor wording change to the doc/README.SIEVE file.
3684 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3685 computation of the current number of files was incorrect.
3687 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3688 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3691 4. Give more explanation in the error message when the command for a transport
3692 filter fails to execute.
3694 5. There are several places where Exim runs a non-Exim command in a
3695 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3696 was being done only for the command run by the queryprogram router. It is
3697 now done for all such subprocesses. The other cases are: ${run, transport
3698 filters, and the commands run by the lmtp and pipe transports.
3700 6. Added CONFIGURE_GROUP build-time option.
3702 7. Some older OS have a limit of 256 on the maximum number of file
3703 descriptors. Exim was using setrlimit() to set 1000 as a large value
3704 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3705 systems. I've change it so that if it can't get 1000, it tries for 256.
3707 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3708 was an oversight, and furthermore, ever since the addition of extra
3709 controls (e.g. 4.43/32), the checks on when to allow different forms of
3710 "control" were broken. There should now be diagnostics for all cases when a
3711 control that does not make sense is encountered.
3713 9. Added the /retain_sender option to "control=submission".
3715 10. $recipients is now available in the predata ACL (oversight).
3717 11. Tidy the search cache before the fork to do a delivery from a message
3718 received from the command line. Otherwise the child will trigger a lookup
3719 failure and thereby defer the delivery if it tries to use (for example) a
3720 cached ldap connection that the parent has called unbind on.
3722 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3723 of $address_data from the recipient verification was clobbered by the
3724 sender verification.
3726 13. The value of address_data from a sender verification is now available in
3727 $sender_address_data in subsequent conditions in the ACL statement.
3729 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3731 15. Added a new option "connect=<time>" to callout options, to set a different
3734 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3735 was its contents. (It was OK if the option was not defined at all.)
3737 17. A "Completed" log line is now written for messages that are removed from
3738 the spool by the -Mrm option.
3740 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3741 information about exactly what failed.
3743 19. Added -dd to debug only the daemon process.
3745 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3746 handles timeouts, both on the server side and network timeouts. Renamed the
3747 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3749 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3750 It is now set to "smtps".
3752 22. $host_address is now set to the target address during the checking of
3753 ignore_target_hosts.
3755 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3756 being passed; this would have caused $sender_host_name to have been used if
3757 matching the list had actually called for a host name (not very likely,
3758 since this list is usually IP addresses). A host name is now passed as
3761 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3762 code that specifies a non-restarting handler (typically sigaction() in
3763 modern systems) in an attempt to fix a rare and obscure crash bug.
3765 25. Narrowed the window for a race in the daemon that could cause it to ignore
3766 SIGCHLD signals. This is not a major problem, because they are used only to
3767 wake it up if nothing else does.
3769 26. A malformed maildirsize file could cause Exim to calculate negative values
3770 for the mailbox size or file count. Odd effects could occur as a result.
3771 The maildirsize information is now recalculated if the size or filecount
3774 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3775 support for a long time. Removed HAVE_SYS_VFS_H.
3777 28. Installed the latest version of exipick from John Jetmore.
3779 29. In an address list, if the pattern was not a regular expression, an empty
3780 subject address (from a bounce message) matched only if the pattern was an
3781 empty string. Non-empty patterns were not even tested. This was the wrong
3782 because it is perfectly reasonable to use an empty address as part of a
3783 database query. An empty address is now tested by patterns that are
3784 lookups. However, all the other forms of pattern expect the subject to
3785 contain a local part and a domain, and therefore, for them, an empty
3786 address still always fails if the pattern is not itself empty.
3788 30. Exim went into a mad DNS loop when attempting to do a callout where the
3789 host was specified on an smtp transport, and looking it up yielded more
3790 than one IP address.
3792 31. Re-factored the code for checking spool and log partition space into a
3793 function that finds that data and another that does the check. The former
3794 is then used to implement four new variables: $spool_space, $log_space,
3795 $spool_inodes, and $log_inodes.
3797 32. The RFC2047 encoding function was originally intended for short strings
3798 such as real names; it was not keeping to the 75-character limit for
3799 encoded words that the RFC imposes. It now respects the limit, and
3800 generates multiple encoded words if necessary. To be on the safe side, I
3801 have increased the buffer size for the ${rfc2047: expansion operator from
3804 33. It is now permitted to omit both strings after an "if" condition; if the
3805 condition is true, the result is "true". As before, when the second string
3806 is omitted, a false condition yields an empty string. This makes it less
3807 cumbersome to write custom ACL and router conditions.
3809 34. Failure to deliver a bounce message always caused it to be frozen, even if
3810 there was an errors_to setting on the router. The errors_to setting is now
3813 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3814 canonical form (fully expanded) before being placed in
3815 $sender_host_address.
3817 36. The table in the code that translates DNS record types into text (T_A to
3818 "A" for instance) was missing entries for NS and CNAME. It is just possible
3819 that this could have caused confusion if both these types were looked up
3820 for the same domain, because the text type is used as part of Exim's
3821 per-process caching. But the chance of anyone hitting this buglet seems
3824 37. The dnsdb lookup has been extended in a number of ways.
3826 (1) There is a new type, "zns", which walks up the domain tree until it
3827 finds some nameserver records. It should be used with care.
3829 (2) There is a new type, "mxh", which is like "mx" except that it returns
3830 just the host names, not the priorities.
3832 (3) It is now possible to give a list of domains (or IP addresses) to be
3833 looked up. The behaviour when one of the lookups defers can be
3834 controlled by a keyword.
3836 (4) It is now possible to specify the separator character for use when
3837 multiple records are returned.
3839 38. The dnslists ACL condition has been extended: it is now possible to supply
3840 a list of IP addresses and/or domains to be looked up in a particular DNS
3843 39. Added log_selector=+queue_time_overall.
3845 40. When running the queue in the test harness, wait just a tad after forking a
3846 delivery process, to get repeatability of debugging output.
3848 41. Include certificate and key file names in error message when GnuTLS fails
3849 to set them up, because the GnuTLS error message doesn't include the name
3850 of the failing file when there is a problem reading it.
3852 42. Allow both -bf and -bF in the same test run.
3854 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3856 44. The "Exiscan patch" is now merged into the mainline Exim source.
3858 45. Sometimes the final signoff response after QUIT could fail to get
3859 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3860 < 0 before doing a fflush(). This bug looks as though it goes back to the
3861 introduction of TLS in release 3.20, but "sometimes" must have been rare
3862 because the tests only now provoked it.
3864 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3865 (this can affect the format of dates).
3867 47. exim_tidydb, when checking for the continued existence of a message for
3868 which it has found a message-specific retry record, was not finding
3869 messages that were in split spool directories. Consequently, it was
3870 deleting retry records that should have stayed in existence.
3872 48. Steve fixed some bugs in eximstats.
3874 49. The SPA authentication driver was not abandoning authentication and moving
3875 on to the next authenticator when an expansion was forced to fail,
3876 contradicting the general specification for all authenticators. Instead it
3877 was generating a temporary error. It now behaves as specified.
3879 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3880 (the order specifies the preference for clients). The order is now AES256,
3881 AES128, 3DES, ARCFOUR128.
3883 51. Small patch to Sieve code - explicitly set From: when generating an
3886 52. Exim crashed if a remote delivery caused a very long error message to be
3887 recorded - for instance if somebody sent an entire SpamAssassin report back
3888 as a large number of 550 error lines. This bug was coincidentally fixed by
3889 increasing the size of one of Exim's internal buffers (big_buffer) that
3890 happened as part of the Exiscan merge. However, to be on the safe side, I
3891 have made the code more robust (and fixed the comments that describe what
3894 53. Now that there can be additional text after "Completed" in log lines (if
3895 the queue_time_overall log selector is set), a one-byte patch to exigrep
3896 was needed to allow it to recognize "Completed" as not the last thing in
3899 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3900 patch that reportedly fixes this has been added. I am not expert enough to
3901 create a test for it. This is what the patch creator wrote:
3903 "I found a little strange behaviour of ldap code when working with
3904 Windows 2003 AD Domain, where users was placed in more than one
3905 Organization Units. When I tried to give exim partial DN, the exit code
3906 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3907 But simultaneously result of request was absolutely normal ldap result,
3908 so I produce this patch..."
3910 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3911 REFERENCE, so I have modified the code to exclude the patch when that macro
3914 55. Some experimental protocols are using DNS PTR records for new purposes. The
3915 keys for these records are domain names, not reversed IP addresses. The
3916 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3917 leaves it alone. Component reversal etc. now happens only for IP addresses.
3920 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3922 57. Double the size of the debug message buffer (to 2048) so that more of very
3923 long debug lines gets shown.
3925 58. The exicyclog utility now does better if the number of log files to keep
3926 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3928 59. Two changes related to the smtp_active_hostname option:
3930 (1) $smtp_active_hostname is now available as a variable.
3931 (2) The default for smtp_banner uses $smtp_active_hostname instead
3932 of $primary_hostname.
3934 60. The host_aton() function is supposed to be passed a string that is known
3935 to be a valid IP address. However, in the case of IPv6 addresses, it was
3936 not checking this. This is a hostage to fortune. Exim now panics and dies
3937 if the condition is not met. A case was found where this could be provoked
3938 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3939 components; fortuitously, this particular loophole had already been fixed
3940 by change 4.50/55 above.
3942 If there are any other similar loopholes, the new check in host_aton()
3943 itself should stop them being exploited. The report I received stated that
3944 data on the command line could provoke the exploit when Exim was running as
3945 exim, but did not say which command line option was involved. All I could
3946 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3947 running as the user.
3950 61. There was a buffer overflow vulnerability in the SPA authentication code
3951 (which came originally from the Samba project). I have added a test to the
3952 spa_base64_to_bits() function which I hope fixes it.
3955 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3956 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3958 63. The daemon start-up calls getloadavg() while still root for those OS that
3959 need the first call to be done as root, but it missed one case: when
3960 deliver_queue_load_max is set with deliver_drop_privilege. This is
3961 necessary for the benefit of the queue runner, because there is no re-exec
3962 when deliver_drop_privilege is set.
3964 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3965 This has been fixed.
3967 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3968 was in use, was not putting the data itself into the right store pool;
3969 consequently, it could be overwritten for a subsequent message in the same
3970 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3973 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3975 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3976 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3977 the function were treating the return as a boolean value, which happened to
3978 work because 0=false and not-0=true, but is not correct code.
3980 68. The host_aton() function was not handling scoped IPv6 addresses (those
3981 with, for example, "%eth0" on the end) correctly.
3983 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3984 negated items (that is, ~something) in unsigned ints. Some compilers
3985 apparently mutter when there is no cast.
3987 70. If an address verification called from an ACL failed, and did not produce a
3988 user-specific message (i.e. there was only a "system" message), nothing was
3989 put in $acl_verify_message. In this situation, it now puts the system
3992 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3993 session; change 4.31/43 added the unwanted input to the log line - except
3994 that it did not do this in the start of session case. It now does.
3996 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3997 This is wrong and can cause the other end to generate a synchronization
3998 error if it is another Exim or anything else that does the synchronization
3999 check. A QUIT command is no longer sent after a timeout.
4001 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4002 during host lookups.
4004 74. The defer_ok option of callout verification was not working if it was used
4005 when verifying addresses in header lines, that is, for this case:
4007 verify = header_sender/callout=defer_ok
4009 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4010 those file descriptors could be used for SMTP connections. If anything
4011 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4012 could be sent to the SMTP client, causing chaos. The daemon now opens
4013 stdin, stdout, and stderr to /dev/null when it puts itself into the
4016 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4017 log by default. The user can override this with suitable Perl magic.
4019 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4020 the log message when discard triggers, was not working for the DATA ACL or
4021 for the non-SMTP ACL.
4023 78. Error message wording change in sieve.c.
4025 79. If smtp_accept_max_per_host was set, the number of connections could be
4026 restricted to fewer than expected, because the daemon was trying to set up
4027 a new connection before checking whether the processes handling previous
4028 connections had finished. The check for completed processes is now done
4029 earlier. On busy systems, this bug wouldn't be noticed because something
4030 else would have woken the daemon, and it would have reaped the completed
4033 80. If a message was submitted locally by a user whose login name contained one
4034 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4035 It caused a spool format error. I have fixed the spool reading code. A
4036 related problem was that the "from" clause in the Received: line became
4037 illegal because of the space(s). It is now covered by ${quote_local_part.
4039 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4042 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4044 83. Patch from Sieve maintainer to fix unterminated string problem in
4045 "vacation" handling.
4047 84. Some minor changes to the Linux configuration files to help with other
4048 OS variants using glibc.
4050 85. One more patch for Sieve to update vacation handling to latest spec.
4053 ----------------------------------------------------
4054 See the note above about the 4.44 and 4.50 releases.
4055 ----------------------------------------------------
4061 1. Change 4.43/35 introduced a bug that caused file counts to be
4062 incorrectly computed when quota_filecount was set in an appendfile
4065 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4066 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4069 3. Give more explanation in the error message when the command for a transport
4070 filter fails to execute.
4072 4. There are several places where Exim runs a non-Exim command in a
4073 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4074 was being done only for the command run by the queryprogram router. It is
4075 now done for all such subprocesses. The other cases are: ${run, transport
4076 filters, and the commands run by the lmtp and pipe transports.
4078 5. Some older OS have a limit of 256 on the maximum number of file
4079 descriptors. Exim was using setrlimit() to set 1000 as a large value
4080 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4081 systems. I've change it so that if it can't get 1000, it tries for 256.
4083 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4084 was an oversight, and furthermore, ever since the addition of extra
4085 controls (e.g. 4.43/32), the checks on when to allow different forms of
4086 "control" were broken. There should now be diagnostics for all cases when a
4087 control that does not make sense is encountered.
4089 7. $recipients is now available in the predata ACL (oversight).
4091 8. Tidy the search cache before the fork to do a delivery from a message
4092 received from the command line. Otherwise the child will trigger a lookup
4093 failure and thereby defer the delivery if it tries to use (for example) a
4094 cached ldap connection that the parent has called unbind on.
4096 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4097 of $address_data from the recipient verification was clobbered by the
4098 sender verification.
4100 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4101 was its contents. (It was OK if the option was not defined at all.)
4103 11. A "Completed" log line is now written for messages that are removed from
4104 the spool by the -Mrm option.
4106 12. $host_address is now set to the target address during the checking of
4107 ignore_target_hosts.
4109 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4110 being passed; this would have caused $sender_host_name to have been used if
4111 matching the list had actually called for a host name (not very likely,
4112 since this list is usually IP addresses). A host name is now passed as
4115 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4116 code that specifies a non-restarting handler (typically sigaction() in
4117 modern systems) in an attempt to fix a rare and obscure crash bug.
4119 15. Narrowed the window for a race in the daemon that could cause it to ignore
4120 SIGCHLD signals. This is not a major problem, because they are used only to
4121 wake it up if nothing else does.
4123 16. A malformed maildirsize file could cause Exim to calculate negative values
4124 for the mailbox size or file count. Odd effects could occur as a result.
4125 The maildirsize information is now recalculated if the size or filecount
4128 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4129 support for a long time. Removed HAVE_SYS_VFS_H.
4131 18. Updated exipick to current release from John Jetmore.
4133 19. Allow an empty sender to be matched against a lookup in an address list.
4134 Previously the only cases considered were a regular expression, or an
4137 20. Exim went into a mad DNS lookup loop when doing a callout where the
4138 host was specified on the transport, if the DNS lookup yielded more than
4141 21. The RFC2047 encoding function was originally intended for short strings
4142 such as real names; it was not keeping to the 75-character limit for
4143 encoded words that the RFC imposes. It now respects the limit, and
4144 generates multiple encoded words if necessary. To be on the safe side, I
4145 have increased the buffer size for the ${rfc2047: expansion operator from
4148 22. Failure to deliver a bounce message always caused it to be frozen, even if
4149 there was an errors_to setting on the router. The errors_to setting is now
4152 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4153 canonical form (fully expanded) before being placed in
4154 $sender_host_address.
4156 24. Updated eximstats to version 1.33
4158 25. Include certificate and key file names in error message when GnuTLS fails
4159 to set them up, because the GnuTLS error message doesn't include the name
4160 of the failing file when there is a problem reading it.
4162 26. Expand error message when OpenSSL has problems setting up cert/key files.
4165 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4166 (this can affect the format of dates).
4168 28. exim_tidydb, when checking for the continued existence of a message for
4169 which it has found a message-specific retry record, was not finding
4170 messages that were in split spool directories. Consequently, it was
4171 deleting retry records that should have stayed in existence.
4173 29. eximstats updated to version 1.35
4174 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4175 1.35 - bugfix such that pie charts by volume are generated correctly
4177 30. The SPA authentication driver was not abandoning authentication and moving
4178 on to the next authenticator when an expansion was forced to fail,
4179 contradicting the general specification for all authenticators. Instead it
4180 was generating a temporary error. It now behaves as specified.
4182 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4183 (the order specifies the preference for clients). The order is now AES256,
4184 AES128, 3DES, ARCFOUR128.
4186 31. Small patch to Sieve code - explicitly set From: when generating an
4189 32. Exim crashed if a remote delivery caused a very long error message to be
4190 recorded - for instance if somebody sent an entire SpamAssassin report back
4191 as a large number of 550 error lines. This bug was coincidentally fixed by
4192 increasing the size of one of Exim's internal buffers (big_buffer) that
4193 happened as part of the Exiscan merge. However, to be on the safe side, I
4194 have made the code more robust (and fixed the comments that describe what
4197 33. Some experimental protocols are using DNS PTR records for new purposes. The
4198 keys for these records are domain names, not reversed IP addresses. The
4199 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4200 leaves it alone. Component reversal etc. now happens only for IP addresses.
4203 34. The host_aton() function is supposed to be passed a string that is known
4204 to be a valid IP address. However, in the case of IPv6 addresses, it was
4205 not checking this. This is a hostage to fortune. Exim now panics and dies
4206 if the condition is not met. A case was found where this could be provoked
4207 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4208 components; fortuitously, this particular loophole had already been fixed
4209 by change 4.50/55 or 4.44/33 above.
4211 If there are any other similar loopholes, the new check in host_aton()
4212 itself should stop them being exploited. The report I received stated that
4213 data on the command line could provoke the exploit when Exim was running as
4214 exim, but did not say which command line option was involved. All I could
4215 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4216 running as the user.
4219 35. There was a buffer overflow vulnerability in the SPA authentication code
4220 (which came originally from the Samba project). I have added a test to the
4221 spa_base64_to_bits() function which I hope fixes it.
4224 36. The daemon start-up calls getloadavg() while still root for those OS that
4225 need the first call to be done as root, but it missed one case: when
4226 deliver_queue_load_max is set with deliver_drop_privilege. This is
4227 necessary for the benefit of the queue runner, because there is no re-exec
4228 when deliver_drop_privilege is set.
4230 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4231 was in use, was not putting the data itself into the right store pool;
4232 consequently, it could be overwritten for a subsequent message in the same
4233 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4236 38. Sometimes the final signoff response after QUIT could fail to get
4237 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4238 < 0 before doing a fflush(). This bug looks as though it goes back to the
4239 introduction of TLS in release 3.20, but "sometimes" must have been rare
4240 because the tests only now provoked it.
4246 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4247 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4248 Now it returns an integer. A number of calls were still expecting a T/F
4249 return. Fortuitously, in all cases, the tests worked in OK situations,
4250 which is the norm. However, things would have gone wrong on any write
4251 failures on the smtp file descriptor. This function is used when sending
4252 messages over SMTP and also when doing verify callouts.
4254 2. When Exim is called to do synchronous delivery of a locally submitted
4255 message (the -odf or -odi options), it no longer closes stderr before doing
4258 3. Implemented the mua_wrapper option.
4260 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4262 5. Implemented the functions header_remove(), header_testname(),
4263 header_add_at_position(), and receive_remove_recipient(), and exported them
4266 6. If an ACL "warn" statement specified the addition of headers, Exim already
4267 inserted X-ACL-Warn: at the start if there was no header name. However, it
4268 was not making this test for the second and subsequent header lines if
4269 there were newlines in the string. This meant that an invalid header could
4270 be inserted if Exim was badly configured.
4272 7. Allow an ACL "warn" statement to add header lines at the start or after all
4273 the Received: headers, as well as at the end.
4275 8. Added the rcpt_4xx retry error code.
4277 9. Added postmaster_mailfrom=xxx to callout verification option.
4279 10. Added mailfrom=xxxx to the callout verification option, for verify=
4282 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4283 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4285 12. Inserted some casts to stop certain compilers warning when using pointer
4286 differences as field lengths or precisions in printf-type calls (mostly
4287 affecting debugging statements).
4289 13. Added optional readline() support for -be (dynamically loaded).
4291 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4292 same clock tick as a message's arrival, so that its received time was the
4293 same as the "first fail" time on the retry record, and that message
4294 remained on the queue past the ultimate address timeout, every queue runner
4295 would try a delivery (because it was past the ultimate address timeout) but
4296 after another failure, the ultimate address timeout, which should have then
4297 bounced the address, did not kick in. This was a "< instead of <=" error;
4298 in most cases the first failure would have been in the next clock tick
4299 after the received time, and all would be well.
4301 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4302 being recognized when the domain list was tested by the match_domain
4303 condition in an expansion string.
4305 16. Added the ${str2b64: operator.
4307 17. Exim was always calling setrlimit() to set a large limit for the number of
4308 processes, without checking whether the existing limit was already
4309 adequate. (It did check for the limit on file descriptors.) Furthermore,
4310 errors from getrlimit() and setrlimit() were being ignored. Now they are
4311 logged to the main and panic logs, but Exim does carry on, to try to do its
4312 job under whatever limits there are.
4314 18. Imported PCRE 5.0.
4316 19. Trivial typo in log message " temporarily refused connection" (the leading
4319 20. If the log selector return_path_on_delivery was set and an address was
4320 redirected to /dev/null, the delivery process crashed because it assumed
4321 that a return path would always be set for a "successful" delivery. In this
4322 case, the whole delivery is bypassed as an optimization, and therefore no
4325 21. Internal re-arrangement: the function for sending a challenge and reading
4326 a response while authentication was assuming a zero-terminated challenge
4327 string. It's now changed to take a pointer and a length, to allow for
4328 binary data in such strings.
4330 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4332 23. Exim was not respecting finduser_retries when seeking the login of the
4333 uid under which it was called; it was always trying 10 times. (The default
4334 setting of finduser_retries is zero.) Also, it was sleeping after the final
4335 failure, which is pointless.
4337 24. Implemented tls_on_connect_ports.
4339 25. Implemented acl_smtp_predata.
4341 26. If the domain in control=submission is set empty, Exim assumes that the
4342 authenticated id is a complete email address when it generates From: or
4343 Sender: header lines.
4345 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4346 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4347 chown and chgrp in /bin and hostname in /usr/bin.
4349 28. Exim was keeping the "process log" file open after each use, just as it
4350 does for the main log. This opens the possibility of it remaining open for
4351 long periods when the USR1 signal hits a daemon. Occasional processlog
4352 errors were reported, that could have been caused by this. Anyway, it seems
4353 much more sensible not to leave this file open at all, so that is what now
4356 29. The long-running daemon process does not normally write to the log once it
4357 has entered its main loop, and it closes the log before doing so. This is
4358 so that log files can straightforwardly be renamed and moved. However,
4359 there are a couple of unusual error situations where the daemon does write
4360 log entries, and I had neglected to close the log afterwards.
4362 30. The text of an SMTP error response that was received during a remote
4363 delivery was being truncated at 512 bytes. This is too short for some of
4364 the long messages that one sometimes sees. I've increased the limit to
4367 31. It is now possible to make retry rules that apply only when a message has a
4368 specific sender, in particular, an empty sender.
4370 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4371 it possible to be selective about when SMTP synchronization is enforced.
4373 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4375 32. Implemented hosts_connection_nolog.
4377 33. Added an ACL for QUIT.
4379 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4382 35. Added mailbox_size and mailbox_filecount to appendfile.
4384 36. Added control = no_multiline_responses to ACLs.
4386 37. There was a bug in the logic of the code that waits for the clock to tick
4387 in the case where the clock went backwards by a substantial amount such
4388 that the microsecond fraction of "now" was more than the microsecond
4389 fraction of "then" (but the whole seconds number was less).
4391 38. Added support for the libradius Radius client library this is found on
4392 FreeBSD (previously only the radiusclient library was supported).
4398 1. When certain lookups returned multiple values in the form name=value, the
4399 quoting of the values was not always being done properly. Specifically:
4400 (a) If the value started with a double quote, but contained no whitespace,
4402 (b) If the value contained whitespace other than a space character (i.e.
4403 tabs or newlines or carriage returns) it was not quoted.
4404 This fix has been applied to the mysql and pgsql lookups by writing a
4405 separate quoting function and calling it from the lookup code. The fix
4406 should probably also be applied to nisplus, ibase and oracle lookups, but
4407 since I cannot test any of those, I have not disturbed their existing code.
4409 2. A hit in the callout cache for a specific address caused a log line with no
4410 reason for rejecting RCPT. Now it says "Previous (cached) callout
4411 verification failure".
4413 3. There was an off-by-one bug in the queryprogram router. An over-long
4414 return line was truncated at 256 instead of 255 characters, thereby
4415 overflowing its buffer with the terminating zero. As well as fixing this, I
4416 have increased the buffer size to 1024 (and made a note to document this).
4418 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4419 when Exim is waiting for an SMTP response from a remote server, Exim
4420 restarts its select() call on the socket, thereby resetting its timeout.
4421 This is not a problem when such interrupts are rare. Somebody set up a cron
4422 job to run exiwhat every 2 minutes, which is less than the normal select()
4423 timeout (5 or 10 minutes). This meant that the select() timeout never
4424 kicked in because it was always reset. I have fixed this by comparing the
4425 time when an interrupt arrives with the time at the start of the first call
4426 to select(). If more time than the timeout has elapsed, the interrupt is
4427 treated as a timeout.
4429 5. Some internal re-factoring in preparation for the addition of Sieve
4430 extensions (by MH). In particular, the "personal" test is moved to a
4431 separate function, and given an option for scanning Cc: and Bcc: (which is
4432 not set for Exim filters).
4434 6. When Exim created an email address using the login of the caller as the
4435 local part (e.g. when creating a From: or Sender: header line), it was not
4436 quoting the local part when it contained special characters such as @.
4438 7. Installed new OpenBSD configuration files.
4440 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4441 try to make them clearer.
4443 9. Callout options, other than the timeout value, were being ignored when
4444 verifying sender addresses in header lines. For example, when using
4446 verify = header_sender/callout=no_cache
4448 the cache was (incorrectly) being used.
4450 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4451 only the Cygwin environment.
4453 11. When return_path_on_delivery was set as a log selector, if different remote
4454 addresses in the same message used different return paths and parallel
4455 remote delivery occurred, the wrong values would sometimes be logged.
4456 (Whenever a remote delivery process finished, the return path value from
4457 the most recently started remote delivery process was logged.)
4459 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4460 lines when AUTH and/or TLS are in use. This is the "received protocol"
4461 field. Exim used to use "asmtp" for authenticated SMTP, without any
4462 indication (in the protocol name) for TLS use. Now it follows the RFC and
4463 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4464 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4465 names appear in log lines as well as in Received: header lines.
4467 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4468 extensions, and comparison tests, and to fix some bugs.
4470 14. Changes to the "personal" filter test:
4472 (1) The test was buggy in that it was just doing the equivalent of
4473 "contains" tests on header lines. For example, if a user's address was
4474 anne@some.where, the "personal" test would incorrectly be true for
4476 To: susanne@some.where
4478 This test is now done by extracting each address from the header in turn,
4479 and checking the entire address. Other tests that are part of "personal"
4480 are now done using regular expressions (for example, to check local parts
4481 of addresses in From: header lines).
4483 (2) The list of non-personal local parts in From: addresses has been
4484 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4485 taken from the Sieve specification recommendations.
4487 (3) If the message contains any header line starting with "List-" it is
4488 treated as non-personal.
4490 (4) The test for "circular" in the Subject: header line has been removed
4491 because it now seems ill-conceived.
4493 15. Minor typos in src/EDITME comments corrected.
4495 16. Installed latest exipick from John Jetmore.
4497 17. If headers_add on a router specified a text string that was too long for
4498 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4499 of string_sprintf() is now avoided.
4501 18. $message_body_size was not set (it was always zero) when running the DATA
4502 ACL and the local_scan() function.
4504 19. For the "mail" command in an Exim filter, no default was being set for
4505 the once_repeat time, causing a random time value to be used if "once" was
4506 specified. (If the value happened to be <= 0, no repeat happened.) The
4507 default is now 0s, meaning "never repeat". The "vacation" command was OK
4508 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4509 (I found it when inspecting the code).
4511 20. There is now an overall timeout for performing a callout verification. It
4512 defaults to 4 times the callout timeout, which applies to individual SMTP
4513 commands during the callout. The overall timeout applies when there is more
4514 than one host that can be tried. The timeout is checked before trying the
4515 next host. This prevents very long delays if there are a large number of
4516 hosts and all are timing out (e.g. when the network connections are timing
4517 out). The value of the overall timeout can be changed by specifying an
4518 additional sub-option for "callout", called "maxwait". For example:
4520 verify = sender/callout=5s,maxwait=20s
4522 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4523 to the end before writing, but this should make it even safer).
4525 22. Exim was forgetting that it had advertised PIPELINING for the second and
4526 subsequent messages on an SMTP connection. It was also not resetting its
4527 memory on STARTTLS and an internal HELO.
4529 23. When Exim logs an SMTP synchronization error within a session, it now
4530 records whether PIPELINING has been advertised or not.
4532 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4533 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4534 rather than long int.
4536 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4538 26. Added the never_mail option to autoreply.
4544 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4545 crash if the getsockname() call failed; this can happen if a connection is
4546 closed very soon after it is established. The problem was simply in the
4547 order in which certain operations were done, causing Exim to try to write
4548 to the SMTP stream before it had set up the file descriptor. The bug has
4549 been fixed by making things happen in the correct order.
4555 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4556 before the connection was closed, thus losing the rejection response.
4558 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4559 some early Solaris releases, but causes trouble in current releases where
4560 socklen_t is defined.
4562 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4565 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4568 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4569 that releasing the top of it at the end releases what was used for sub-
4570 expansions (unless the block got too big). However, discard this block if
4571 the first thing is a variable or header, so that we can use its block when
4572 it is dynamic (useful for very large $message_headers, for example).
4574 6. Lookups now cache *every* query, not just the most recent. A new, separate
4575 store pool is used for this. It can be recovered when all lookup caches are
4576 flushed. Lookups now release memory at the end of their result strings.
4577 This has involved some general refactoring of the lookup sources.
4579 7. Some code has been added to the store_xxx() functions to reduce the amount
4580 of flapping under certain conditions.
4582 8. log_incoming_interface used to affect only the <= reception log lines. Now
4583 it causes the local interface and port to be added to several more SMTP log
4584 lines, for example "SMTP connection from", and rejection lines.
4586 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4588 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4590 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4591 could be overwritten at the end of the current message (or the start of a
4592 new message if it was set in a HELO ACL). The value is now preserved for
4593 the duration of the SMTP connection.
4595 12. If a transport had a headers_rewrite setting, and a matching header line
4596 contained an unqualified address, that address was qualified, even if it
4597 did not match any rewriting rules. The underlying bug was that the values
4598 of the flags that permit the existence of unqualified sender and recipient
4599 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4600 non-local messages, and by -bnq for local messages) were not being
4601 preserved with the message after it was received.
4603 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4604 "next input=" as part of the text comprising the host identity instead of
4605 the correct text. The code was using the same buffer for two different
4606 strings. However, depending on which order the printing function evaluated
4607 its arguments, the bug did not always show up. Under Linux, for example, my
4608 test suite worked just fine.
4610 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4611 doesn't work with some older versions of Perl. It has been changed to "my",
4612 which in any case is probably the better facility to use.
4614 15. A really picky compiler found some instances of statements for creating
4615 error messages that either had too many or two few arguments for the format
4618 16. The size of the buffer for calls to the DNS resolver has been increased
4619 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4620 for addresses that have a lot of PTR records. This alleviates a problem; it
4621 does not fully solve it.
4623 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4624 buffer now truncates the list and logs the incident, which is the same
4625 action as happens when Exim is looking up a host name and its aliases.
4626 Previously in this situation something unpredictable would happen;
4627 sometimes it was "internal error: store_reset failed".
4629 18. If a server dropped the connection unexpectedly when an Exim client was
4630 using GnuTLS and trying to read a response, the client delivery process
4631 crashed while trying to generate an error log message.
4633 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4634 string, for example:
4636 warn message = H1: something\nH2: something
4638 the text was added as a single header line from Exim's point of view
4639 though it ended up OK in the delivered message. However, searching for the
4640 second and subsequent header lines using $h_h2: did not work. This has been
4641 fixed. Similarly, if a system filter added multiple headers in this way,
4642 the routers could not see them.
4644 20. Expanded the error message when iplsearch is called with an invalid key to
4645 suggest using net-iplsearch in a host list.
4647 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4648 ACLs are no longer actually imposed (and a message to that effect is
4651 22. If a "gecos" field in a passwd entry contained escaped characters, in
4652 particular, if it contained a \" sequence, Exim got it wrong when building
4653 a From: or a Sender: header from that name. A second bug also caused
4654 incorrect handling when an unquoted " was present following a character
4655 that needed quoting.
4657 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4658 was not being matched caselessly.
4660 24. Arranged for all hyphens in the exim.8 source to be escaped with
4663 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4664 properly. Recipient callout cache records were still being keyed to include
4665 the sender, even when use_sender was set false. This led to far more
4666 callouts that were necessary. The sender is no longer included in the key
4667 when use_sender is false.
4669 26. Added "control = submission" modifier to ACLs.
4671 27. Added the ${base62d: operator to decode base 62 numbers.
4673 28. dnsdb lookups can now access SRV records.
4675 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4676 the configuration file.
4678 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4679 (-v) mode. This makes the output for a verbose queue run more intelligible.
4681 31. Added a use_postmaster feature to recipient callouts.
4683 32. Added the $body_zerocount variable, containing the number of binary zero
4684 bytes in the message body.
4686 33. The time of last modification of the "new" subdirectory is now used as the
4687 "mailbox time last read" when there is a quota error for a maildir
4690 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4692 35. Added +ignore_unknown as a special item in host lists.
4694 36. Code for decoding IPv6 addresses in host lists is now included, even if
4695 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4696 address was recognized as an IP address, but was then not correctly decoded
4697 into binary, causing unexpected and incorrect effects when compared with
4704 1. Very minor rewording of debugging text in manualroute to say "list of
4705 hosts" instead of "hostlist".
4707 2. If verify=header_syntax was set, and a header line with an unqualified
4708 address (no domain) and a large number of spaces between the end of the
4709 name and the colon was received, the reception process suffered a buffer
4710 overflow, and (when I tested it) crashed. This was caused by some obsolete
4711 code that should have been removed. The fix is to remove it!
4713 3. When running in the test harness, delay a bit after writing a bounce
4714 message to get a bit more predictability in the log output.
4716 4. Added a call to search_tidyup() just before forking a reception process. In
4717 theory, someone could use a lookup in the expansion of smtp_accept_max_
4718 per_host which, without the tidyup, could leave open a database connection.
4720 5. Added the variables $recipient_data and $sender_data which get set from a
4721 lookup success in an ACL "recipients" or "senders" condition, or a router
4722 "senders" option, similar to $domain_data and $local_part_data.
4724 6. Moved the writing of debug_print from before to after the "senders" test
4727 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4728 problems for message scanning, either using a data ACL, or using
4729 local_scan() because the Received: header was not generated till after they
4730 were called (in order to set the time as the time of reception completion).
4731 I have revised the way this works. The header is now generated after the
4732 body is received, but before the ACL or local_scan() are called. After they
4733 are run, the timestamp in the header is updated.
4739 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4740 before starting a queue runner without re-exec. This happened only when
4741 deliver_drop_privilege was set or when the Exim user was set to root. The
4742 effect of the bug was that timeouts during subsequent deliveries caused
4743 crashes instead of being properly handled. The handler is now left at its
4744 default (and expected) setting.
4746 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4747 message, again when deliver_drop_privilege is set or Exim is run as root.
4748 The bug described in (1) was not present in this case, but the tidying up
4749 of the other signals was missing. I have made the two cases consistent.
4751 3. The ignore_target_hosts setting on a manualroute router was being ignored
4752 for hosts that were looked up using the /MX notation.
4754 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4757 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4758 operated on the sender address. After changing the $sender_address to <>
4759 for the sender address verify, Exim was re-instated it as the original
4760 (before rewriting) address, but remembering that it had rewritten it, so it
4761 wasn't rewriting it again. This bug also had the effect of breaking the
4762 sender address verification caching when the sender address was rewritten.
4764 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4765 This has been changed so that if the ip literal address matches
4766 ignore_target_hosts, the router declines.
4768 7. Added expansion conditions match_domain, match_address, and match_local_
4769 part (NOT match_host).
4771 8. The placeholder for the Received: header didn't have a length field set.
4773 9. Added code to Exim itself and to exim_lock to test for a specific race
4774 condition that could lead to file corruption when using MBX delivery. The
4775 issue is with the lockfile that is created in /tmp. If this file is removed
4776 after a process has opened it but before that process has acquired a lock,
4777 there is the potential for a second process to recreate the file and also
4778 acquire a lock. This could lead to two Exim processes writing to the file
4779 at the same time. The added code performs the same test as UW imapd; it
4780 checks after acquiring the lock that its file descriptor still refers to
4781 the same named file.
4783 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4784 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4787 11. Added the smtp_active_hostname option. If used, this will typically be made
4788 to depend on the incoming interface address. Because $interface_address is
4789 not set up until the daemon has forked a reception process, error responses
4790 that can happen earlier (such as "too many connections") no longer contain
4793 12. If an expansion in a condition on a "warn" statement fails because a lookup
4794 defers, the "warn" statement is abandoned, and the next ACL statement is
4795 processed. Previously this caused the whole ACL to be aborted.
4797 13. Added the iplsearch lookup type.
4799 14. Added ident_timeout as a log selector.
4801 15. Added tls_certificate_verified as a log selector.
4803 16. Added a global option tls_require_ciphers (compare the smtp transport
4804 option of the same name). This controls incoming TLS connections.
4806 17. I finally figured out how to make tls_require_ciphers do a similar thing
4807 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4808 before starting the TLS session.
4810 18. Tabs are now shown as \t in -bP output.
4812 19. If the log selector return_path_on_delivery was set, Exim crashed when
4813 bouncing a message because it had too many Received: header lines.
4815 20. If two routers both had headers_remove settings, and the first one included
4816 a superfluous trailing colon, the final name in the first list and the
4817 first name in the second list were incorrectly joined into one item (with a
4818 colon in the middle).
4824 1. Added -C and -D options to the exinext utility, mainly to make it easier
4825 to include in the automated testing, but these could be helpful when
4826 multiple configurations are in use.
4828 2. The exinext utility was not formatting the output nicely when there was
4829 an alternate port involved in the retry record key, nor when there was a
4830 message id as well (for retries that were specific to a specific message
4831 and a specific host). It was also confused by IPv6 addresses, because of
4832 the additional colons they contain. I have fixed the IPv4 problem, and
4833 patched it up to do a reasonable job for IPv6.
4835 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4836 delivery, the log line now contains "pipelined" if PIPELINING was used.
4838 4. An SMTP transport process used to panic and die if the bind() call to set
4839 an explicit outgoing interface failed. This has been changed; it is now
4840 treated in the same way as a connect() failure.
4842 5. A reference to $sender_host_name in the part of a conditional expansion
4843 that was being skipped was still causing a DNS lookup. This no longer
4846 6. The def: expansion condition was not recognizing references to header lines
4847 that used bh_ and bheader_.
4849 7. Added the _cache feature to named lists.
4851 8. The code for checking quota_filecount in the appendfile transport was
4852 allowing one more file than it should have been.
4854 9. For compatibility with Sendmail, the command line option
4862 and sets the incoming protocol and host name (for trusted callers). The
4863 host name and its colon can be omitted when only the protocol is to be set.
4864 Note the Exim already has two private options, -pd and -ps, that refer to
4865 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4866 "s", but I don't think that's a major issue.
4868 10. A number of refactoring changes to the code, none of which should affect
4871 (a) The number of logging options was getting close to filling up the
4872 32-bit word that was used as a bit map. I have split them into two classes:
4873 those that are passed in the argument to log_write(), and those that are
4874 only ever tested independently outside of that function. These are now in
4875 separate 32-bit words, so there is plenty of room for expansion again.
4876 There is no change in the user interface or the logging behaviour.
4878 (b) When building, for example, log lines, the code previously used a
4879 macro that called string_cat() twice, in order to add two strings. This is
4880 not really sufficiently general. Furthermore, there was one instance where
4881 it was actually wrong because one of the argument was used twice, and in
4882 one call a function was used. (As it happened, calling the function twice
4883 did not affect the overall behaviour.) The macro has been replaced by a
4884 function that can join an arbitrary number of extra strings onto a growing
4887 (c) The code for expansion conditions now uses a table and a binary chop
4888 instead of a serial search (which was left over from when there were very
4889 few conditions). Also, it now recognizes conditions like "pam" even when
4890 the relevant support is not compiled in: a suitably worded error message is
4891 given if an attempt is made to use such a condition.
4893 11. Added ${time_interval:xxxxx}.
4895 12. A bug was causing one of the ddress fields not to be passed back correctly
4896 from remote delivery subprocesses. The field in question was not being
4897 subsequently used, so this caused to problems in practice.
4899 13. Added new log selectors queue_time and deliver_time.
4901 14. Might have fixed a bug in maildirsizefile handling that threw up
4902 "unexpected character" debug warnings, and recalculated the data
4903 unnecessarily. In any case, I expanded the warning message to give more
4906 15. Added the message "Restricted characters in address" to the statements in
4907 the default ACL that block characters like @ and % in local parts.
4909 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4910 Three changes have been made:
4912 (a) There was a serious bug; a negative response to MAIL caused the whole
4913 recipient domain to be cached as invalid, thereby blocking all messages
4914 to all local parts at the same domain, from all senders. This bug has
4915 been fixed. The domain is no longer cached after a negative response to
4916 MAIL if the sender used is not empty.
4918 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4921 (c) A new callout option, "use_sender" has been added for people who want
4922 the modified behaviour.
4928 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4931 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4932 indeed breaks things for older releases.
4934 3. Added additional logging to the case where there is a problem reading data
4935 from a filter that is running in a subprocess using a pipe, in order to
4936 try to track down a specific problem.
4938 4. Testing facility fudge: when running in the test harness and attempting
4939 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4940 getting "No route to host". Convert this to a timeout.
4942 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4945 6. Some OS don't have socklen_t but use size_t instead. This affects the
4946 fifth argument of getsockopt() amongst other things. This is now
4947 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4948 can be set for individual OS. I have set it for SunOS5, OSF1, and
4949 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4950 some earlier ones do not.
4952 7. Change 4.30/15 was not doing the test caselessly.
4954 8. The standard form for an IPv6 address literal was being rejected by address
4955 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4956 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4957 this, as well as the form without the "IPv6" on the front (but only when
4958 address literals are enabled, of course).
4960 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4962 10. Exim crashed if a message with an empty sender address specified by -f
4963 encountered a router with an errors_to setting. This could be provoked only
4964 by a command such as
4968 where an empty string was supplied; "<>" did not hit this bug.
4970 11. Installed PCRE release 4.5.
4972 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4973 remained set. It is now erased.
4975 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4976 times from message ids (which are base 36 rather than the normal 62).
4978 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4979 were being counted as actual protocol errors, and logged if the log
4980 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4981 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4982 and DATA following a set of rejected RCPTs do not count as protocol errors.
4983 In other words, Exim assumes they were pipelined, though this may not
4984 actually be the case. Of course, in all cases the client gets an
4985 appropriate error code.
4987 15. If a lookup fails in an ACL condition, a message about the failure may
4988 be available; it is used if testing the ACL cannot continue, because most
4989 such messages specify what the cause of the deferral is. However, some
4990 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4991 that caused an old message to be retained and used if a later statement
4992 caused a defer, replacing the real cause of the deferral.
4994 16. If an IP address had so many PTR records that the DNS lookup buffer
4995 was not large enough to hold them, Exim could crash while trying to process
4996 the truncated data. It now detects and logs this case.
4998 17. Further to 4.21/58, another change has been made: if (and only if) the
4999 first line of a message (the first header line) ends with CRLF, a bare LF
5000 in a subsequent header line has a space inserted after it, so as not to
5001 terminate the header.
5003 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5004 unnecessarily, used atoi() instead of strtol(), and didn't check the
5005 termination when getting file sizes from file names by regex.
5007 19. Completely re-implemented the support for maildirsize files, in the light
5008 of a number of problems with the previous contributed implementation
5009 (4.30/29). In particular:
5011 . If the quota is zero, the maildirsize file is maintained, but no quota is
5014 . If the maildir directory does not exist, it is created before any attempt
5015 to write a maildirsize file.
5017 . The quota value in the file is just a cache; if the quota is changed in
5018 the transport, the new value overrides.
5020 . A regular expression is available for excluding directories from the
5023 20. The autoreply transport checks the characters in options that define the
5024 message's headers; it allows continued headers, but it was checking with
5025 isspace() after an embedded newline instead of explicitly looking for a
5028 21. If all the "regular" hosts to which an address was routed had passed their
5029 expiry times, and had not reached their retry times, the address was
5030 bounced, even if fallback hosts were defined. Now Exim should go on to try
5033 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5034 equivalent code in the SMTP transport. Some hosts send humungous responses
5035 to HELO/EHLO, more than 1024 it seems.
5037 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5038 gives compiler warnings in some environments. I've now done it "properly",
5041 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5042 (because of problems with the built-in one) was declared to return uschar *
5043 instead of char *, causing compiler failure.
5045 25. Fixed a file descriptor leak when processing alias/forward files.
5047 26. Fixed a minor format string issue in dbfn.c.
5049 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5051 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5052 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5055 29. When a sender address is verified, it is cached, to save repeating the test
5056 when there is more than one recipient in a message. However, when the
5057 verification involves a callout, it is possible for different callout
5058 options to be set for different recipients. It is too complicated to keep
5059 track of this in the cache, so now Exim always runs a verification when a
5060 callout is required, relying on the callout cache for the optimization.
5061 The overhead is duplication of the address routing, but this should not be
5064 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5065 to be verified with callout=postmaster, and the main callout worked but the
5066 postmaster check failed, the verification correctly failed. However, if a
5067 subsequent RCPT command asked for sender verification *without* the
5068 postmaster check, incorrect caching caused this verification also to fail,
5071 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5072 it was not caching the DNS options (qualify_single, search_parents) that
5073 were used when the lookup failed. A subsequent lookup with different
5074 options therefore always gave the same answer, though there were cases
5075 where it should not have. (Example: a "domains = !$mx_any" option on a
5076 dnslookup router: the "domains" option is always processed without any
5077 widening, but the router might have qualify_single set.) Now Exim uses the
5078 cached value only when the same options are set.
5080 32. Added John Jetmore's "exipick" utility to the distribution.
5082 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5083 than a timeout (e.g. a certificate is required, and is not provided), an
5084 Exim server now closes the connection immediately. Previously it waited for
5085 the client to close - but if the client is SSL, it seems that they each
5086 wait for each other, leading to a delay before one of them times out.
5088 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5089 maintained 0.8.x compatibility because I don't think many are using it, and
5090 it is clearly obsolete.
5092 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5095 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5096 client certificate was expired. A simple patch fixes this, though I don't
5097 understand the full logic of why the verify callback is called multiple
5100 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5101 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5102 which causes problems with some clients (such as the Certicom SSL Plus
5103 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5104 disables the coutermeasure allowing Eudora to connect."
5106 38. Exim was not checking that a write() to a log file succeeded. This could
5107 lead to Bad Things if a log got too big, in particular if it hit a file
5108 size limit. Exim now panics and dies if it cannot write to a log file, just
5109 as it does if it cannot open a log file.
5111 39. Modified OS/Makefile-Linux so that it now contains
5113 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5115 The two -D definitions ensure that Exim is compiled with large file
5116 support, which makes it possible to handle log files that are bigger than
5119 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5120 instance) a domain was checked against a named list that involved a lookup,
5121 causing $domain_data to be set, then another domain was checked against the
5122 same list, then the first domain was re-checked, the value of $domain_data
5123 after the final check could be wrong. In particular, if the second check
5124 failed, it could be set empty. This bug probably also applied to
5127 41. The strip_trailing_dot option was not being applied to the address given
5128 with the -f command-line option.
5130 42. The code for reading a message's header from the spool was incrementing
5131 $received_count, but never initializing it. This meant that the value was
5132 incorrect (doubled) while delivering a message in the same process in which
5133 it was received. In the most common configuration of Exim, this never
5134 happens - a fresh exec is done - but it can happen when
5135 deliver_drop_privilege is set.
5137 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5138 it now includes up to 150 characters of the unexpected data in the log
5141 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5142 and building data strings. The size of both of these buffers was 10 000
5143 bytes - far larger than anybody would *ever* want, thought I. Needless to
5144 say, somebody hit the limit. I have increased the maximum line length to
5145 20 000 and the maximum data length of concatenated lines to 100 000. I have
5146 also fixed two bugs, because there was no checking on these buffers. Tsk,
5147 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5148 buffer is too small.
5150 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5151 lsearch lookups. Now it does.
5153 46. When parsing a route_list item in a manualroute router, a fixed-length
5154 buffer was used for the list of hosts. I made this 1024 bytes long,
5155 thinking that nobody would ever have a list of hosts that long. Wrong.
5156 Somebody had a whole pile of complicated expansion conditions, and the
5157 string was silently truncated, leading to an expansion error. It turns out
5158 that it is easier to change to an unlimited length (owing to other changes
5159 that have happened since this code was originally written) than to build
5160 structure for giving a limitation error. The length of the item that
5161 expands into the list of hosts is now unlimited.
5163 47. The lsearch lookup could not handle data where the length of text line was
5164 more than 4095 characters. Such lines were truncated, leading to shortened
5165 data being returned. It should now handle lines of any length.
5167 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5168 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5171 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5172 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5173 no longer happens in comment lines. A list of replacements is now placed
5174 at the head of all of the source files, except those whose only change is
5175 to replace PERL_COMMAND in the very first #! line.
5177 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5178 messages on the queue, with a bottom-up merge sort, using code contributed
5179 by Michael Haardt. This should make operations like -bp somewhat faster on
5180 large queues. It won't affect queue runners, except when queue_run_in_order
5183 51. Installed eximstats 1.31 in the distribution.
5185 52. Added support for SRV lookups to the dnslookup router.
5187 53. If an ACL referred to $message_body or $message_body_end, the value was not
5188 reset for any messages that followed in the same SMTP session.
5190 54. The store-handling optimization for building very long strings was not
5191 differentiating between the different store pools. I don't think this
5192 actually made any difference in practice, but I've tidied it.
5194 55. While running the routers to verify a sender address, $sender_address
5195 was still set to the sender address. This is wrong, because when routing to
5196 send a bounce to the sender, it would be empty. Therefore, I have changed
5197 it so that, while verifying a sender address, $sender_address is set to <>.
5198 (There is no change to what happens when verifying a recipient address.)
5200 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5201 target A or AAAA records (if not already returned) without resetting the
5202 qualify_single or search_parents options of the DNS resolver. These are
5203 inappropriate in this case because the targets of MX and SRV records must
5204 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5205 target that, when qualified, matched something in the local domain. These
5206 two options are now turned off when doing these lookups.
5208 57. It seems that at least some releases of Reiserfs (which does not have the
5209 concept of a fixed number of inodes) returns zero and not -1 for the
5210 number of available inodes. This interacted badly with check_spool_inodes,
5211 which assumed that -1 was the "no such thing" setting. What I have done is
5212 to check that the total number of inodes is greater than zero before doing
5213 the test of how many are available.
5215 58. When a "warn" ACL statement has a log_message modifier, the message is
5216 remembered, and not repeated. This is to avoid a lot of repetition when a
5217 message has many recipients that cause the same warning to be written.
5218 Howewer, Exim was preserving the list of already written lines for an
5219 entire SMTP session, which doesn't seem right. The memory is now reset if a
5220 new message is started.
5222 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5223 debugging output unless log_rewrite was also set.
5225 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5226 of (int)(handle) when we know that handle contains (void *)(-1).
5228 61. The Exim daemon panic-logs an error return when it closes the incoming
5229 connection. However "connection reset by peer" seems to be common, and
5230 isn't really an error worthy of noting specially, so that particular error
5233 62. When Exim is trying to find all the local interfaces, it used to panic and
5234 die if the ioctl to get the interface flags failed. However, it seems that
5235 on at least one OS (Solaris 9) it is possible to have an interface that is
5236 included in the list of interfaces, but for which you get a failure error
5237 for this call. This happens when the interface is not "plumbed" into a
5238 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5239 failure of the "get flags" call assumes that the interface is down.
5241 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5242 makes life easier for people who are doing arithmetic on fields extracted
5243 from dates, where you often get leading zeros that should not be
5244 interpreted as octal.
5246 64. Added qualify_domain to the redirect router, to override the global
5249 65. If a pathologically long header line contained very many addresses (the
5250 report of this problem mentioned 10 000) and each of them was rewritten,
5251 Exim could use up a very large amount of memory. (It kept on making new
5252 copies of the header line as it rewrote, and never released the old ones.)
5253 At the expense of a bit more processing, the header rewriting function has
5254 been changed so that it no longer eats memory in this way.
5256 66. The generation of the Received: header has been moved from the time that a
5257 message starts to be received, to the time that it finishes. The timestamp
5258 in the Received: header should now be very close to that of the <= log
5259 line. There are two side-effects of this change:
5261 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5262 logged header lines no longer include the local Received: line, because
5263 it has not yet been created. The same applies to a copy of the message
5264 that is returned to a non-SMTP sender when a message is rejected.
5266 (b) When a filter file is tested using -bf, no additional Received: header
5267 is added to the test message. After some thought, I decided that this
5270 This change does not affect the value of $received_for. It is still set
5271 after address rewriting, but before local_scan() is called.
5273 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5275 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5276 gave an unhelpful panic error message, and a defer error. I have managed to
5277 change this behaviour so that it now rejects any supplied certificate,
5278 which seems right, as the list of acceptable certificates is empty.
5280 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5281 gave an unhelpful defer error. I have not managed to make this reject any
5282 supplied certificates, but the error message it gives is "no certificate
5283 supplied", which is not helpful.
5285 70. exigrep's output now also includes lines that are not associated with any
5286 message, but which match the given pattern. Implemented by a patch from
5287 Martin Sluka, which also tidied up the Perl a bit.
5289 71. Recipient callout verification, like sender verification, was using <> in
5290 the MAIL FROM command. This isn't really the right thing, since the actual
5291 sender may affect whether the remote host accepts the recipient or not. I
5292 have changed it to use the actual sender in the callout; this means that
5293 the cache record is now keyed on a recipient/sender pair, not just the
5294 recipient address. There doesn't seem to be a real danger of callout loops,
5295 since a callout by the remote host to check the sender would use <>.
5296 [SEE ABOVE: changed after hitting problems.]
5298 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5299 temporary errors. However, in the case of such a code being given after
5300 the end of a data transmission (i.e. after ".") Exim was failing to write
5301 a retry record for the message. (Yes, there was some broken host that was
5302 actually sending 8xx at this point.)
5304 73. An unknown lookup type in a host list could cause Exim to panic-die when
5305 the list was checked. (An example that provoked this was putting <; in the
5306 middle of a list instead of at the start.) If this happened during a DATA
5307 ACL check, a -D file could be left lying around. This kind of configuration
5308 error no longer causes Exim to die; instead it causes a defer errror. The
5309 incident is still logged to the main and panic logs.
5311 74. Buglet left over from Exim 3 conversion. The message "too many messages
5312 in one connection" was written to the rejectlog but not the mainlog, except
5313 when address rewriting (yes!) was being logged.
5315 75. Added write_rejectlog option.
5317 76. When a system filter was run not as root (that is, when system_filter_user
5318 was set), the values of the $n variables were not being returned to the
5319 main process; thus, they were not subsequently available in the $sn
5322 77. Added +return_path_on_delivery log selector.
5324 78. A connection timeout was being treated differently from recipients deferred
5325 when testing hosts_max_try with a message that was older than the host's
5326 retry timeout. (The host should not be counted, thus allowing all hosts to
5327 be tried at least once before bouncing.) This may have been the cause of an
5328 occasionally reported bug whereby a message would remain on the queue
5329 longer than the retry timeout, but would be bounced if a delivery was
5330 forced. I say "may" because I never totally pinned down the problem;
5331 setting up timeout/retry tests is difficult. See also the next item.
5333 79. The ultimate address timeout was not being applied to errors that involved
5334 a combination of host plus message (for example, a timeout on a MAIL
5335 command). When an address resolved to a number of possible hosts, and they
5336 were not all tried for each delivery (e.g. because of hosts_max_try), a
5337 message could remain on the queue longer than the retry timeout.
5339 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5342 81. Fixed an obscure SMTP outgoing bug which required at least the following
5343 conditions: (a) there was another message waiting for the same server;
5344 (b) the server returned 5xx to all RCPT commands in the first message so
5345 that the message was not completed; (c) the server dropped the connection
5346 or gave a negative response to the RSET that Exim sends to abort the
5347 transaction. The observed case was a dropped connection after DATA that had
5348 been sent in pipelining mode. That is, the server had advertised PIPELINING
5349 but was not implementing it correctly. The effect of the bug was incorrect
5350 behaviour, such as trying another host, and this could lead to a crash.
5356 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5357 and daemon.c were passed as pointers to ints; they should have been
5358 pointers to socklen_t variables (which are typically unsigned ints).
5360 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5363 3. Fixed a really odd bug that affected only the testing scheme; patching a
5364 certain fixed string in the binary changed the value of another string that
5365 happened to be identical to the end of the original first string.
5367 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5368 name", it returns that address as the IP address. On some operating
5369 systems (e.g. Solaris), it also passes back the IP address string as the
5370 "host name". However, on others (e.g. Linux), it passes back an empty
5371 string. Exim wasn't checking for this, and was changing the host name to an
5372 empty string, assuming it had been canonicized.
5374 5. Although rare, it is permitted to have more than one PTR record for a given
5375 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5376 all the names associated with an address, because they do in Solaris.
5377 However, it seems that they do not in Linux for data that comes from the
5378 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5379 I found this out when I moved to a new Linux workstation and tried to run
5380 the Exim test suite.
5382 To get round this problem I have changed the code so that it now does its
5383 own call to the DNS to look up PTR records when searching for a host name.
5384 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5385 addresses that are only in /etc/hosts are still found.
5387 This behaviour is, however, controlled by an option called host_lookup_
5388 order, which defaults to "bydns:byaddr". If people want to use the other
5389 order, or indeed, just use one or the other means of lookup, they can
5390 specify it in this variable.
5392 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5393 some operating systems, this comes back from gethostbyaddr() as an empty
5394 string, and this is what Exim used to test for. However, it seems that in
5395 other systems, "." is yielded. Exim now tests for this case too.
5397 7. The values of check_spool_space and check_log_space are now held internally
5398 as a number of kilobytes instead of an absolute number of bytes. If a
5399 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5400 kilobyte. This means that much larger values can be stored.
5402 8. Exim monitor: an attempt to get the action menu when not actually pointing
5403 at a message produces an empty menu entitled "No message selected". This
5404 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5405 no entries in it ("Shell widget menu has zero width and/or height"). So I
5406 have added a single, blank menu entry in this case.
5408 9. Added ${quote_local_part.
5410 10. MIME decoding is now applied to the contents of Subject: header lines when
5413 11. Now that a reference to $sender_host_address automatically causes a reverse
5414 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5415 host lookup before query-style lookups in lists that might use this
5416 variable. This has therefore been abolished, and the "net-" prefix is no
5417 longer necessary for query-style lookups.
5419 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5420 have been a typo for LFLAGS, so it has been changed.
5422 13. The install script calls Exim with "-C /dev/null" in order to find the
5423 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5424 to be output. Howeve, since Exim outputs its version number before the
5425 error, it didn't break the script. It just looked ugly. I fixed this by
5426 always allowing "-C /dev/null" if the caller is root.
5428 14. Ignore overlarge ACL variable number when reading spool file - insurance
5429 against a later release with more variables having written the file.
5431 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5432 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5433 this, as well as the form without the "IPv6" on the front.
5435 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5436 OS/Makefile-Darwin file.
5438 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5439 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5441 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5442 variable. However, some OS use ints for inodes. I've added cast to long int
5443 to get rid of the compiler warning.
5445 19. I had forgotten to lock out "/../" in configuration file names when
5446 ALT_CONFIG_PREFIX was set.
5448 20. Routers used for verification do not need to specify transports. However,
5449 if such a router generated a host list, and callout was configured, Exim
5450 crashed, because it could not find a port number from the (non-existent)
5451 transport. It now assumes port 25 in this circumstance.
5453 21. Added the -t option to exigrep.
5455 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5456 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5457 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5458 them is set, LOOKUP_LSEARCH is forced.
5460 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5462 24. Added sender and host information to the "rejected by local_scan()" log
5463 line; previously there was no indication of these.
5465 25. Added .include_if_exists.
5467 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5468 Linux. It turns out that not all file systems support this. Apparently some
5469 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5470 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5471 not supported on the file descriptor, is now ignored when Exim is trying to
5472 sync a directory. This applies only to Linux.
5474 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5476 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5477 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5480 29. Added support for maildirsize files from supplied patch (modified a bit).
5482 30. The use of :fail: followed by an empty string could lead Exim to respond to
5483 sender verification failures with (e.g.):
5485 550 Verification failed for <xxx>
5486 550 Sender verify failed
5488 where the first response line was missing the '-' that indicates it is not
5489 the final line of the response.
5491 31. The loop for finding the name of the user that called Exim had a hardwired
5492 limit of 10; it now uses the value of finduser_retries, which is used for
5493 all other user lookups.
5495 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5498 33. Exim was neglecting to zero errno before one call of strtol() when
5499 expanding a string and expecting an integer value. On some systems this
5500 resulted in spurious "integer overflow" errors. Also, it was casting the
5501 result into an int without checking.
5503 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5504 did not work. The code looks as if it has *never* worked, though it appears
5505 to have been documented since at least releast 1.62. I have made it work.
5507 35. The "timeout_DNS" error in retry rules, also documented since at least
5508 1.62, also never worked. As it isn't clear exactly what this means, and
5509 clearly it isn't a major issue, I have abolished the feature by treating it
5510 as "timeout", and writing a warning to the main and panic logs.
5512 36. The display of retry rules for -brt wasn't always showing the error code
5515 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5516 timeout_connect_A, timeout_connect_MX.
5518 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5519 to the empty sender.
5521 39. The daemon was not analysing the content of -oX till after it had closed
5522 stderr and disconnected from the controlling terminal. This meant that any
5523 syntax errors were only noted on the panic log, and the return code from
5524 the command was 0. By re-arranging the code a little, I've made the
5525 decoding happen first, so such errors now appear on stderr, and the return
5526 code is 1. However, the actual setting up of the sockets still happens in
5527 the disconnected process, so errors there are still only recorded on the
5530 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5531 connections (as happens on some IP stacks) was logged at start up time as
5532 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5533 it from "IPv6 and IPv4", which means that two separate sockets are being
5536 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5537 says whether AF_INET or AF_INET6 was passed as an argument.
5539 42. Exiwhat output was messed up when time zones were included in log
5542 43. Exiwhat now gives more information about the daemon's listening ports,
5543 and whether -tls-on-connect was used.
5545 44. The "port" option of the smtp transport is now expanded.
5547 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5548 silently ignored. Now an error message is written to the main and panic
5551 46. There's a new ACL modifier called "logwrite" which writes to a log file
5552 as soon as it is encountered.
5554 47. Added $local_user_uid and $local_user_gid at routing time.
5556 48. Exim crashed when trying to verify a sender address that was being
5559 49. Exim was recognizing only a space character after ".include". It now also
5560 recognizes a tab character.
5562 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5563 extracting the relevant information from the specification. The man page no
5564 longer contains scrambled data for the -d option, and I've added a section
5565 at the front about calling Exim under different names.
5567 51. Added "extra_headers" argument to the "mail" command in filter files.
5569 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5572 53. Installed eximstats 1.29.
5574 54. Added transport_filter_timeout as a generic transport option.
5576 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5577 Cc: header lines. This was required by RFC 822, but it not required by RFC
5580 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5581 incoming messages that did not have them. Now it does so only if the
5582 message originates locally, that is, if there is no associated remote host
5583 address. When Resent- header lines are present, this applies to the Resent-
5584 lines rather than the non-Resent- lines.
5586 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5587 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5589 58. Messages for configuration errors now include the name of the main
5590 configuration file - useful now that there may be more than one file in a
5591 list (.included file names were always shown).
5593 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5594 for those rare installations that do not start the daemon as root or run it
5595 setuid root. I've cut out the call to initgroups() if the daemon is not
5598 60. The Exim user and group can now be bound into the binary as text strings
5599 that are looked up at the start of Exim's processing.
5601 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5603 62. Added $mailstore_basename variable.
5605 63. Installed patch to sieve.c from Michael Haardt.
5607 64. When Exim failed to open the panic log after failing to open the main log,
5608 the original message it was trying to log was written to stderr and debug
5609 output, but if they were not available (the usual case in production), it
5610 was lost. Now it is written to syslog before the two lines that record the
5611 failures to open the logs.
5613 65. Users' Exim filters run in subprocesses under the user's uid. It is
5614 possible for a "deliver" command or an alias in a "personal" command to
5615 provoke an address rewrite. If logging of address rewriting is configured,
5616 this fails because the process is not running as root or exim. There may be
5617 a better way of dealing with this, but for the moment (because 4.30 needs
5618 to be released), I have disabled address rewrite logging when running a
5619 filter in a non-root, non-exim process.
5625 1. The buildconfig auxiliary program wasn't quoting the value set for
5626 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5627 not defined. This bug was masked in 4.22 by the effect that was fixed in
5630 2. Some messages that were rejected after a message id was allocated were
5631 shown as "incomplete" by exigrep. It no longer does this for messages that
5632 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5634 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5635 have allow_domain_literals set, the ID did not get logged in the <= line.
5636 Domain literals are now always recognized in Message-ID: header lines.
5638 4. The first argument for a ${extract expansion item is the key name or field
5639 number. Leading and trailing spaces in this item were not being ignored,
5640 causing some misleading effects.
5642 5. When deliver_drop_privilege was set, single queue runner processes started
5643 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5644 same command in the process it spins off) were not dropping privilege.
5646 6. When the daemon running as "exim" started a queue runner, it always
5647 re-executed Exim in the spun-off process. This is a waste of effort when
5648 deliver_drop_privilege is set. The new process now just calls the
5649 queue-runner function directly.
5655 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5658 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5659 ignored. Though the use of -oP was forcing the writing of a pid file, it
5660 was always written to the default place.
5662 3. If the message "no IP address found for host xxxx" is generated during
5663 incoming verification, it is now followed by identification of the incoming
5664 connection (so you can more easily find what provoked it).
5666 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5668 5. Added some features to "harden" Exim a bit more against certain attacks:
5670 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5671 be put in Local/Makefile. This is like the never_users runtime option,
5672 but it cannot be overridden. The default setting is "root".
5674 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5675 prefix string with which any file named in a -C command line option
5678 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5679 is retained for -C and -D only if the caller of Exim is root. Without
5680 it, the exim user may also use -C and -D and retain privilege.
5682 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5683 command line option is disabled.
5685 6. Macro names set by the -D option must start with an upper case letter, just
5686 like macro names defined in the configuration file.
5688 7. Added "dereference=" facility to LDAP.
5690 8. Two instances of the typo "uknown" in the source files are fixed.
5692 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5693 the Configure-Makefile script screwed up while processing it.
5695 10. Incorporated PCRE 4.4.
5697 11. The SMTP synchronization check was not operating right at the start of an
5698 SMTP session. For example, it could not catch a HELO sent before the client
5699 waited for the greeting. There is now a check for outstanding input at the
5700 point when the greeting is written. Because of the duplex, asynchronous
5701 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5702 way, but not yet received, when the check is performed.
5704 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5705 on TCP/IP sockets, because this apparently causes some broken clients to
5708 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5709 unchanged) from the Cygwin maintainer.
5711 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5712 was supported instead of testing for mbx. Effectively a typo.
5714 15. The spa authenticator server code was not checking that the input it
5715 received was valid base64.
5717 16. The debug output line for the "set" modifier in ACLs was not showing the
5718 name of the variable that was being set.
5720 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5722 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5723 lookup on its own. Something else was needed to trigger the lookup. For
5724 example, a match in host_lookup or the need for a host name in a host list.
5725 Now, if $sender_host_name is referenced and the host name has not yet been
5726 looked up, a lookup is performed. If the lookup fails, the variable remains
5727 empty, and $host_lookup_failed is set to "1".
5729 19. Added "eqi" as a case-independent comparison operator.
5731 20. The saslauthd authentication condition could segfault if neither service
5732 nor realm was specified.
5734 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5735 error message that was logged was misleading, and incoming SMTP
5736 connections were dropped. The message is now more accurate, and temporary
5737 errors are given to SMTP connections.
5739 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5740 (except RSET) with a 421 error, until QUIT is received. However, it was
5741 failing to send a response to QUIT.
5743 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5744 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5745 ACL is now run after the helo_try_verify_hosts code.
5747 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5748 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5749 case-independent, so other case variants are also recognized). Apparently
5750 some systems use these upper case variants.
5752 25. If more than two messages were waiting for the same host, and a transport
5753 filter was specified for the transport, Exim sent two messages over the
5754 same TCP/IP connection, and then failed with "socket operation on non-
5755 socket" when it tried to send the third.
5757 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5759 27. The extern definition of crypt16() in expand.c was not being excluded when
5760 the OS had its own crypt16() function.
5762 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5763 as a preferred synonym for return_size_limit, both as an option and as an
5766 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5768 30. Changed the default configuration ACL to relax the local part checking rule
5769 for addresses that are not in any local domains. For these addresses,
5770 slashes and pipe symbols are allowed within local parts, but the sequence
5771 /../ is explicitly forbidden.
5773 31. SPA server authentication was not clearing the challenge buffer before
5776 32. log_message in a "warn" ACL statement was writing to the reject log as
5777 well as to the main log, which contradicts the documentation and doesn't
5778 seem right (because no rejection is happening). So I have stopped it.
5780 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5781 However, I am unable to do any testing of this.
5783 34. Fixed an infelicity in the appendfile transport. When checking directories
5784 for a mailbox, to see if any needed to be created, it was accidentally
5785 using path names with one or more superfluous leading slashes; tracing
5786 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5788 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5789 contents are added to the log line that is written for every discarded
5790 recipient. (Previously a log_message setting was ignored.)
5792 36. The ${quote: operator now quotes the string if it is empty.
5794 37. The install script runs exim in order to find its version number. If for
5795 some reason other than non-existence or emptiness, which it checks, it
5796 could not run './exim', it was installing it with an empty version number,
5797 i.e. as "exim-". This error state is now caught, and the installation is
5800 38. An argument was missing from the function that creates an error message
5801 when Exim fails to connect to the socket for saslauthd authentication.
5802 This could cause Exim to crash, or give a corrupted message.
5804 39. Added isip, isip4, and isip6 to ${if conditions.
5806 40. The ACL variables $acl_xx are now saved with the message, and can be
5807 accessed later in routers, transports, and filters.
5809 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5810 strings in the file are not string-expanded.
5812 42. If a MAIL command specified a SIZE value that was too large to fit into an
5813 int variable, the check against message_size_limit failed. Such values are
5814 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5815 day this will have to be increased, but I don't think I want to be around
5816 when emails are that large.
5823 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5824 iconv() is not standard in FreeBSD.
5826 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5827 IPv6 enabled. The observed symptom was a segmentation fault on return from
5828 the function os_common_find_running_interfaces() in src/os.c.
5830 3. In the check_special_case() function in daemon.c I had used "errno" as an
5831 argument name, which causes warnings on some systems. This was basically a
5832 typo, since it was named "eno" in the comments!
5834 4. The code that waits for the clock to tick (at a resolution of some fraction
5835 of a second) so as to ensure message-id uniqueness was always waiting for
5836 at least one whole tick, when it could have waited for less. [This is
5837 almost certainly not relevant at current processor speeds, where it is
5838 unlikely to ever wait at all. But we try to future-proof.]
5840 5. The function that sleeps for a time interval that includes fractions of a
5841 second contained a race. It did not block SIGALRM between setting the
5842 timer, and suspending (a couple of lines later). If the interval was short
5843 and the sigsuspend() was delayed until after it had expired, the suspension
5844 never ended. On busy systems this could lead to processes getting stuck for
5847 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5848 process, before it forks any delivery processes. The open lookup caching
5849 mechanism meant that the open file or database connection was passed into
5850 the delivery process. The problem was that delivery processes always tidy
5851 up cached lookup data. This could cause a problem for the next delivery
5852 process started by the queue runner, because the external queue runner
5853 process does not know about the closure. So the next delivery process
5854 still has data in the lookup cache. In the case of a file lookup, there was
5855 no problem because closing a file descriptor in a subprocess doesn't affect
5856 the parent. However, if the lookup was caching a connection to a database,
5857 the connection was closed, and the second delivery process was likely to
5858 see errors such as "PGSQL: query failed: server closed the connection
5859 unexpectedly". The problem has been fixed by closing all cached lookups
5860 in a queue runner before running a delivery process.
5862 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5863 seem to have the "const" qualifier which it has on other OS. I've
5866 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5867 *of the same type* (client or server) with the same public name that an
5868 error should be diagnosed.
5870 9. When Exim looked up a host name for an IP address, but failed to find the
5871 original IP address when looking up the host name (a safety check), it
5872 output the message "<ip address> does not match any IP for NULL", which was
5873 confusing, to say the least. The bug was that the host name should have
5874 appeared instead of "NULL".
5876 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5877 user that is built into the binary, and the -C or -D options is used, root
5878 privilege is dropped before the configuration file is read. In addition,
5879 logging is switched to stderr instead of the normal log files. If the
5880 configuration then re-defines the Exim user, the unprivileged environment
5881 is probably not what is expected, so Exim logs a panic warning message (but
5884 However, if deliver_drop_privilege is set, the unprivileged state may well
5885 be exactly what is intended, so the warning has been cut out in that case,
5886 and Exim is allowed to try to write to its normal log files.
5892 1. smtp_return_error_details was not giving details for temporary sender
5893 or receiver verification errors.
5895 2. Diagnose a configuration error if two authenticators have the same public
5898 3. Exim used not to create the message log file for a message until the first
5899 delivery attempt. This could be confusing when incoming messages were held
5900 for policy or load reasons. The message log file is now created at the time
5901 the message is received, and an initial "Received" line is written to it.
5903 4. The automatically generated man page for command line options had a minor
5904 bug that caused no ill effects; however, a more serious problem was that
5905 the procedure for building the man page automatically didn't always
5906 operate. Consequently, release 4.20 contains an out-of-date version. This
5907 shouldn't happen again.
5909 5. When building Exim with embedded Perl support, the script that builds the
5910 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5911 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5913 6. The freeze_tell option was not being used for messages that were frozen on
5914 arrival, either by an ACL or by local_scan().
5916 7. Added the smtp_incomplete_transaction log selector.
5918 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5919 was accepting AUTH without a new EHLO.
5921 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5922 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5925 10. Logging of TCP/IP connections (when configured) now happens in the main
5926 daemon process instead of the child process, so that the TCP/IP connection
5927 count is more accurate (but it can never be perfect).
5929 11. The use of "drop" in a nested ACL was not being handled correctly in the
5930 outer ACL. Now, if condition failure induced by the nested "drop" causes
5931 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5932 or "require"), the connection is dropped.
5934 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5935 that yield "discard" can now be used with an "accept" or a "discard" verb,
5936 but an error is generated for any others (because I can't see a useful way
5937 to define what should happen).
5939 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5940 are now processed in the same way as lines in the Exim configuration file.
5941 In particular, continuation lines are supported.
5943 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5945 15. Added -ti meaning -t -i.
5947 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5948 domains, and warn by logging if others are found.
5950 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5951 returned from GIFCONF when Exim is trying to find the list of interfaces on
5952 a host. The code in os.c has been modified to copy each ifreq to an aligned
5953 structure in all cases.
5955 Also, in some cases, the returned ifreq's were being copied to a 'struct
5956 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5957 means the last couple of bytes of an IPv6 address could be chopped if the
5958 ifreq contained only a normal sockaddr (14 bytes storage).
5960 18. Named domain lists were not supported in the hosts_treat_as_local option.
5961 An entry such as +xxxx was not recognized, and was treated as a literal
5964 19. Ensure that header lines added by a DATA ACL are included in the reject log
5965 if the ACL subsequently rejects the message.
5967 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5968 MD5 (which is deprecated).
5970 21. When testing a filter file using -bf, Exim was writing a message when it
5971 took the sender from a "From " line in the message, but it was not doing so
5972 when it took $return_path from a Return-Path: header line. It now does.
5974 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5975 with a valid header line field name (a series of printing characters
5976 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5978 23. Changed "disc" in the source to "disk" to conform to the documentation and
5979 the book and for uniformity.
5981 24. Ignore Sendmail's -Ooption=value command line item.
5983 25. When execve() failed while trying to run a command in a pipe transport,
5984 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5985 could be confused with a return value of 69 from the command itself. This
5986 has been changed to 127, the value the shell returns if it is asked to run
5987 a non-existent command. The wording for the related log line suggests a
5988 non-existent command as the problem.
5990 26. If received_header_text expands to an empty string, do not add a Received:
5991 header line to the message. (Well, it adds a token one on the spool, but
5992 marks it "old" so that it doesn't get used or transmitted.)
5994 27. Installed eximstats 1.28 (addition of -nt option).
5996 28. There was no check for failure on the call to getsockname() in the daemon
5997 code. This can fail if there is a shortage of resources on the system, with
5998 ENOMEM, for example. A temporary error is now given on failure.
6000 29. Contrary to the C standard, it seems that in some environments, the
6001 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6002 program. Exim now does this explicitly; it affects the formatting of
6003 timestamps using strftime().
6005 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6006 complaints. I've now initialized all the variables, to avoid this.
6008 32. Header lines added by a system filter were not being "seen" during
6009 transport-time rewrites.
6011 33. The info_callback() function passed to OpenSSL is set up with type void
6012 (*)(SSL *, int, int), as described somewhere. However, when calling the
6013 function (actually a macro) that sets it up, the type void(*)() is
6014 expected. I've put in a cast to prevent warnings from picky compilers.
6016 34. If a DNS black list lookup found a CNAME record, but there were no A
6017 records associated with the domain it pointed at, Exim crashed.
6019 35. If a DNS black list lookup returned more than one A record, Exim ignored
6020 all but the first. It now scans all returned addresses if a particular IP
6021 value is being sought. In this situation, the contents of the
6022 $dnslist_value variable are a list of all the addresses, separated by a
6025 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6026 to accept a host name and all its aliases if the forward lookup for any of
6027 them yielded the IP address of the incoming connection. Now it accepts only
6028 those names whose forward lookup yields the correct IP address. Any other
6029 names are discarded. This closes a loophole whereby a rogue DNS
6030 administrator could create reverse DNS records to break through a
6031 wildcarded host restriction in an ACL.
6033 37. If a user filter or a system filter that ran in a subprocess used any of
6034 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6035 the wrong values were passed to the pipe command ($thisaddress had the
6036 value of $0, $0 had the value of $1, etc). This bug was introduced by
6037 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6039 38. Improved the line breaking for long SMTP error messages from ACLs.
6040 Previously, if there was no break point between 40 and 75 characters, Exim
6041 left the rest of the message alone. Two changes have been made: (a) I've
6042 reduced the minimum length to 35 characters; (b) if it can't find a break
6043 point between 35 and 75 characters, it looks ahead and uses the first one
6044 that it finds. This may give the occasional overlong line, but at least the
6045 remaining text gets split now.
6047 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6048 file descriptors might be low, and that setting 1000 would always raise it.
6049 It turns out that in some environments, the limit is already over 1000 and
6050 that lowering it causes trouble. So now Exim takes care not to decrease it.
6052 40. When delivering a message, the value of $return_path is set to $sender_
6053 address at the start of routing (routers may change the value). By an
6054 oversight, this default was not being set up when an address was tested by
6055 -bt or -bv, which affected the outcome if any router or filter referred to
6058 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6059 log only when "message" or "log_message" are set. However, if one of the
6060 conditions was an address verification, or a call to a nested ACL, the
6061 messages generated by the underlying test were being passed through. This
6062 no longer happens. The underlying message is available in $acl_verify_
6063 message for both "message" and "log_message" expansions, so it can be
6064 passed through if needed.
6066 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6067 new expansion $bh_ to give the encoded byte string without charset
6068 translation. Translation happens only if iconv() is available; HAVE_ICONV
6069 indicates this at build time. HEADERS_CHARSET gives the charset to
6070 translate to; headers_charset can change it in the configuration, and
6071 "headers charset" can change it in an individual filter file.
6073 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6074 that creates RFC 2047 encoded "words" labels them as that charset instead
6075 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6076 expansion operator; (ii) when Exim creates a From: line for a local
6077 message; (iii) when a header line is rewritten to include a "phrase" part.
6079 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6080 buggy, causing it to skip the first lines of messages whose message ID
6081 ended in 'D'. This would not have bitten before Exim release 4.14, because
6082 message IDs were unlikely to end in 'D' before then. The effect was to have
6083 incorrect size information for certain domains.
6085 45. #include "config.h" was missing at the start of the crypt16.c module. This
6086 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6089 46. If there was a timeout during a "random" callout check, Exim treated it as
6090 a failure of the random address, and carried on sending RSET and the real
6091 address. If the delay was just some slowness somewhere, the response to the
6092 original RCPT would be taken as a response to RSET and so on, causing
6093 mayhem of various kinds.
6095 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6096 when I implemented it. It didn't allow for the fact that some option values
6097 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6098 the right test for positive values.
6100 48. Domain names in DNS records are case-independent. Exim always looks them up
6101 in lower case. Some resolvers return domain names in exactly the case they
6102 appear in the zone file, that is, they may contain uppercase letters. Not
6103 all resolvers do this - some return always lower case. Exim was treating a
6104 change of case by a resolver as a change of domain, similar to a widening
6105 of a domain abbreviation. This triggered its re-routing code and so it was
6106 trying to route what was effectively the same domain again. This normally
6107 caused routing to fail (because the router wouldn't handle the domain
6108 twice). Now Exim checks for this case specially, and just changes the
6109 casing of the domain that it ultimately uses when it transmits the message
6112 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6115 50. If a filter generated a file delivery with a non-absolute name (possible if
6116 no home directory exists for the router), the forbid_file option was not
6119 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6120 the existing equality matching.
6122 52. Exim was using ints instead of ino_t variables in some places where it was
6123 dealing with inode numbers.
6125 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6126 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6127 called TMPDIR, and if it finds it is different, it changes its value.
6129 54. The smtp_printf() function is now made available to local_scan() so
6130 additional output lines can be written before returning. There is also an
6131 smtp_fflush() function to enable the detection of a dropped connection.
6132 The variables smtp_input and smtp_batched_input are exported to
6135 55. Changed the default runtime configuration: the message "Unknown user"
6136 has been removed from the ACL, and instead placed on the localuser router,
6137 using the cannot_route_message feature. This means that any verification
6138 failures that generate their own messages won't get overridden. Similarly,
6139 the "Unrouteable address" message that was in the ACL for unverifiable
6140 relay addresses has also been removed.
6142 56. Added hosts_avoid_esmtp to the smtp transport.
6144 57. The exicyclog script was not checking for the esoteric option
6145 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6146 will work only if exicyclog is run under the appropriate euid.
6148 58. Following a discussion on the list, the rules by which Exim recognises line
6149 endings on incoming messages have been changed. The -dropcr and drop_cr
6150 options are now no-ops, retained only for backwards compatibility. The
6151 following line terminators are recognized: LF CRLF CR. However, special
6152 processing applies to CR:
6154 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6155 nor a local message in the state where . is a terminator.
6157 (ii) If a bare CR is encountered in a header line, an extra space is added
6158 after the line terminator so as not to end the header. The reasoning
6159 behind this is that bare CRs in header lines are most likely either
6160 to be mistakes, or people trying to play silly games.
6162 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6163 was being incorrectly given as 18 bytes larger than it should have been.
6164 This is a VOB (very old bug).
6166 60. This may never have affected anything current, but just in case it has:
6167 When the local host is found other than at the start of a list of hosts,
6168 the local host, those with the same MX, and any that follow, are discarded.
6169 When the list in question was part of a longer list of hosts, the following
6170 hosts (not currently being processed) were also being discarded. This no
6171 longer happens. I'm not sure if this situation could ever has previously
6174 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6177 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6178 header. This is something that is recommended in a new Internet Draft, and
6179 is something that is documented as being done by Sendmail. There are two
6180 possible values. For messages generated by the autoreply transport, Exim
6183 Auto-Submitted: auto-replied
6185 whereas for all other generated messages (e.g. bounces) it adds
6187 Auto-Submitted: auto-generated
6189 63. The "personal" condition in filters now includes a test for the
6190 Auto-Submitted: header. If it contains the string "auto-" the message it
6191 not considered personal.
6193 64. Added rcpt_include_affixes as a generic transport option.
6195 65. Added queue_only_override (default true).
6197 66. Added the syslog_duplication option.
6199 67. If what should have been the first header line of a message consisted of
6200 a space followed by a colon, Exim was mis-interpreting it as a header line.
6201 It isn't of course - it is syntactically invalid and should therefore be
6202 treated as the start of the message body. The misbehaviour could have
6203 caused a number of strange effects, including loss of data in subsequent
6204 header lines, and spool format errors.
6206 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6207 client host had authenticated. This control can now be exercised by an ACL
6208 for more flexibility.
6210 69. By default, callouts do not happen when testing with -bh. There is now a
6211 variant, -bhc, which does actually run the callout code, including
6212 consulting and updating the callout cache.
6214 70. Added support for saslauthd authentication, courtesy of Alexander
6217 71. If statvfs() failed on the spool or log directories while checking their
6218 size for availability, Exim confusingly gave the error "space shortage".
6219 Furthermore, in debugging mode it crashed with a floating point exception.
6220 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6221 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6222 serious problem, Exim now writes to the main and panic logs when this
6223 happens, with details of the failure. It then refuses to accept the
6224 incoming message, giving the message "spool directory problem" or "log
6225 directory problem" with a 421 code for SMTP messages.
6227 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6228 0, 1, and 2 exist, because some OS complain for execs without them (see
6229 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6230 descriptors. However, the code omitted to check that the open succeeded,
6231 causing mysterious errors if for some reason the permissions on /dev/null
6232 got screwed. Now Exim writes a message to the main and panic logs, and
6233 bombs out if it can't open /dev/null.
6235 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6236 interact so that it is all more flexible. It is supposed to remain
6237 backwards compatible. Also added extra_local_interfaces.
6239 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6240 to bomb out with an assertion failure - to the client this appears as a
6241 connection drop. This problem occurs in the part of the code that was taken
6242 from the Samba project. Fortunately, the assertion is in a very simple
6243 function, so I have fixed this by reproducing the function inline in the
6244 one place where it is called, and arranging for authentication to fail
6245 instead of killing the process with assert().
6247 75. The SPA client code was not working when the server requested OEM rather
6248 than Unicode encoding.
6250 76. Added code to make require_files with a specific uid setting more usable in
6251 the case where statting the file as root fails - usually a non-root-mounted
6252 NFS file system. When this happens and the failure is EACCES, Exim now
6253 forks a subprocess and does the per-uid checking as the relevant uid.
6255 77. Added process_log_path.
6257 78. If log_file_path was not explicitly set, a setting of check_log_space or
6258 check_log_inodes was ignored.
6260 79. If a space check for the spool or log partitions fails, the incident is now
6261 logged. Of course, in the latter case the data may get lost...
6263 80. Added the %p formatting code to string_format() so that it can be used to
6264 print addresses in debug_print(). Adjusted all the address printing in the
6265 debugging in store.c to use %p rather than %d.
6267 81. There was a concern that a line of code in smtp_in.c could overflow a
6268 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6269 initially expressed, the concern was not well-founded, because trailing
6270 spaces are removed early. However, if the trailing spaces were followed by
6271 a NULL, they did not get removed, so the overflow was possible. Two fixes
6274 (a) I re-wrote the offending code in a cleaner fashion.
6275 (b) If an incoming SMTP command contains a NULL character, it is rejected
6278 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6279 runs initgroups(), so that if the Exim user is in any additional groups,
6280 they will be used during message reception.
6286 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob