1 Change log file for Exim from version 4.21
2 -------------------------------------------
11 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
14 JH/02 The smtp transport option "multi_domain" is now expanded.
16 JH/03 The smtp transport now requests PRDR by default, if the server offers
19 JH/04 Certificate name checking on server certificates, when exim is a client,
20 is now done by default. The transport option tls_verify_cert_hostnames
21 can be used to disable this per-host. The build option
22 EXPERIMENTAL_CERTNAMES is withdrawn.
24 JH/05 The value of the tls_verify_certificates smtp transport and main options
25 default to the word "system" to access the system default CA bundle.
26 For GnuTLS, only version 3.0.20 or later.
28 JH/06 Verification of the server certificate for a TLS connection is now tried
29 (but not required) by default. The verification status is now logged by
30 default, for both outbound TLS and client-certificate supplying inbound
33 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
36 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
37 Status Notification (bounce) messages are now MIME format per RFC 3464.
38 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
39 under the control of the dsn_advertise_hosts option, and routers may
40 have a dsn_lasthop option.
42 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
43 default, modifiable by a malware= option. The list separator for
44 the options can now be changed in the usual way. Bug 68.
46 JH/10 The smtp_receive_timeout main option is now expanded before use.
48 JH/11 The incoming_interface log option now also enables logging of the
49 local interface on delivery outgoing connections.
51 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
52 if the interface and destination host and port all match.
54 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
57 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
58 Patch from Andrew Lewis.
60 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
61 now supports optional time-restrictions, weighting, and priority
62 modifiers per server. Patch originally by <rommer@active.by>.
64 JH/16 The spamd_address main option now supports a mixed list of local
65 and remote servers. Remote servers can be IPv6 addresses, and
68 JH/17 Bug 68: The spamd_address main option now supports an optional
69 timeout value per server.
71 JH/18 Bug 1581: Router and transport options headers_add/remove can
72 now have the list separator specified.
74 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
77 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
80 JH/21 Support for the A6 type of dns record is withdrawn.
82 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
83 rather than the verbs used.
85 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
86 from 255 to 1024 chars.
88 JH/24 Verification callouts now attempt to use TLS by default.
90 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
91 are generic router options now. The defaults didn't change.
93 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
94 Original patch from Alexander Shikoff, worked over by JH.
96 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
97 files if ZCAT_COMMAND is not executable.
99 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
101 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
103 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
104 Normally benign, it bites when the pair was led to by a CNAME;
105 modern usage is to not canoicalize the domain to a CNAME target
106 (and we were inconsistent anyway for A-only vs AAAA+A).
108 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
110 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
111 when evaluating $sender_host_dnssec.
113 JH/31 Check the HELO verification lookup for DNSSEC, adding new
114 $sender_helo_dnssec variable.
116 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
118 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
120 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
122 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
123 documented as working, but never had. Support all but $spam_report.
125 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
126 added for tls authenticator.
131 TL/01 When running the test suite, the README says that variables such as
132 no_msglog_check are global and can be placed anywhere in a specific
133 test's script, however it was observed that placement needed to be near
134 the beginning for it to behave that way. Changed the runtest perl
135 script to read through the entire script once to detect and set these
136 variables, reset to the beginning of the script, and then run through
137 the script parsing/test process like normal.
139 TL/02 The BSD's have an arc4random API. One of the functions to induce
140 adding randomness was arc4random_stir(), but it has been removed in
141 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
142 function when detected.
144 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
145 cause callback expansion.
147 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
148 syntax errors in an expansion can be treated as a string instead of
149 logging or causing an error, due to the internal use of bool_lax
150 instead of bool when processing it.
152 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
153 server certificates when making smtp deliveries.
155 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
157 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
159 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
161 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
162 Merged patch from Sebastian Wiedenroth.
164 JH/05 Fix results-pipe from transport process. Several recipients, combined
165 with certificate use, exposed issues where response data items split
166 over buffer boundaries were not parsed properly. This eventually
167 resulted in duplicates being sent. This issue only became common enough
168 to notice due to the introduction of conection certificate information,
169 the item size being so much larger. Found and fixed by Wolfgang Breyha.
171 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
172 size buffer was used, resulting in syntax errors when an expansion
175 JH/07 Add support for directories of certificates when compiled with a GnuTLS
176 version 3.3.6 or later.
178 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
179 is EXPERIMENTAL_EVENT, the main-configuration and transport options
180 both become "event_action", the variables become $event_name, $event_data
181 and $event_defer_errno. There is a new variable $verify_mode, usable in
182 routers, transports and related events. The tls:cert event is now also
183 raised for inbound connections, if the main configuration event_action
186 TL/06 In test suite, disable OCSP for old versions of openssl which contained
187 early OCSP support, but no stapling (appears to be less than 1.0.0).
189 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
190 server certificate names available under the smtp transport option
191 "tls_verify_cert_hostname" now do not permit multi-component wildcard
194 JH/10 Time-related extraction expansions from certificates now use the main
195 option "timezone" setting for output formatting, and are consistent
196 between OpenSSL and GnuTLS compilations. Bug 1541.
198 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
199 encoded parameter in the incoming message. Bug 1558.
201 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
202 include certificate info, eximon was claiming there were spoolfile
205 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
207 JH/14 Log delivery-related information more consistently, using the sequence
208 "H=<name> [<ip>]" wherever possible.
210 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
211 are problematic for Debian distribution, omit them from the release
214 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
216 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
219 JH/17 Fix a null-indirection in certextract expansions when a nondefault
220 output list separator was used.
225 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
226 checkers that were complaining about end of non-void function with no
229 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
230 This was a regression intruduced in 4.83 by another bugfix.
232 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
234 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
235 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
241 TF/01 Correctly close the server side of TLS when forking for delivery.
243 When a message was received over SMTP with TLS, Exim failed to clear up
244 the incoming connection properly after forking off the child process to
245 deliver the message. In some situations the subsequent outgoing
246 delivery connection happened to have the same fd number as the incoming
247 connection previously had. Exim would try to use TLS and fail, logging
248 a "Bad file descriptor" error.
250 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
251 utilities have not been installed.
253 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
254 temporary space as the ACL may create new global variables.
256 TL/01 LDAP support uses per connection or global context settings, depending
257 upon the detected version of the libraries at build time.
259 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
260 to extract and use the src ip:port in logging and expansions as if it
261 were a direct connection from the outside internet. PPv2 support was
262 updated based on HAProxy spec change in May 2014.
264 JH/02 Add ${listextract {number}{list}{success}{fail}}.
266 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
267 Properly escape header and check for NULL return.
269 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
272 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
274 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
275 characters in header names, implemented as a verify condition.
276 Contributed by Michael Fischer v. Mollard.
278 TL/05 Rename SPF condition results err_perm and err_temp to standardized
279 results permerror and temperror. Previous values are deprecated but
280 still accepted. In a future release, err_perm and err_temp will be
281 completely removed, which will be a backward incompatibility if the
282 ACL tests for either of these two old results. Patch contributed by
283 user bes-internal on the mailing list.
285 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
287 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
288 selectors, in both main and reject logs.
290 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
293 JH/07 Add malware type "sock" for talking to simple daemon.
295 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
297 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
298 routers/transports under cutthrough routing.
300 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
301 numbers. Touch up "bool" conditional to keep the same definition.
303 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
305 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
307 JH/12 Expand items in router/transport headers_add or headers_remove lists
308 individually rather than the list as a whole. Bug 1452.
310 Required for reasonable handling of multiple headers_ options when
311 they may be empty; requires that headers_remove items with embedded
312 colons must have them doubled (or the list-separator changed).
314 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
315 view the policy declared in the DMARC record. Currently, $dmarc_status
316 is a combined value of both the record presence and the result of the
319 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
321 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
322 dnslookup router and the smtp transport (applying to the forward
325 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
326 of ldap servers used for a specific lookup. Patch provided by Heiko
329 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
330 New variable $lookup_dnssec_authenticated for observability.
332 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
333 Patch submitted by Lars Timman.
335 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
337 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
338 Requires trusted mode and valid format message id, aborts otherwise.
339 Patch contributed by Heiko Schlichting.
341 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
342 certextract with support for various fields. Bug 1358.
344 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
345 is requested by default, modifiable by smtp transport option
348 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
349 operate on certificate variables to give certificate fingerprints
350 Also new ${sha256:cert_variable}.
352 JH/23 The PRDR feature is moved from being Experimental into the mainline.
354 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
355 Christian Aistleitner.
357 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
359 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
360 file. Patch from Wolfgang Breyha.
362 JH/25 Expand the coverage of the delivery $host and $host_address to
363 client authenticators run in verify callout. Bug 1476.
365 JH/26 Port service names are now accepted for tls_on_connect_ports, to
366 align with daemon_smtp_ports. Bug 72.
368 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
369 support and error reporting did not work properly.
371 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
372 and is readable. Patch from Andrew Colin Kissa.
374 TL/14 Enhance documentation of ${run expansion and how it parses the
375 commandline after expansion, particularly in the case when an
376 unquoted variable expansion results in an empty value.
378 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
380 PP/02 Fix internal collision of T_APL on systems which support RFC3123
381 by renaming away from it. Addresses GH issue 15, reported by
384 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
386 TL/15 SECURITY: prevent double expansion in math comparison functions
387 (can expand unsanitized data). Not remotely exploitable.
394 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
396 PP/02 Make -n do something, by making it not do something.
397 When combined with -bP, the name of an option is not output.
399 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
402 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
403 $sender_host_name and config options to manage this, and basic check
406 PP/05 DSCP support for outbound connections and control modifier for inbound.
408 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
409 (Only plugin which currently uses this is kerberos4, which nobody should
410 be using, but we should make it available and other future plugins might
411 conceivably use it, even though it would break NAT; stuff *should* be
412 using channel bindings instead).
414 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
415 name; added for Sendmail compatibility; requires admin caller.
416 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
417 just ignore it); requires trusted caller.
418 Also parse but ignore: -Ac -Am -X<logfile>
421 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
423 TL/02 Add +smtp_confirmation as a default logging option.
425 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
426 Patch by Magnus Holmgren from 2007-02-20.
428 TL/04 Bugzilla 1281 - Spec typo.
429 Bugzilla 1283 - Spec typo.
430 Bugzilla 1290 - Spec grammar fixes.
432 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
434 TL/06 Add Experimental DMARC support using libopendmarc libraries.
436 TL/07 Fix an out of order global option causing a segfault. Reported to dev
437 mailing list by by Dmitry Isaikin.
439 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
441 JH/02 Support "G" suffix to numbers in ${if comparisons.
443 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
445 NM/01 Bugzilla 1197 - Spec typo
446 Bugzilla 1196 - Spec examples corrections
448 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
450 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
451 gnutls_enable_pkcs11, but renamed to more accurately indicate its
454 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
455 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
457 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
458 "acl {{name}{arg}...}", and optional args on acl condition
461 JH/05 Permit multiple router/transport headers_add/remove lines.
463 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
465 JH/07 Avoid using a waiting database for a single-message-only transport.
466 Performance patch from Paul Fisher. Bugzilla 1262.
468 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
471 JH/09 Add $headers_added variable, with content from use of ACL modifier
472 add_header (but not yet added to the message). Bugzilla 199.
474 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
475 Pulled from Bugzilla 817 by Wolfgang Breyha.
477 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
479 (nb: this is the same fix as in Exim 4.80.1)
481 JH/11 Add A= logging on delivery lines, and a client_set_id option on
484 JH/12 Add optional authenticated_sender logging to A= and a log_selector
487 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
489 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
490 advertise SMTP AUTH mechanism to us, instead of a generic
491 protocol violation error. Also, make Exim more robust to bad
492 data from the Dovecot auth socket.
494 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
496 When a queue runner is handling a message, Exim first routes the
497 recipient addresses, during which it prunes them based on the retry
498 hints database. After that it attempts to deliver the message to
499 any remaining recipients. It then updates the hints database using
502 So if a recipient address works intermittently, it can get repeatedly
503 deferred at routing time. The retry hints record remains fresh so the
504 address never reaches the final cutoff time.
506 This is a fairly common occurrence when a user is bumping up against
507 their storage quota. Exim had some logic in its local delivery code
508 to deal with this. However it did not apply to per-recipient defers
509 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
511 This change adds a proper retry rule check during routing so that the
512 final cutoff time is checked against the message's age. We only do
513 this check if there is an address retry record and there is not a
514 domain retry record; this implies that previous attempts to handle
515 the address had the retry_use_local_parts option turned on. We use
516 this as an approximation for the destination being like a local
517 delivery, as in LMTP.
519 I suspect this new check makes the old local delivery cutoff check
520 redundant, but I have not verified this so I left the code in place.
522 TF/02 Correct gecos expansion when From: is a prefix of the username.
524 Test 0254 submits a message to Exim with the header
528 When I ran the test suite under the user fanf2, Exim expanded
529 the header to contain my full name, whereas it should have added
530 a Resent-Sender: header. It erroneously treats any prefix of the
531 username as equal to the username.
533 This change corrects that bug.
535 GF/01 DCC debug and logging tidyup
536 Error conditions log to paniclog rather than rejectlog.
537 Debug lines prefixed by "DCC: " to remove any ambiguity.
539 TF/03 Avoid unnecessary rebuilds of lookup-related code.
541 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
542 Bug spotted by Jeremy Harris; was flawed since initial commit.
543 Would have resulted in OCSP responses post-SNI triggering an Exim
544 NULL dereference and crash.
546 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
548 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
549 Bug detection, analysis and fix by Samuel Thibault.
550 Bugzilla 1331, Debian bug #698092.
552 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
554 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
555 Server implementation by Todd Lyons, client by JH.
556 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
557 config variable "prdr_enable" controls whether the server
558 advertises the facility. If the client requests PRDR a new
559 acl_data_smtp_prdr ACL is called once for each recipient, after
560 the body content is received and before the acl_smtp_data ACL.
561 The client is controlled by bolth of: a hosts_try_prdr option
562 on the smtp transport, and the server advertisement.
563 Default client logging of deliveries and rejections involving
564 PRDR are flagged with the string "PRDR".
566 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
567 fclose(). Diagnosis by Todd Lyons.
569 PP/17 Update configure.default to handle IPv6 localhost better.
570 Patch by Alain Williams (plus minor tweaks).
573 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
574 This is now consistent with GnuTLS, and is now documented: the
575 previous undocumented portable approach to treating the option as
576 unset was to force an expansion failure. That still works, and
577 an empty string is now equivalent.
579 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
580 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
581 not performing validation itself.
583 PP/20 Added force_command boolean option to pipe transport.
584 Patch from Nick Koston, of cPanel Inc.
586 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
589 TF/04 Added udpsend ACL modifer and hexquote expansion operator
591 PP/21 Fix eximon continuous updating with timestamped log-files.
592 Broken in a format-string cleanup in 4.80, missed when I repaired the
593 other false fix of the same issue.
594 Report and fix from Heiko Schlichting.
597 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
598 Report from Prashanth Katuri.
600 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
601 It's SecureTransport, so affects any MacOS clients which use the
602 system-integrated TLS libraries, including email clients.
604 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
605 using a MIME ACL for non-SMTP local injection.
606 Report and assistance in diagnosis by Warren Baker.
608 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
610 JH/16 Fix comparisons for 64b. Bugzilla 1385.
612 TL/09 Add expansion variable $authenticated_fail_id to keep track of
613 last id that failed so it may be referenced in subsequent ACL's.
615 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
618 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
619 ldap library initialization, allowing self-signed CA's to be
620 used. Also properly sets require_cert option later in code by
621 using NULL (global ldap config) instead of ldap handle (per
622 session). Bug diagnosis and testing by alxgomz.
624 TL/12 Enhanced documentation in the ratelimit.pl script provided in
625 the src/util/ subdirectory.
627 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
628 renamed to Transport Post Delivery Action by Jeremy Harris, as
631 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
632 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
633 redis_servers = needs to be configured which will be used by the redis
634 lookup. Patch from Warren Baker, of The Packet Hub.
636 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
638 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
639 hostname or reverse DNS when processing a host list. Used suggestions
640 from multiple comments on this bug.
642 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
644 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
645 Missed a few lines, added it to make the runtest require no keyboard
648 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
649 contains upper case chars. Make router use caseful_local_part.
651 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
652 support when GnuTLS has been built with p11-kit.
658 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
660 This, or similar/improved, will also be change PP/11 of 4.82.
666 PP/01 Handle short writes when writing local log-files.
667 In practice, only affects FreeBSD (8 onwards).
668 Bugzilla 1053, with thanks to Dmitry Isaikin.
670 NM/01 Bugzilla 949 - Documentation tweak
672 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
675 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
677 PP/02 Implemented gsasl authenticator.
679 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
681 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
682 `pkg-config foo` for cflags/libs.
684 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
685 with rest of GSASL and with heimdal_gssapi.
687 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
688 `pkg-config foo` for cflags/libs for the TLS implementation.
690 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
691 properties get this fed in as external SSF. A number of robustness
692 and debugging improvements to the cyrus_sasl authenticator.
694 PP/08 cyrus_sasl server now expands the server_realm option.
696 PP/09 Bugzilla 1214 - Log authentication information in reject log.
697 Patch by Jeremy Harris.
699 PP/10 Added dbmjz lookup type.
701 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
703 PP/12 MAIL args handles TAB as well as SP, for better interop with
704 non-compliant senders.
705 Analysis and variant patch by Todd Lyons.
707 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
708 Bug report from Lars Müller <lars@samba.org> (via SUSE),
709 Patch from Dirk Mueller <dmueller@suse.com>
711 PP/13 tls_peerdn now print-escaped for spool files.
712 Observed some $tls_peerdn in wild which contained \n, which resulted
713 in spool file corruption.
715 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
716 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
717 or write after TLS renegotiation, which otherwise led to messages
720 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
721 as a tracking header (ie: a signed header comes before the signature).
722 Patch from Wolfgang Breyha.
724 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
725 comma-sep list; embedded commas doubled.
727 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
729 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
731 Report and patch from Dmitry Banschikov.
733 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
734 Removed SSL_clear() after SSL_new() which led to protocol negotiation
735 failures. We appear to now support TLS1.1+ with Exim.
737 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
738 lets Exim select keys and certificates based upon TLS SNI from client.
739 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
740 before an outbound SMTP session. New log_selector, +tls_sni.
742 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
743 NULL dereference. Report and patch from Alun Jones.
745 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
746 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
747 Not seeing resolver debug output on NetBSD, but suspect this is a
748 resolver implementation change.
750 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
751 Left warnings. Added "eximon gdb" invocation mode.
753 PP/21 Defaulting "accept_8bitmime" to true, not false.
755 PP/22 Added -bw for inetd wait mode support.
757 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
758 locate the relevant includes and libraries. Made this the default.
760 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
761 Bugzilla 1246, report and most of solution from Tomasz Kusy.
763 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
764 This may cause build issues on older platforms.
766 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
767 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
768 gnutls_require_mac & gnutls_require_protocols (no longer supported).
769 Added SNI support via GnuTLS too.
770 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
772 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
774 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
775 Applied second patch from Janne, implementing suggestion to default
776 multiple-strings-in-record handling to match SPF spec.
778 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
780 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
781 read-only, out of scope).
782 Patch from Wolfgang Breyha, report from Stuart Northfield.
784 PP/29 Fix three issues highlighted by clang analyser static analysis.
785 Only crash-plausible issue would require the Cambridge-specific
786 iplookup router and a misconfiguration.
787 Report from Marcin Mirosław.
789 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
791 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
792 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
793 As part of this, removing so much warning spew let me fix some minor
794 real issues in debug logging.
796 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
797 assignment on my part. Fixed.
799 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
800 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
801 Janne Snabb (who went above and beyond: thank you).
803 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
804 string otherwise requires a connection and a bunch more work and it's
805 relatively easy to get wrong. Should also expose TLS library linkage
808 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
809 64-bit ${eval} (JH/03).
811 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
812 GNU libc to support some of the 64-bit stuff, should not lead to
813 conflicts. Defined before os.h is pulled in, so if a given platform
814 needs to override this, it can.
816 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
817 protection layer was required, which is not implemented.
818 Bugzilla 1254, patch from Wolfgang Breyha.
820 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
821 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
822 tls_dhparam take prime identifiers. Also unbreak combination of
823 OpenSSL+DH_params+TLSSNI.
825 PP/39 Disable SSLv2 by default in OpenSSL support.
831 PP/01 Solaris build fix for Oracle's LDAP libraries.
832 Bugzilla 1109, patch from Stephen Usher.
834 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
836 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
839 TF/02 Fix a couple more cases where we did not log the error message
840 when unlink() failed. See also change 4.74-TF/03.
842 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
843 lock up or crash if it happened to be inside a call to libc when it
844 got a SIGUSR1 from exiwhat.
846 The SIGUSR1 handler appends the current process status to the process
847 log which is later printed by exiwhat. It used to use the general
848 purpose logging code to do this, but several functions it calls are
849 not safe for signals.
851 The new output code in the SIGUSR1 handler is specific to the process
852 log, and simple enough that it's easy to inspect for signal safety.
853 Removing some special cases also simplifies the general logging code.
854 Removing the spurious timestamps from the process log simplifies
857 TF/04 Improved ratelimit ACL condition.
859 The /noupdate option has been deprecated in favour of /readonly which
860 has clearer semantics. The /leaky, /strict, and /readonly update modes
861 are mutually exclusive. The update mode is no longer included in the
862 database key; it just determines when the database is updated. (This
863 means that when you upgrde Exim will forget old rate measurements.)
865 Exim now checks that the per_* options are used with an update mode that
866 makes sense for the current ACL. For example, when Exim is processing a
867 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
868 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
869 must specify per_mail/readonly. If you omit the update mode it defaults to
870 /leaky where that makes sense (as before) or /readonly where required.
872 The /noupdate option is now undocumented but still supported for
873 backwards compatibility. It is equivalent to /readonly except that in
874 ACLs where /readonly is required you may specify /leaky/noupdate or
875 /strict/noupdate which are treated the same as /readonly.
877 A useful new feature is the /count= option. This is a generalization
878 of the per_byte option, so that you can measure the throughput of other
879 aggregate values. For example, the per_byte option is now equivalent
880 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
882 The per_rcpt option has been generalized using the /count= mechanism
883 (though it's more complicated than the per_byte equivalence). When it is
884 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
885 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
886 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
887 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
888 Note that using per_rcpt with a non-readonly update mode in more than
889 one ACL will cause the recipients to be double-counted. (The per_mail
890 and per_byte options don't have this problem.)
892 The handling of very low rates has changed slightly. If the computed rate
893 is less than the event's count (usually one) then this event is the first
894 after a long gap. In this case the rate is set to the same as this event's
895 count, so that the first message of a spam run is counted properly.
897 The major new feature is a mechanism for counting the rate of unique
898 events. The new per_addr option counts the number of different
899 recipients that someone has sent messages to in the last time period. It
900 behaves like per_rcpt if all the recipient addresses are different, but
901 duplicate recipient addresses do not increase the measured rate. Like
902 the /count= option this is a general mechanism, so the per_addr option
903 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
904 example, measure the rate that a client uses different sender addresses
905 with the options per_mail/unique=$sender_address. There are further
906 details in the main documentation.
908 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
910 TF/06 Removed a few PCRE remnants.
912 TF/07 Automatically extract Exim's version number from tags in the git
913 repository when doing development or release builds.
915 PP/02 Raise smtp_cmd_buffer_size to 16kB.
916 Bugzilla 879. Patch from Paul Fisher.
918 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
919 Heavily based on revision 40f9a89a from Simon Arlott's tree.
922 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
924 PP/05 Variable $av_failed, true if the AV scanner deferred.
925 Bugzilla 1078. Patch from John Horne.
927 PP/06 Stop make process more reliably on build failure.
928 Bugzilla 1087. Patch from Heiko Schlittermann.
930 PP/07 Make maildir_use_size_file an _expandable_ boolean.
931 Bugzilla 1089. Patch from Heiko Schlittermann.
933 PP/08 Handle ${run} returning more data than OS pipe buffer size.
934 Bugzilla 1131. Patch from Holger Weiß.
936 PP/09 Handle IPv6 addresses with SPF.
937 Bugzilla 860. Patch from Wolfgang Breyha.
939 PP/10 GnuTLS: support TLS 1.2 & 1.1.
941 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
944 PP/11 match_* no longer expand right-hand-side by default.
945 New compile-time build option, EXPAND_LISTMATCH_RHS.
946 New expansion conditions, "inlist", "inlisti".
948 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
950 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
952 PP/14 fix log_write() format string regression from TF/03.
953 Bugzilla 1152. Patch from Dmitry Isaikin.
959 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
961 PP/02 Harmonised TLS library version reporting; only show if debugging.
962 Layout now matches that introduced for other libraries in 4.74 PP/03.
964 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
966 PP/04 New "dns_use_edns0" global option.
968 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
971 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
972 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
974 TK/01 Updated PolarSSL code to 0.14.2.
975 Bugzilla 1097. Patch from Andreas Metzler.
977 PP/07 Catch divide-by-zero in ${eval:...}.
980 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
983 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
984 format-string attack -- SECURITY: remote arbitrary code execution.
986 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
987 time unintentionally subject to list matching rules, letting the header
988 cause arbitrary Exim lookups (of items which can occur in lists, *not*
989 arbitrary string expansion). This allowed for information disclosure.
991 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
992 INT_MIN/-1 -- value coerced to INT_MAX.
998 NM/01 Workround for PCRE version dependency in version reporting
1001 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1002 This fixes portability to compilers other than gcc, notably
1003 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1005 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1006 makefiles for portability to HP-UX and POSIX correctness.
1008 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1009 Also via indented variable definition in the Makefile.
1010 (Debugging by Oliver Heesakkers).
1012 PP/02 Restore caching of spamd results with expanded spamd_address.
1013 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1015 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1016 Improves build reliability. Fix from: Frank Elsner
1018 NM/02 Fix wide character breakage in the rfc2047 coding
1019 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1021 NM/03 Allow underscore in dnslist lookups
1022 Fixes bug 1026. Patch from Graeme Fowler
1024 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1025 Code patches from Adam Ciarcinski of NetBSD.
1027 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1030 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1031 is logged, to avoid truncation. Patch from John Horne.
1033 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1034 Patch from Jakob Hirsch.
1036 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1037 SQL string expansion failure details.
1038 Patch from Andrey Oktyabrski.
1040 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1041 Patch from Simon Arlott.
1043 PP/09 New lookups functionality failed to compile on old gcc which rejects
1044 extern declarations in function scope.
1045 Patch from Oliver Fleischmann
1047 PP/10 Use sig_atomic_t for flags set from signal handlers.
1048 Check getgroups() return and improve debugging.
1049 Fixed developed for diagnosis in bug 927 (which turned out to be
1052 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1053 Patch from Mark Zealey.
1055 PP/12 Bugzilla 1056: Improved spamd server selection.
1056 Patch from Mark Zealey.
1058 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1059 Based on patch from Heiko Schlittermann.
1061 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1062 Patch from Uwe Doering, sign-off by Michael Haardt.
1064 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1065 variable declaration deep within a block. Bug and patch from
1068 PP/15 lookups-Makefile IRIX compatibilty coercion.
1070 PP/16 Make DISABLE_DKIM build knob functional.
1072 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1073 Patch by Simon Arlott
1075 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1076 variable argument macros. Our copy now differs from upstream.
1082 TF/01 Failure to get a lock on a hints database can have serious
1083 consequences so log it to the panic log.
1085 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1086 controlled using the smtp_confirmation log selector.
1088 TF/03 Include the error message when we fail to unlink a spool file.
1090 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1091 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1092 for maintaining out-of-tree patches for some time.
1094 PP/01 Bugzilla 139: Documentation and portability issues.
1095 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1096 Handle per-OS dynamic-module compilation flags.
1098 PP/02 Let /dev/null have normal permissions.
1099 The 4.73 fixes were a little too stringent and complained about the
1100 permissions on /dev/null. Exempt it from some checks.
1101 Reported by Andreas M. Kirchwitz.
1103 PP/03 Report version information for many libraries, including
1104 Exim version information for dynamically loaded libraries. Created
1105 version.h, now support a version extension string for distributors
1106 who patch heavily. Dynamic module ABI change.
1108 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1109 privilege escalation vulnerability whereby the Exim run-time user
1110 can cause root to append content of the attacker's choosing to
1113 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1116 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1117 If dropping privileges for untrusted macros, we disabled normal logging
1118 on the basis that it would fail; for the Exim run-time user, this is not
1119 the case, and it resulted in successful deliveries going unlogged.
1120 Fixed. Reported by Andreas Metzler.
1126 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1127 only prepend for the Resent-* case. Fixes regression introduced in
1128 Exim 4.70 by NM/22 for Bugzilla 607.
1130 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1131 increasing numbers of administrators be bitten by this.
1133 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1135 PP/03 Bugzilla 994: added openssl_options main configuration option.
1137 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1139 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1141 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1143 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1144 without a peer certificate, leading to a segfault because of an
1145 assumption that peers always have certificates. Be a little more
1146 paranoid. Problem reported by Martin Tscholak.
1148 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1149 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1150 NB: ClamAV planning to remove STREAM in "middle of 2010".
1151 CL also introduces -bmalware, various -d+acl logging additions and
1152 more caution in buffer sizes.
1154 PP/09 Implemented reverse_ip expansion operator.
1156 PP/10 Bugzilla 937: provide a "debug" ACL control.
1158 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1160 PP/12 Bugzilla 973: Implement --version.
1162 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1164 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1166 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1168 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1169 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1170 ignore trailing whitespace.
1172 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1174 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1177 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1178 Notification from Dr Andrew Aitchison.
1180 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1181 ExtendedDetectionInfo response format.
1182 Notification from John Horne.
1184 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1187 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1188 XSL and documented dependency on system catalogs, with examples of how
1191 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1194 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1195 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1196 configuration file which is writeable by the Exim user or group.
1198 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1199 of configuration files to cover files specified with the -C option if
1200 they are going to be used with root privileges, not just the default
1203 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1204 option (effectively making it always true).
1206 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1207 files to be used while preserving root privileges.
1209 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1210 that rogue child processes cannot use them.
1212 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1213 run-time user, instead of root.
1215 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1216 Exim run-time user without dropping privileges.
1218 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1219 result string, instead of calling string_vformat() twice with the same
1222 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1223 for other users. Others should always drop root privileges if they use
1224 -C on the command line, even for a whitelisted configure file.
1226 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1228 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1234 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1235 $data_path, and $header_path variables; fixed documentation bugs and
1238 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1239 exipick to access non-standard spools, including the "frozen" queue
1242 NM/01 Bugzilla 965: Support mysql stored procedures.
1243 Patch from Alain Williams
1245 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1247 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1248 Patch from Andreas Metzler
1250 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1251 Patch from Kirill Miazine
1253 NM/05 Bugzilla 671: Added umask to procmail example.
1255 JJ/03 installed exipick 20100323.0, fixing doc bug
1257 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1258 directory. Notification and patch from Dan Rosenberg.
1260 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1262 TK/02 Improve log output when DKIM signing operation fails.
1264 MH/01 Treat the transport option dkim_domain as a colon separated
1265 list, not as a single string, and sign the message with each element,
1266 omitting multiple occurences of the same signer.
1268 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1269 Bugzilla 985, 986. Patch by Simon Arlott
1271 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1272 Patch by Simon Arlott
1274 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1275 MBX locking. Notification from Dan Rosenberg.
1281 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1283 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1285 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1287 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1289 NM/04 Bugzilla 915: Flag broken perl installation during build.
1295 TK/01 Added patch by Johannes Berg that expands the main option
1296 "spamd_address" if it starts with a dollar sign.
1298 TK/02 Write list of recipients to X-Envelope-Sender header when building
1299 the mbox-format spool file for content scanning (suggested by Jakob
1302 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1303 (http://www.dcc-servers.net/) support via dccifd. Activated by
1304 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1306 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1307 by Mark Daniel Reidel <mr@df.eu>.
1309 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1310 When building exim an external PCRE library is now needed -
1311 PCRE is a system library on the majority of modern systems.
1312 See entry on PCRE_LIBS in EDITME file.
1314 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1315 conversation. Added nologin parameter to request.
1316 Patch contributed by Kirill Miazine.
1318 TF/01 Do not log submission mode rewrites if they do not change the address.
1320 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1322 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1323 log files in place. Contributed by Roberto Lima.
1325 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1327 TF/03 Bugzilla 615: When checking the local_parts router precondition
1328 after a local_part_suffix or local_part_prefix option, Exim now
1329 does not use the address's named list lookup cache, since this
1330 contains cached lookups for the whole local part.
1332 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1333 Robert Millan. Documentation is in experimental-spec.txt.
1335 TF/04 Bugzilla 668: Fix parallel build (make -j).
1337 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1339 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1340 Patch provided by Jan Srzednicki.
1342 TF/05 Leading white space used to be stripped from $spam_report which
1343 wrecked the formatting. Now it is preserved.
1345 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1346 that they are available at delivery time.
1348 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1350 TF/08 TLS error reporting now respects the incoming_interface and
1351 incoming_port log selectors.
1353 TF/09 Produce a more useful error message if an SMTP transport's hosts
1354 setting expands to an empty string.
1356 NM/06 Bugzilla 744: EXPN did not work under TLS.
1357 Patch provided by Phil Pennock.
1359 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1360 Patch provided by Richard Godbee.
1362 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1363 acl_smtp_notquit, added index entry.
1365 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1366 Patch provided by Eugene Bujak.
1368 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1369 accept(). Patch provided by Maxim Dounin.
1371 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1372 Patch provided by Phil Pennock.
1374 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1376 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1377 Patch provided by Brad "anomie" Jorsch.
1379 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1380 Patch provided by Dean Brooks.
1382 TK/05 Add native DKIM support (does not depend on external libraries).
1384 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1385 Patch provided by Graeme Fowler.
1387 NM/16 Bugzilla 851: Documentation example syntax fix.
1389 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1391 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1394 NM/19 Bugzilla 745: TLS version reporting.
1395 Patch provided by Phil Pennock.
1397 NM/20 Bugzilla 167: bool: condition support.
1398 Patch provided by Phil Pennock.
1400 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1401 clients. Patch provided by Phil Pennock.
1403 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1404 Patch provided by Brad "anomie" Jorsch.
1406 NM/23 Bugzilla 687: Fix misparses in eximstats.
1407 Patch provided by Heiko Schlittermann.
1409 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1410 Patch provided by Heiko Schlittermann.
1412 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1413 plus update to original patch.
1415 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1417 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1418 Patch provided by David Brownlee.
1420 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1422 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1424 NM/30 Bugzilla 888: TLS documentation bugfixes.
1426 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1428 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1429 Unlike the original bugzilla I have changed all shell scripts in src tree.
1431 NM/33 Bugzilla 898: Transport filter timeout fix.
1432 Patch by Todd Rinaldo.
1434 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1435 Patch by Serge Demonchaux.
1437 NM/35 Bugzilla 39: Base64 decode bug fixes.
1438 Patch by Jakob Hirsch.
1440 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1442 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1444 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1446 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1452 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1453 ALT-N's libdkim that I have put here:
1454 http://duncanthrax.net/exim-experimental/
1456 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1457 were called 'true' and it seems that C99 defines that as a reserved
1458 keyword to be used with 'bool' variable types. That means you could
1459 not include C99-style headers which use bools without triggering
1460 build errors in sieve.c.
1462 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1463 as mailq or other aliases. Changed the --help handling significantly
1464 to do whats expected. exim_usage() emits usage/help information.
1466 SC/01 Added the -bylocaldomain option to eximstats.
1468 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1470 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1472 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1478 PH/01 Another patch from the Sieve maintainer.
1480 PH/02 When an IPv6 address is converted to a string for single-key lookup
1481 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1482 dots are used instead of colons so that keys in lsearch files need not
1483 contain colons. This was done some time before quoting was made available
1484 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1485 (notated using the quote facility) so as to distinguish them from IPv4
1486 keys. This meant that lookups for IP addresses in host lists did not work
1487 for iplsearch lookups.
1489 This has been fixed by arranging for IPv6 addresses to be expressed with
1490 colons if the lookup type is iplsearch. This is not incompatible, because
1491 previously such lookups could never work.
1493 The situation is now rather anomolous, since one *can* have colons in
1494 ordinary lsearch keys. However, making the change in all cases is
1495 incompatible and would probably break a number of configurations.
1497 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1500 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1501 conversion specification without a maximum field width, thereby enabling
1502 a rogue spamd server to cause a buffer overflow. While nobody in their
1503 right mind would setup Exim to query an untrusted spamd server, an
1504 attacker that gains access to a server running spamd could potentially
1505 exploit this vulnerability to run arbitrary code as the Exim user.
1507 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1508 $primary_hostname instead of what libspf2 thinks the hosts name is.
1510 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1511 a directory entry by the name of the lookup key. Previously, if a
1512 symlink pointed to a non-existing file or a file in a directory that
1513 Exim lacked permissions to read, a lookup for a key matching that
1514 symlink would fail. Now it is enough that a matching directory entry
1515 exists, symlink or not. (Bugzilla 503.)
1517 PH/03 The body_linecount and body_zerocount variables are now exported in the
1520 PH/04 Added the $dnslist_matched variable.
1522 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1523 This means they are set thereafter only if the connection becomes
1526 PH/06 Added the client_condition to authenticators so that some can be skipped
1527 by clients under certain conditions.
1529 PH/07 The error message for a badly-placed control=no_multiline_responses left
1530 "_responses" off the end of the name.
1532 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1534 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1535 (without spaces) instead of just copying the configuration text.
1537 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1539 PH/11 Added $max_received_linelength.
1541 PH/12 Added +ignore_defer and +include_defer to host lists.
1543 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1544 way in which PCRE > 7.0 is built.
1546 PH/14 Implemented queue_only_load_latch.
1548 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1549 MAIL command. The effect was to mangle the value on 64-bit systems.
1551 PH/16 Another patch from the Sieve maintainer.
1553 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1555 PH/18 If a system quota error occurred while trying to create the file for
1556 a maildir delivery, the message "Mailbox is full" was not appended to the
1557 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1558 only to a quota excession during the actual writing of the file.
1560 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1561 characters?) which causes problems in log lines. The DN values are now
1562 passed through string_printing() before being added to log lines.
1564 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1565 and InterBase are left for another time.)
1567 PH/21 Added message_body_newlines option.
1569 PH/22 Guard against possible overflow in moan_check_errorcopy().
1571 PH/23 POSIX allows open() to be a macro; guard against that.
1573 PH/24 If the recipient of an error message contained an @ in the local part
1574 (suitably quoted, of course), incorrect values were put in $domain and
1575 $local_part during the evaluation of errors_copy.
1581 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1582 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1585 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1586 issue a MAIL command.
1588 PH/02 In an ACL statement such as
1590 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1592 if a client was not listed at all, or was listed with a value other than
1593 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1594 the condition was not true (as it should be), so access was not denied.
1595 The bug was that the ! inversion was incorrectly passed on to the second
1596 item. This has been fixed.
1598 PH/03 Added additional dnslists conditions == and =& which are different from
1599 = and & when the dns lookup returns more than one IP address.
1601 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1602 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1604 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1605 FSYNC, which compiles an option called disable_fsync that allows for
1606 bypassing fsync(). The documentation is heavily laced with warnings.
1608 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1610 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1611 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1612 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1613 including adding "make clean"; (3) Added -fPIC when compiling the test
1614 dynamically loaded module, to get rid of a warning.
1616 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1617 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1618 = 0s. The bug is otherwise harmless.
1620 PH/07 There was a bug in the dovecot authenticator such that the value of
1621 $auth1 could be overwritten, and so not correctly preserved, after a
1622 successful authentication. This usually meant that the value preserved by
1623 the server_setid option was incorrect.
1625 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1627 PH/09 Installed PCRE release 7.0.
1629 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1630 run for batched SMTP input. It is now run at the start of every message
1631 in the batch. While fixing this I discovered that the process information
1632 (output by running exiwhat) was not always getting set for -bs and -bS
1633 input. This is fixed, and it now also says "batched" for BSMTP.
1635 PH/11 Added control=no_pipelining.
1637 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1638 patch, slightly modified), and move the expansion of helo_data till after
1639 the connection is made in the smtp transport (so it can use these
1642 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1644 PH/14 Added log_selector = +pid.
1646 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1648 PH/16 Add ${if forany and ${if forall.
1650 PH/17 Added dsn_from option to vary the From: line in DSNs.
1652 PH/18 Flush SMTP output before performing a callout, unless control =
1653 no_callout_flush is set.
1655 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1656 was true (the default) a successful delivery failed to delete the retry
1657 item, thus causing premature timeout of the address. The bug is now
1660 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1662 PH/21 Long custom messages for fakedefer and fakereject are now split up
1663 into multiline reponses in the same way that messages for "deny" and
1664 other ACL rejections are.
1666 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1667 with slight modification.
1669 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1670 draft, changing the syntax and factoring some duplicate code".
1672 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1673 for deliveries of the second and subsequent messages over the same SMTP
1676 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1677 ${reduce, with only minor "tidies".
1679 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1681 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1682 expansion side effects.
1684 PH/27 When a message times out after an over-quota error from an Exim-imposed
1685 quota, the bounce message says "mailbox is full". This message was not
1686 being given when it was a system quota that was exceeded. It now should
1689 MH/03 Made $recipients available in local_scan(). local_scan() already has
1690 better access to the recipient list through recipients_list[], but
1691 $recipients can be useful in postmaster-provided expansion strings.
1693 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1694 in the case of a MAIL command with additional options following the
1695 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1696 were accidentally chopped off.
1698 PH/29 SMTP synchronization checks are implemented when a command is read -
1699 there is a check that no more input is waiting when there shouldn't be
1700 any. However, for some commands, a delay in an ACL can mean that it is
1701 some time before the response is written. In this time, more input might
1702 arrive, invalidly. So now there are extra checks after an ACL has run for
1703 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1704 pipelining has not been advertised.
1706 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1708 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1709 returned in the SMTP response when smtp_return_error_details was false.
1710 This has been fixed.
1712 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1713 instead of the C I/O that was originally supplied, because problems were
1714 reported on Solaris.
1716 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1717 Exim which did not show up earlier: it was assuming that a call to
1718 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1719 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1720 was a macro that became an assignment, so it seemed to work. This has
1721 changed to a proper function call with a void return, hence the compile
1722 error. Exim's code has been fixed.
1724 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1727 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1729 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1731 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1732 because it turns out to be too big or there is a timeout) there is no
1733 "Completed" line in the log. When some messages of this type were
1734 selected by exigrep, they were listed as "not completed". Others were
1735 picked up by some special patterns. I have improved the selection
1736 criteria to be more general.
1738 PH/38 The host_find_failed option in the manualroute router can now be set
1739 to "ignore", to completely ignore a host whose IP address cannot be
1740 found. If all hosts are ignored, the behaviour is controlled by the new
1741 host_all_ignored option.
1743 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1744 homing or because of multiple MX records with /mx) generated more than
1745 one IP address, and the following item turned out to be the local host,
1746 all the secondary addresses of the first item were incorrectly removed
1747 from the list, along with the local host and any following hosts (which
1748 is what is supposed to happen).
1750 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1751 whoever called Exim into the -H file. In the case of the daemon it was
1752 behaving confusingly. When first started, it used values for whoever
1753 started the daemon, but after a SIGHUP it used the Exim user (because it
1754 calls itself on a restart). I have changed the code so that it now always
1757 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1758 message are rejected with the same error (e.g. no authentication or bad
1759 sender address), and a DATA command is nevertheless sent (as can happen
1760 with PIPELINING or a stupid MUA), the error message that was given to the
1761 RCPT commands is included in the rejection of the DATA command. This is
1762 intended to be helpful for MUAs that show only the final error to their
1765 PH/42 Another patch from the Sieve maintainer.
1767 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1768 Eximstats - Fixed some broken HTML links and added missing column headers
1770 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1771 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1773 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1775 PH/43 Yet another patch from the Sieve maintainer.
1777 PH/44 I found a way to check for a TCP/IP connection going away before sending
1778 the response to the final '.' that terminates a message, but only in the
1779 case where the client has not sent further data following the '.'
1780 (unfortunately, this is allowed). However, in many cases there won't be
1781 any further data because there won't be any more messages to send. A call
1782 to select() can be used: if it shows that the input is "ready", there is
1783 either input waiting, or the socket has been closed. An attempt to read
1784 the next input character can distinguish the two cases. Previously, Exim
1785 would have sent an OK response which the client would never have see.
1786 This could lead to message repetition. This fix should cure that, at
1787 least in a lot of common cases.
1789 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1790 advertised in response to EHLO.
1796 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1797 fixed by 4.65/MH/01 (is this a record?) are fixed:
1799 (i) An empty string was always treated as zero by the numeric comparison
1800 operators. This behaviour has been restored.
1802 (ii) It is documented that the numeric comparison operators always treat
1803 their arguments as decimal numbers. This was broken in that numbers
1804 starting with 0 were being interpreted as octal.
1806 While fixing these problems I realized that there was another issue that
1807 hadn't been noticed. Values of message_size_limit (both the global option
1808 and the transport option) were treated as octal if they started with 0.
1809 The documentation was vague. These values are now always treated as
1810 decimal, and I will make that clear in the documentation.
1816 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1817 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1820 MH/01 Don't check that the operands of numeric comparison operators are
1821 integers when their expansion is in "skipping" mode (fixes bug
1822 introduced by 4.64-PH/07).
1824 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1825 child addresses, Exim now panics and dies. Previously, because the count
1826 is held in a short int, deliveries were likely to be lost. As such a
1827 large number of recipients for a single message is ridiculous
1828 (performance will be very, very poor), I have chosen to impose a limit
1829 rather than extend the field.
1835 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1836 leftover -K file (the existence of which was triggered by #402).
1837 While we were at it, introduced process PID as part of the -K
1838 filename. This should rule out race conditions when creating
1841 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1842 processing considerably. Previous code took too long for large mails,
1843 triggering a timeout which in turn triggers #401.
1845 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1846 in the DK code in transports.c. sendfile() is not really portable,
1847 hence the _LINUX specificness.
1849 TF/01 In the add_headers option to the mail command in an Exim filter,
1850 there was a bug that Exim would claim a syntax error in any
1851 header after the first one which had an odd number of characters
1854 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1855 callout verification, Exim cached a "reject" for the entire domain. This
1856 is correct for most verifications, but it is not correct for a recipient
1857 verification with use_sender or use_postmaster set, because in that case
1858 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1859 case of MAIL FROM:<> rejection from other early rejections (e.g.
1860 rejection of HELO). When verifying a recipient using a non-null MAIL
1861 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1862 Whatever the result of the callout, the value of the domain cache is
1863 left unchanged (for any other kind of callout, getting as far as trying
1864 RCPT means that the domain itself is ok).
1866 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1869 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1870 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1871 session, the daemon ignores ECONNECT errors and logs others; it now
1872 ignores EPIPE as well.
1874 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1875 (quoted-printable decoding).
1877 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1878 later the small subsequent patch to fix an introduced bug.
1880 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1882 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1884 PH/08 An error is now given if message_size_limit is specified negative.
1886 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1887 to be given (somewhat) arbitrary names.
1889 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1892 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1893 miscellaneous code fixes
1895 PH/10 Added the log_reject_target ACL modifier to specify where to log
1898 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1899 hostname. This is wrong, because it relates to the incoming message (and
1900 probably the interface on which it is arriving) and not to the outgoing
1901 callout (which could be using a different interface). This has been
1902 changed to use the value of the helo_data option from the smtp transport
1903 instead - this is what is used when a message is actually being sent. If
1904 there is no remote transport (possible with a router that sets up host
1905 addresses), $smtp_active_hostname is used.
1907 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1908 tweaks were necessary in order to get it to work (see also 21 below):
1909 (a) The code assumed that strncpy() returns a negative number on buffer
1910 overflow, which isn't the case. Replaced with Exim's string_format()
1912 (b) There were several signed/unsigned issues. I just did the minimum
1913 hacking in of casts. There is scope for a larger refactoring.
1914 (c) The code used strcasecmp() which is not a standard C function.
1915 Replaced with Exim's strcmpic() function.
1916 (d) The code set only $1; it now sets $auth1 as well.
1917 (e) A simple test gave the error "authentication client didn't specify
1918 service in request". It would seem that Dovecot has changed its
1919 interface. Fortunately there's a specification; I followed it and
1920 changed what the client sends and it appears to be working now.
1922 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1925 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1926 address A is aliased to B and C, where B exists and C does not. Without
1927 -v the output is "A verified" because verification stops after a
1928 successful redirection if more than one address is generated. However,
1929 with -v the child addresses are also verified. Exim was outputting "A
1930 failed to verify" and then showing the successful verification for C,
1931 with its parentage. It now outputs "B failed to verify", showing B's
1932 parentage before showing the successful verification of C.
1934 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1935 look up a TXT record in a specific list after matching in a combined
1938 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1939 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1940 they consult the DNS. I had assumed they would set it the way they
1941 wanted; and indeed my experiments on Linux seem to show that in some
1942 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1943 To be on the safe side, however, I have now made the interface to
1944 host_find_byname() similar to host_find_bydns(), with an argument
1945 containing the DNS resolver options. The host_find_byname() function now
1946 sets these options at its start, just as host_find_bydns() does. The smtp
1947 transport options dns_qualify_single and dns_search_parents are passed to
1948 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1949 of host_find_byname() use the default settings of RES_DEFNAMES
1950 (qualify_single) but not RES_DNSRCH (search_parents).
1952 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1953 spool_read_header() do less string testing, by means of a preliminary
1954 switch on the second character of optional "-foo" lines. (This is
1955 overdue, caused by the large number of possibilities that now exist.
1956 Originally there were few.) While I was there, I also converted the
1957 str(n)cmp tests so they don't re-test the leading "-" and the first
1958 character, in the hope this might squeeze out yet more improvement.
1960 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1961 flag allowing group syntax was set by the header_syntax check but not
1962 turned off, possible causing trouble later; (2) The flag was not being
1963 set at all for the header_verify test, causing "group"-style headers to
1964 be rejected. I have now set it in this case, and also caused header_
1965 verify to ignore an empty address taken from a group. While doing this, I
1966 came across some other cases where the code for allowing group syntax
1967 while scanning a header line wasn't quite right (mostly, not resetting
1968 the flag correctly in the right place). These bugs could have caused
1969 trouble for malformed header lines. I hope it is now all correct.
1971 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1972 with the "reply" argument non-NULL. The code, however (which originally
1973 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1974 but it didn't always do it. This confused somebody who was copying the
1975 code for some other use. I have removed all the tests.
1977 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1978 feature that was used to support insecure browsers during the U.S. crypto
1979 embargo. It requires special client support, and Exim is probably the
1980 only MTA that supported it -- and would never use it because real RSA is
1981 always available. This code has been removed, because it had the bad
1982 effect of slowing Exim down by computing (never used) parameters for the
1983 RSA_EXPORT functionality.
1985 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1986 authenticator to fail if there's a tab character in the incoming data
1987 (there should never be unless someone is messing about, as it's supposed
1988 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1989 if the connection is using TLS or if the remote IP is the same as the
1990 local IP, and the "valid-client-cert option" if a client certificate has
1993 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1994 authenticators. This can be used for authorization after authentication
1995 succeeds. (In the case of plaintext, it servers for both authentication
1998 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1999 if any retry times were supplied.
2001 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2002 connection, where there is no client IP address to check. In this
2003 situation, the verify now always succeeds.
2005 PH/25 Applied John Jetmore's -Mset patch.
2007 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2009 PH/27 In a string expansion for a processed (not raw) header when multiple
2010 headers of the same name were present, leading whitespace was being
2011 removed from all of them, but trailing whitespace was being removed only
2012 from the last one. Now trailing whitespace is removed from each header
2013 before concatenation. Completely empty headers in a concatenation (as
2014 before) are ignored.
2016 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2017 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2019 PH/29 [Removed. This was a change that I later backed out, and forgot to
2020 correct the ChangeLog entry (that I had efficiently created) before
2021 committing the later change.]
2023 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2024 address errors (4xx response to RCPT) over the same connection as other
2025 messages routed to the same hosts. Such deliveries are always "forced",
2026 so retry times are not inspected. This resulted in far too many retries
2027 for the affected addresses. The effect occurred only when there were more
2028 hosts than the hosts_max_try setting in the smtp transport when it had
2029 the 4xx errors. Those hosts that it had tried were not added to the list
2030 of hosts for which the message was waiting, so if all were tried, there
2031 was no problem. Two fixes have been applied:
2033 (i) If there are any address or message errors in an SMTP delivery, none
2034 of the hosts (tried or untried) are now added to the list of hosts
2035 for which the message is waiting, so the message should not be a
2036 candidate for sending over the same connection that was used for a
2037 successful delivery of some other message. This seems entirely
2038 reasonable: after all the message is NOT "waiting for some host".
2039 This is so "obvious" that I'm not sure why it wasn't done
2040 previously. Hope I haven't missed anything, but it can't do any
2041 harm, as the worst effect is to miss an optimization.
2043 (ii) If, despite (i), such a delivery is accidentally attempted, the
2044 routing retry time is respected, so at least it doesn't keep
2045 hammering the server.
2047 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2048 in ${readsocket because some servers need this prod.
2050 PH/32 Added some extra debug output when updating a wait-xxx database.
2052 PH/33 The hint "could be header name not terminated by colon", which has been
2053 given for certain expansion errors for a long time, was not being given
2054 for the ${if def:h_colon_omitted{... case.
2056 PH/34 The spec says: "With one important exception, whenever a domain list is
2057 being scanned, $domain contains the subject domain." There was at least
2058 one case where this was not true.
2060 PH/35 The error "getsockname() failed: connection reset by peer" was being
2061 written to the panic log as well as the main log, but it isn't really
2062 panic-worthy as it just means the connection died rather early on. I have
2063 removed the panic log writing for the ECONNRESET error when getsockname()
2066 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2067 runs only) independently of the message's sender address. This meant
2068 that, if the 4xx error was in fact related to the sender, a different
2069 message to the same recipient with a different sender could confuse
2070 things. In particualar, this can happen when sending to a greylisting
2071 server, but other circumstances could also provoke similar problems.
2072 I have changed the default so that the retry time for these errors is now
2073 based a combination of the sender and recipient addresses. This change
2074 can be overridden by setting address_retry_include_sender=false in the
2077 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2078 remote server are returned as part of bounce messages. This was not
2079 happening for LMTP over a pipe (the lmtp transport), but now it is the
2080 same for both kinds of LMTP.
2082 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2083 in header lines that were in fact CNAMEs. This is no longer the case.
2085 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2086 and queue runs started by the daemon processed all messages. This has
2087 been fixed so that -R and -S can now usefully be given with -q<time>.
2089 PH/40 Import PCRE release 6.7 (fixes some bugs).
2091 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2093 PH/42 Give an error if -q is specified more than once.
2095 PH/43 Renamed the variables $interface_address and $interface_port as
2096 $received_ip_address and $received_port, to make it clear that these
2097 values apply to message reception, and not to the outgoing interface when
2098 a message is delivered. (The old names remain recognized, of course.)
2100 PH/44 There was no timeout on the connect() call when using a Unix domain
2101 socket in the ${readsocket expansion. There now is.
2103 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2104 be meaningful with "accept".
2106 SC/01 Eximstats V1.43
2107 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2109 SC/02 Eximstats V1.44
2110 Use a glob alias rather than an array ref in the generated
2111 parser. This improves both readability and performance.
2113 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2114 Collect SpamAssassin and rejection statistics.
2115 Don't display local sender or destination tables unless
2116 there is data to show.
2117 Added average volumes into the top table text output.
2119 SC/04 Eximstats V1.46
2120 Collect data on the number of addresses (recipients)
2121 as well as the number of messages.
2123 SC/05 Eximstats V1.47
2124 Added 'Message too big' to the list of mail rejection
2125 reasons (thanks to Marco Gaiarin).
2127 SC/06 Eximstats V1.48
2128 Mainlog lines which have GMT offsets and are too short to
2129 have a flag are now skipped.
2131 SC/07 Eximstats V1.49 (Alain Williams)
2132 Added the -emptyok flag.
2134 SC/08 Eximstats V1.50
2135 Fixes for obtaining the IP address from reject messages.
2137 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2138 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2139 whitesspace changes from 4.64-PH/27
2141 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2144 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2145 are found, allow negative numbers in numeric criteria)
2147 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2149 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2150 to match changes made in 4.64-PH/43
2152 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2154 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2155 "domain" is actually an IP address, and force a failure. This locks out
2156 those revolvers/nameservers that support "A-for-A" lookups, in
2157 contravention of the specifications.
2159 PH/48 When a host name was looked up from an IP address, and the subsequent
2160 forward lookup of the name timed out, the host name was left in
2161 $sender_host_name, contrary to the specification.
2163 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2164 restricted to single-key lookups, Exim was not diagnosing an error if
2165 * or *@ was used with a query-style lookup.
2167 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2169 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2170 long ago, but noone interested enough thought of it. Let's just say that
2171 the "1.1" means that there are some new functions that weren't there at
2172 some point in the past.
2174 PH/51 Error processing for expansion failure of helo_data from an smtp
2175 transport during callout processing was broken.
2177 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2178 tested/used via the -bh/-bhc/-bs options.
2180 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2181 bug, fixed in subsequent PCRE releases).
2183 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2184 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2186 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2192 SC/01 Use a glob alias rather than an array ref in eximstats generated
2193 parser. This improves both readability and performance.
2195 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2196 Don't display local sender or destination tables in eximstats unless
2197 there is data to show.
2198 Added average volumes into the eximstats top table text output.
2200 SC/03 Collect data on the number of addresses (recipients) as well
2201 as the number of messages in eximstats.
2203 TF/01 Correct an error in the documentation for the redirect router. Exim
2204 does (usually) call initgroups() when daemonizing.
2206 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2207 with consistent privilege compared to when running as a daemon.
2209 TF/03 Note in the spec that $authenticated_id is not set for local
2210 submissions from trusted users.
2212 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2213 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2215 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2216 by adding some example configuration directives to the default
2217 configuration file. A little bit of work is required to uncomment the
2218 directives and define how usernames and passwords are checked, but
2219 there is now a framework to start from.
2221 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2222 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2223 without this. I don't know how relevant this is to other LDAP libraries.
2225 PH/02 Add the verb name to the "unknown ACL verb" error.
2227 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2229 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2231 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2232 directory not expanded when it should be if an expanded home directory
2233 was set for the address (which is overridden by the transport).
2235 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2238 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2239 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2240 because it is too late at that time, and has no effect.
2242 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2243 security issue with \' (bugzilla #107). I could not use the
2244 PQescapeStringConn() function, because it needs a PGconn value as one of
2247 PH/08 When testing addresses using -bt, indicate those final addresses that
2248 are duplicates that would not cause an additional delivery. At least one
2249 person was confused, thinking that -bt output corresponded to deliveries.
2250 (Suppressing duplicates isn't a good idea as you lose the information
2251 about possibly different redirections that led to the duplicates.)
2253 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2254 systems where poll() doesn't work, in particular OS X.
2256 PH/10 Added more information to debugging output for retry time not reached.
2258 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2259 operations in malware.c.
2261 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2264 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2265 syslog_duplication set false, log lines that would normally be written
2266 both the the main log and to the reject log were not written to syslog at
2269 PH/14 In the default configuration, change the use of "message" in ACL warn
2270 statements to "add_header".
2272 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2273 not followed by a command (e.g. "seen endif").
2275 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2276 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2279 PH/17 Added extra conditions to the default value of delay_warning_condition
2283 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2284 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2285 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2288 The Auto-Submitted: and various List- headers are standardised, whereas I
2289 don't think Precedence: ever was.
2291 PH/18 Refactored debugging code in route_finduser() to show more information,
2292 in particular, the error code if getpwnam() issues one.
2294 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2295 This is apparently needed in addition to the PH/07 change above to avoid
2296 any possible encoding problems.
2298 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2299 but not after initializing Perl.
2301 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2302 output them only if debugging. By default they are written stderr,
2303 apparently, which is not desirable.
2305 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2308 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2311 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2313 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2314 authenticated or an ident call has been made. Suppress the default
2315 values for $authenticated_id and $authenticated_sender (but permit -oMai
2316 and -oMas) when testing with -bh.
2318 PH/24 Re-jigged the order of the tests in the default configuration so that the
2319 tests for valid domains and recipients precede the DNS black list and CSA
2320 tests, on the grounds that those ones are more expensive.
2322 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2323 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2324 This bug exists in every version of Exim that I still have, right back to
2327 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2328 However, an attempt to turn on case-sensitivity in a regex key by
2329 including (?-i) didn't work because the subject string was already
2330 lowercased, and the effects were non-intuitive. It turns out that a
2331 one-line patch can be used to allow (?-i) to work as expected.
2337 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2338 other effects) broke the use of negated acl sub-conditions.
2340 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2343 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2344 "Deny" causes Exim to reject the incoming connection with a 554 error.
2345 Unfortunately, if there is a major crisis, such as a disk failure,
2346 tcp-wrappers gives "deny", whereas what one would like would be some
2347 kind of temporary error. A kludge has been added to help with this.
2348 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2349 554 error is used if errno is still zero or contains ENOENT (which occurs
2350 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2353 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2355 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2356 errors. Otherwise a message that provokes a temporary error (when other
2357 messages do not) can cause a whole host to time out.
2359 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2360 addresses were routed directly to files or pipes from a redirect router.
2361 File deliveries just didn't batch; pipe deliveries might have suffered
2364 PH/06 A failure to get a lock for a hints database would erroneously always say
2365 "Failed to get write lock", even when it was really a read lock.
2367 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2368 of 0600. This has been changed to use the value of the lockfile_mode
2369 option (which defaults to 0600).
2371 PH/08 Applied small patch from the Sieve maintainer.
2373 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2374 folder from quota calculations, a direct delivery into this folder messed
2375 up the contents of the maildirsize file. This was because the regex was
2376 used only to exclude .Trash (or whatever) when the size of the mailbox
2377 was calculated. There was no check that a delivery was happening into an
2378 excluded directory. This bug has been fixed by ignoring all quota
2379 processing for deliveries into excluded directories.
2381 PH/10 Added the maildirfolder_create_regex option to appendfile.
2387 PH/01 The code for finding all the local interface addresses on a FreeBSD
2388 system running IPv6 was broken. This may well have applied to all BSD
2389 systems, as well as to others that have similar system calls. The broken
2390 code found IPv4 interfaces correctly, but gave incorrect values for the
2391 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2392 that it would not match correctly against @[] and not recognize the IPv6
2395 PH/02 The ipliteral router was not recognizing addresses of the form user@
2396 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2398 PH/03 Added disable_ipv6.
2400 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2401 decoded form, because it is most often used to construct To: headers
2402 lines in autoreplies, and the decoded form may well be syntactically
2403 invalid. However, $reply_address has leading white space removed, and all
2404 newlines turned into spaces so that the autoreply transport does not
2407 PH/05 If group was specified without a user on a router, and no group or user
2408 was specified on a transport, the group from the router was ignored.
2410 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2411 for visible compile-time settings that can be used to change these
2412 numbers, for those that want even more. Backwards compatibility with old
2413 spool files has been maintained. However, going back to a previous Exim
2414 release will lost any variables that are in spool files.
2416 PH/07 Two small changes when running in the test harness: increase delay when
2417 passing a TCP/IP connection to a new process, in case the original
2418 process has to generate a bounce, and remove special handling of
2419 127.0.0.2 (sic), which is no longer necessary.
2421 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2422 be the same on different OS.
2424 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2427 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2428 whether --show-vars was specified or not
2430 JJ/02 exipick: Added support for new ACL variable spool format introduced
2433 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2434 syntactically invalid From: or Reply-to: line, and a filter used this to
2435 generate an autoreply, and therefore failed to obtain an address for the
2436 autoreply, Exim could try to deliver to a non-existent relative file
2437 name, causing unrelated and misleading errors. What now happens is that
2438 it logs this as a hard delivery error, but does not attempt to create a
2441 PH/11 The exinext utility has a -C option for testing purposes, but although
2442 the given file was scanned by exinext itself; it wasn't being passed on
2443 when Exim was called.
2445 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2446 an end-of-file indication when reading a command response.
2448 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2449 compiled. In many other places in Exim, IPv6 addresses are always
2450 recognized, so I have changed this. It also means that IPv4 domain
2451 literals of the form [IPV4:n.n.n.n] are now always recognized.
2453 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2454 used if the router is not running as root, for example, when verifying at
2455 ACL time, or when using -bh. The debugging output from this situation was
2456 non-existent - all you got was a failure to exec. I have made two
2459 (a) Failures to set uid/gid, the current directory, or a process leader
2460 in a subprocess such as that created by queryprogram now generate
2461 suitable debugging ouput when -d is set.
2463 (b) The queryprogram router detects when it is not running as root,
2464 outputs suitable debugging information if -d is set, and then runs
2465 the subprocess without attempting to change uid/gid.
2467 PH/15 Minor change to Makefile for building test_host (undocumented testing
2470 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2471 additional section of a DNS packet that returns MX or SRV records.
2472 Instead, it always explicitly searches for A/AAAA records. This avoids
2473 major problems that occur when a DNS server includes only records of one
2474 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2475 fixed another bug: if SRV records were looked up and the corresponding
2476 address records were *not* found in the additional section, the port
2477 values from the SRV records were lost.
2479 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2480 using the correct key (the original address) when searching the retry
2481 rules in order to find which one to use for generating the retry hint.
2483 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2484 adding the default one. Similarly, if it contains a Reply-To: header, the
2485 errors_reply_to option, if set, is not used.
2487 PH/19 When calculating a retry time, Exim used to measure the "time since
2488 failure" by looking at the "first failed" field in the retry record. Now
2489 it does not use this if it is later than than the arrival time of the
2490 message. Instead it uses the arrival time. This makes for better
2491 behaviour in cases where some deliveries succeed, thus re-setting the
2492 "first failed" field. An example is a quota failure for a huge message
2493 when small messages continue to be delivered. Without this change, the
2494 "time since failure" will always be short, possible causing more frequent
2495 delivery attempts for the huge message than are intended.
2496 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2498 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2499 $1, $2, $3) because the numerical variables can be reset during some
2500 expansion items (e.g. "match"), thereby losing the authentication data.
2502 PH/21 Make -bV show the size of off_t variables so that the test suite can
2503 decide whether to run tests for quotas > 2G.
2505 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2506 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2507 filecount value is greater than 2G or if a quota value is greater than 2G
2508 on a system where the size of off_t is not greater than 4, a panic error
2511 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2512 never match. The debug and -bh output now contains an explicit error
2513 message indicating a malformed IPv4 address or mask.
2515 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2516 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2517 PH/23 above applies.
2519 PH/25 Do not write to syslog when running in the test harness. The only
2520 occasion when this arises is a failure to open the main or panic logs
2521 (for which there is an explicit test).
2523 PH/26 Added the /no_tell option to "control=freeze".
2525 PH/27 If a host name lookup failed very early in a connection, for example, if
2526 the IP address matched host_lookup and the reverse lookup yielded a name
2527 that did not have a forward lookup, an error message of the form "no IP
2528 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2529 could be logged. Now it outputs the IP address instead of "NULL".
2531 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2532 allows the sender and the authenticated sender to be set when
2533 submitting a message from within Exim. Since child_open_exim() is
2534 documented for local_scan(), the new function should be too.
2536 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2537 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2538 results in an empty string is now treated as unset.
2540 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2542 PH/31 Added sender_verify_fail logging option.
2544 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2545 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2546 tidied the source and removed it altogether.
2548 PH/33 When a queue run was abandoned because the load average was too high, a
2549 log line was always written; now it is written only if the queue_run log
2550 selector is set. In addition, the log line for abandonment now contains
2551 information about the queue run such as the pid. This is always present
2552 in "start" and "stop" lines but was omitted from the "abandon" line.
2554 PH/34 Omit spaces between a header name and the colon in the error message that
2555 is given when verify = headers_syntax fails (if there are lots of them,
2556 the message gets confusing).
2558 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2559 names, as there are now some PTR records that contain slashes. This check
2560 is only to protect against broken name servers that fall over on strange
2561 characters, so the fact that it applies to all lookups doesn't matter.
2563 PH/36 Now that the new test suite is complete, we can remove some of the
2564 special code in Exim that was needed for the old test suite. For example,
2565 sorting DNS records because real resolvers return them in an arbitrary
2566 order. The new test suite's fake resolver always returns records in the
2569 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2570 bounces) except when queue_only is set, to avoid logging races between
2571 the different processes.
2573 PH/38 Panic-die if .include specifies a non-absolute path.
2575 PH/39 A tweak to the "H" retry rule from its user.
2577 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2578 a label. They prevented compilation on older perls.
2580 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2581 a warning to be raised on newish perls.
2583 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2584 on queue. Changes to match documented behaviour of showing count of
2585 messages matching specified criteria.
2587 PH/40 Changed the default ident timeout from 30s to 5s.
2589 PH/41 Added support for the use of login_cap features, on those BSD systems
2590 that have them, for controlling the resources used by pipe deliveries.
2592 PH/42 The content-scanning code uses fopen() to create files in which to put
2593 message data. Previously it was not paying any attention to the mode of
2594 the files. Exim runs with umask(0) because the rest of the code creates
2595 files with open(), and sets the required mode explicitly. Thus, these
2596 files were ending up world-writeable. This was not a big issue, because,
2597 being within the spool directory, they were not world-accessible. I have
2598 created a function called modefopen, which takes an additional mode
2599 argument. It sets umask(777), creates the file, chmods it to the required
2600 mode, then resets the umask. All the relevant calls to fopen() in the
2601 content scanning code have been changed to use this function.
2603 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2604 to 24 hours. This avoids potential overflow problems when processing G
2605 and H retry rules. I suspect nobody ever tinkers with this value.
2607 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2609 PH/45 When the plaintext authenticator is running as a client, the server's
2610 challenges are checked to ensure they are valid base64 strings. By
2611 default, the authentication attempt is cancelled if an invalid string is
2612 received. Setting client_ignore_invalid_base64 true ignores these errors.
2613 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2614 they are received. Thus, the responses can be made to depend on the
2615 challenges. If an invalid string is ignored, an empty string is placed in
2618 PH/46 Messages that are created by the autoreply transport now contains a
2619 References: header, in accordance with RFCs 2822 and 3834.
2621 PH/47 Added authenticated_sender_force to the smtp transport.
2623 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2625 PH/49 Installed latest patch from the Sieve maintainer.
2627 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2628 was also set, the appendfile transport was unnecessarily scanning a
2629 directory of message files (e.g. for maildir delivery) to find the count
2630 of files (along with the size), even though it did not need this
2631 information. It now does the scan only if it needs to find either the
2632 size of the count of files.
2634 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2636 PH/52 Two bugs concerned with error handling when the smtp transport is
2639 (i) Exim was not creating retry information for temporary errors given
2640 for individual recipients after the DATA command when the smtp transport
2641 was used in LMTP mode. This meant that they could be retried too
2642 frequently, and not timed out correctly.
2644 (ii) Exim was setting the flag that allows error details to be returned
2645 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2646 recipients that were returned after the DATA command.
2648 PH/53 This is related to PH/52, but is more general: for any failing address,
2649 when detailed error information was permitted to be returned to the
2650 sender, but the error was temporary, then after the final timeout, only
2651 "retry timeout exceeded" was returned. Now it returns the full error as
2652 well as "retry timeout exceeded".
2654 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2655 do this, and (what is worse) MTAs that accept it.
2657 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2658 will now be deprecated.
2660 PH/56 New os.c-cygwin from the Cygwin maintainer.
2662 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2663 formats (previously only available in exim formats via -bpr, -bpru,
2664 and -bpra. Now also available in native and exiqgrep formats)
2666 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2667 with very large, slow to parse queues
2669 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2671 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2673 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2674 responses to authentication challenges, though it was showing the
2675 challenges; (ii) I've removed the CR characters from the debug output for
2678 PH/58 Allow for the insertion of a newline as well as a space when a string
2679 is turned into more than one encoded-word during RFC 2047 encoding. The
2680 Sieve code now uses this.
2682 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2683 data_4xx, lost_connection, tls_required.
2685 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2686 message was being sent as an SMTP response.
2688 PH/61 Add -l and -k options to exicyclog.
2690 PH/62 When verifying, if an address was redirected to one new address, so that
2691 verification continued, and the new address failed or deferred after
2692 having set something in $address_data, the value of $address_data was not
2693 passed back to the ACL. This was different to the case when no
2694 redirection occurred. The value is now passed back in both cases.
2696 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2697 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2698 use login_cap.h, so on its own it isn't the distinguishing feature. The
2699 new name refers directly to the setclassresources() function.
2701 PH/65 Added configuration files for NetBSD3.
2703 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2705 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2706 is preferred over IPv4.
2708 PH/68 The bounce_return_message and bounce_return_body options were not being
2709 honoured for bounces generated during the reception of non-SMTP messages.
2710 In particular, this applied to messages rejected by the ACL. This bug has
2711 been fixed. However, if bounce_return_message is true and bounce_return_
2712 body is false, the headers that are returned for a non-SMTP message
2713 include only those that have been read before the error was detected.
2714 (In the case of an ACL rejection, they have all been read.)
2716 PH/69 The HTML version of the specification is now built in a directory called
2717 spec_html instead of spec.html, because the latter looks like a path with
2718 a MIME-type, and this confuses some software.
2720 PH/70 Catch two compiler warnings in sieve.c.
2722 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2723 function verify_get_ident() calls ip_connect() to connect a socket, but
2724 if the "connect()" function timed out, ip_connect() used to close the
2725 socket. However, verify_get_ident() also closes the socket later, and in
2726 between Exim writes to the log, which may get opened at this point. When
2727 the socket was closed in ip_connect(), the log could get the same file
2728 descriptor number as the socket. This naturally causes chaos. The fix is
2729 not to close the socket in ip_connect(); the socket should be closed by
2730 the function that creates it. There was only one place in the code where
2731 this was missing, in the iplookup router, which I don't think anybody now
2732 uses, but I've fixed it anyway.
2734 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2735 well as to direct DNS lookups. Otherwise the handling of names in host
2736 lists is inconsistent and therefore confusing.
2742 PH/01 Two changes to the default runtime configuration:
2744 (1) Move the checks for relay_from_hosts and authenticated clients from
2745 after to before the (commented out) DNS black list checks.
2747 (2) Add control=submission to the relay_from_hosts and authenticated
2748 clients checks, on the grounds that messages accepted by these
2749 statements are most likely to be submissions.
2751 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2753 (1) Generate an error if the third argument for the ${prvs expansion is
2756 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2759 (3) Reset the variables that are obtained from the first argument of
2760 ${prvscheck and used in the second argument before leaving the code,
2761 because their memory is reclaimed, so using them afterwards may do
2764 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2765 one (it's much easier than Tom thought :-).
2767 (5) Because of (4), we can now allow for the use of $prvscheck_result
2768 inside the third argument.
2770 PH/03 For some reason, the default setting of PATH when running a command from
2771 a pipe transport was just "/usr/bin". I have changed it to
2774 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2775 anything to be listed in the output from -bV.
2777 PH/05 When a filter generated an autoreply, the entire To: header line was
2778 quoted in the delivery log line, like this:
2780 => >A.N.Other <ano@some.domain> <original@ddress> ...
2782 This has been changed so that it extracts the operative address. There
2783 may be more than one such address. If so, they are comma-separated, like
2786 => >ano@some.domain,ona@other.domain <original@ddress> ...
2788 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2789 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2790 not being looked up in the rDNS to get a host name, Exim was showing the
2791 IP address twice in Received: lines, even though the IP addresses were
2792 identical. For example:
2794 Received: from [1.2.3.4] (helo=[1.2.3.4])
2796 However, if the real host name was known, it was omitting the HELO data
2797 if it matched the actual IP address. This has been tidied up so that it
2798 doesn't show the same IP address twice.
2800 PH/07 When both +timestamp and +memory debugging was on, the value given by
2801 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2802 called by the debug printing, thereby overwriting the timestamp buffer.
2803 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2805 PH/08 When the original message was included in an autoreply transport, it
2806 always said "this is a copy of the message, including all the headers",
2807 even if body_only or headers_only was set. It now gives an appropriate
2810 PH/09 Applied a patch from the Sieve maintainer which:
2812 o fixes some comments
2813 o adds the (disabled) notify extension core
2814 o adds some debug output for the result of if/elsif tests
2815 o points to the current vacation draft in the documentation
2816 and documents the missing references header update
2820 o fixes a bug in processing the envelope test (when testing
2821 multiple envelope elements, the last element determinted the
2824 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2825 Electronic Mail") by including:
2827 Auto-submitted: auto-generated
2829 in the messages that it generates (bounce messages and others, such as
2830 warnings). In the case of bounce messages for non-SMTP mesages, there was
2831 also a typo: it was using "Auto_submitted" (underscore instead of
2832 hyphen). Since every message generated by Exim is necessarily in response
2833 to another message, thes have all been changed to:
2835 Auto-Submitted: auto-replied
2837 in accordance with these statements in the RFC:
2839 The auto-replied keyword:
2841 - SHOULD be used on messages sent in direct response to another
2842 message by an automatic process,
2844 - MUST NOT be used on manually-generated messages,
2846 - MAY be used on Delivery Status Notifications (DSNs) and Message
2847 Disposition Notifications (MDNs),
2849 - MUST NOT be used on messages generated by automatic or periodic
2850 processes, except for messages which are automatic responses to
2853 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2854 to the default Received: header definition.
2856 PH/12 Added log selector acl_warn_skipped (default on).
2858 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2859 variables because (a) they are in the wrong storage pool and (b) even if
2860 they were copied, it wouldn't work properly because of the caching.
2862 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2863 checking when decoding. Apparently there are clients that generate
2864 overlong encoded strings. Why am I not surprised?
2866 PH/15 If the first argument of "${if match_address" was not empty, but did not
2867 contain an "@" character, Exim crashed. Now it writes a panic log message
2868 and treats the condition as false.
2870 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2872 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2873 extension "envelope-auth". The code is finished and in agreement with
2874 other implementations, but there is no documentation so far and in fact,
2875 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2876 not changing the active code.
2878 Print executed "if" and "elsif" statements when debugging is used. This
2879 helps a great deal to understand what a filter does.
2881 Document more things not specified clearly in RFC3028. I had all this
2882 sorted out, when out of a sudden new issues came to my mind. Oops."
2884 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2887 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2888 canonical form (as documented). However, after a host name lookup from
2889 the IP address, check_host() was doing a simple string comparison with
2890 addresses acquired from the DNS when checking that the found name did
2891 have the original IP as one of its addresses. Since any found IPv6
2892 addresses are likely to be in abbreviated form, the comparison could
2893 fail. Luckily, there already exists a function for doing the comparison
2894 by converting both addresses to binary, so now that is used instead of
2895 the text comparison.
2897 PH/20 There was another similar case to PH/19, when a complete host name was
2898 given in a host list; looking up its IP address could give an abbreviated
2899 form, whereas the current host's name might or might not be abbreviated.
2900 The same fix has been applied.
2906 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2907 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2910 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2911 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2913 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2915 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2916 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2917 Also, for "Auto-Submitted", treat anything other than "no" as
2918 non-personal, in accordance with RFC 3834. (Previously it treated
2919 anything starting "auto-" as non-personal.)
2921 TF/01 The control=submission/name=... option had a problem with syntax
2922 errors if the name included a slash character. The /name= option
2923 now slurps the rest of the string, so it can include any characters
2924 but it must come last in the list of options (after /sender_retain
2927 PH/05 Some modifications to the interface to the fake nameserver for the new
2935 TK/01 Added the "success_on_redirect" address verification option. See
2936 NewStuff for rationale and an example.
2938 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2940 PH/02 Patch to exigrep to allow it to work on syslog lines.
2942 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2943 fread() to skip over the body file's header line, because in Cygwin the
2944 header line is locked and is inaccessible.
2946 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2947 co-exist for some time) to make it clear that it is the Exim ID that is
2948 referenced, not the Message-ID: header line.
2950 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2951 string_format() function, because snprintf() does not exist on all
2954 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2955 ${stat: expansion item.
2957 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2958 protocol synchronization error", to keep the pedants happy.
2960 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2961 well as for IRIX systems, when gcc is being used. See the host.c source
2964 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2966 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2969 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2970 transport and to the smtp transport in LMTP mode.
2972 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2974 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2975 run a filter in a subprocess. This could lead to confusion in subsequent
2976 lookups in the parent process. There should also be a search_tidyup() at
2977 the end of the subprocess.
2979 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2980 only if the host matched helo_try_verify_hosts, which caused the
2981 verification to occur when the EHLO/HELO command was issued. The ACL just
2982 tested the remembered result. Now, if a previous verification attempt has
2983 not happened, "verify = helo" does it there and then.
2985 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2987 TK/03 Fix log output including CR from clamd.
2989 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2990 exist provoked a memory error which could cause a segfault.
2992 PH/15 Installed PCRE 6.2
2994 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2996 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2997 of the problem. Specifically, suggested +O2 rather than +O1 for the
3000 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3002 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3003 record, the port was not correctly logged when the outgoing_port log
3004 selector was set (it logged the transort's default port).
3006 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3007 fallback_hosts, and "hosts" in the smtp transport.
3009 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3010 host errors such as "Connection refused".
3012 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3013 authentication with radiusclient 0.4.9:
3015 - Error returned from rc_read_config was caught wrongly
3016 - Username/password not passed on to radius server due to wrong length.
3018 The presumption is that some radiusclient API changes for 4.51/PH/17
3019 were not taken care of correctly. The code is still untested by me (my
3020 Linux distribution still has 0.3.2 of radiusclient), but it was
3021 contributed by a Radius user.
3023 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3024 expanding the "port" option of the smtp transport.
3026 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3027 while reading a MIME header. Thanks to Tom Hughes for a patch.
3029 PH/24 Include config.h inside local_scan.h so that configuration settings are
3032 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3033 that in an ACL for RCPT (for example), you can examine exactly what was
3036 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3037 commands, but it was not correctly comparing the address with the actual
3038 client host address. Thus, it would show the EHLO address in Received:
3039 header lines when this was not necessary.
3041 PH/27 Added the % operator to ${eval:}.
3043 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3044 it should be ignoring failures (because with -C, for example, it has lost
3045 privilege). It wasn't ignoring creation failures other than "already
3048 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3051 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3052 because an input line was too long, either on its own, or by virtue of
3053 too many continuations, the temporary file was not being removed, and the
3054 return code was incorrect.
3056 PH/31 Missing "BOOL" in function definition in filtertest.c.
3058 PH/32 Applied Sieve patches from the maintainer.
3060 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3062 PH/33 Added "verify = not_blind".
3064 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3065 Local/Makefile (with some defaults set). These are used in built scripts
3066 such as exicyclog, but they have never been used in the exim_install
3067 script (though there are many overriding facilities there). I have
3068 arranged that the exim_install script now takes note of these two
3071 PH/35 Installed configuration files for Dragonfly.
3073 PH/36 When a locally submitted message by a trusted user did not contain a
3074 From: header, and the sender address was obtained from -f or from an SMTP
3075 MAIL command, and the trusted user did not use -F to supply a sender
3076 name, $originator_name was incorrectly used when constructing a From:
3077 header. Furthermore, $originator_name was used for submission mode
3078 messages from external hosts without From: headers in a similar way,
3079 which is clearly wrong.
3081 PH/37 Added control=suppress_local_fixups.
3083 PH/38 When log_selector = +received_sender was set, and the addition of the
3084 sender made the log line's construction buffer exactly full, or one byte
3085 less than full, an overflow happened when the terminating "\n" was
3088 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3089 when the result of a list match is failure because a DNS lookup failed.
3091 PH/40 RM_COMMAND is now used in the building process.
3093 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3094 the "build-* directories that it finds.
3096 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3097 address in a domain literal was a prefix of an interface address.
3099 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3100 when verifying a sender address, unless rewrite_headers is false.
3102 PH/44 Wrote a long comment about why errors_to addresses are verified as
3103 recipients, not senders.
3105 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3106 the ratelimit ACL was added.
3108 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3110 PH/46 Added extra information about PostgreSQL errors to the error string.
3112 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3113 avoiding the need to install special zones in a real server. This is
3114 backwards compatible; if it can't find the fake resolver, it drops back.
3115 Thus, both old and new test suites can be run.
3117 TF/02 Added util/ratelimit.pl
3119 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3120 clock is set back in time.
3122 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3123 Candler <B.Candler@pobox.com>.
3125 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3126 OK for addresses that are the result of redirections.
3128 PH/48 A number of further additions for the benefit of the new test suite,
3129 including a fake gethostbyname() that interfaces to the fake DNS resolver
3132 TF/06 The fix for widen_domains has also been applied to qualify_single and
3133 search_parents which are the other dnslookup options that can cause
3136 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3139 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3141 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3142 service" if it encountered an MX record with an empty target hostname.
3143 The message is now "an MX or SRV record indicated no SMTP service".
3145 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3146 if the DNS of the sending site is misconfigured. This is quite a
3147 common situation. This change restores the behaviour of treating a
3148 helo verification defer as a failure.
3150 PH/51 If self=fail was set on a router, the bounce message did not include the
3151 actual error message.
3157 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3159 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3160 command itself ended in error, the underlying message about the transport
3161 filter timeout was being overwritten with the pipe command error. Now the
3162 underlying error message should be appended to the second error message.
3164 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3166 PH/02 Reduce the amount of output that "make" produces by default. Full output
3167 can still be requested.
3169 PH/03 The warning log line about a condition test deferring for a "warn" verb
3170 was being output only once per connection, rather than after each
3171 occurrence (because it was using the same function as for successful
3172 "warn" verbs). This seems wrong, so I have changed it.
3174 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3175 it should not have, which might have caused a crash in the right
3176 circumstances, but probably never did.
3178 PH/04 Installed a modified version of Tony Finch's patch to make submission
3179 mode fix the return path as well as the Sender: header line, and to
3180 add a /name= option so that you can make the user's friendly name appear
3183 TF/03 Added the control = fakedefer ACL modifier.
3185 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3186 Mark Lowes for thorough testing.
3188 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3190 TK/03 Merged latest SRS patch from Miles Wilton.
3192 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3193 with the definition in sysexits.h (which is #included earlier).
3194 Fortunately, Exim does not actually use EX_OK. The code used to try to
3195 preserve the sysexits.h value, by assumimg that macro definitions were
3196 scanned for macro replacements. I have been disabused of this notion,
3197 so now the code just undefines EX_OK before #including unistd.h.
3199 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3200 in the smtp transport. When a block could not be written in a single
3201 write() function, the timeout was being re-applied to each part-write.
3202 This seems wrong - if the receiver was accepting one byte at a time it
3203 would take for ever. The timeout is now adjusted when this happens. It
3204 doesn't have to be particularly precise.
3206 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3207 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3209 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3210 without a domain if the check to <postmaster@domain> fails.
3212 SC/01 Eximstats: added -xls and the ability to specify output files
3213 (patch written by Frank Heydlauf).
3215 SC/02 Eximstats: use FileHandles for outputing results.
3217 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3219 SC/04 Eximstats: fixed display of large numbers with -nvr option
3221 SC/05 Eximstats: fixed merging of reports with empty tables.
3223 SC/06 Eximstats: added the -include_original_destination flag
3225 SC/07 Eximstats: removed tabs and trailing whitespace.
3227 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3229 TK/06 MBOX spool code: Add real "From " MBOX separator line
3230 so the .eml file is really in mbox format (even though
3231 most programs do not really care). Patch from Alex Miller.
3233 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3234 The latter is generated from $received_to and is only set if the
3235 message has one envelope recipient. SA can use these headers,
3236 obviously out-of-the-box. Patch from Alex Miller.
3238 PH/08 The ${def test on a variable was returning false if the variable's
3239 value was "0", contrary to what the specification has always said!
3240 The result should be true unless the variable is empty.
3242 PH/09 The syntax error of a character other than { following "${if
3243 def:variable_name" (after optional whitespace) was not being diagnosed.
3244 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3245 accidental colon was present, for example, could give incorrect results.
3247 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3248 result is used (not including appendfile, where other changes are about
3251 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3252 This involved changing a lot of size variables from int to off_t. It
3253 should work with maildirs and everything.
3255 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3256 spamd dying while we are connected to it.
3258 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3261 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3262 function for BDB 4.3.
3264 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3266 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3267 the sources for PCRE have been re-arranged and more files are now
3270 PH/15 The code I had for printing potentially long long variables in PH/11
3271 above was not the best (it lost precision). The length of off_t variables
3272 is now inspected at build time, and an appropriate printing format (%ld
3273 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3274 to be "long long int" or "long int". This is needed for the internal
3275 formatting function string_vformat().
3277 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3278 the configuration file to be ":syslog", then the script "guesses" where
3279 the logs files are, rather than using the compiled in default. In our
3280 case the guess is not the same as the compiled default, so the script
3281 suddenly stopped working when I started to use syslog. The patch checks
3282 to see if log_file_path is "". If so, it attempts to read it from exim
3283 with no configuration file to get the compiled in version, before it
3284 falls back to the previous guessing code."
3286 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3287 implementing BATV in an Exim configuration. See NewStuff for the gory
3290 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3291 Makefile that are specific to HP-UX.
3293 PH/18 If the "use_postmaster" option was set for a recipient callout together
3294 with the "random" option, the postmaster address was used as the MAIL
3295 FROM address for the random test, but not for the subsequent recipient
3296 test. It is now used for both.
3298 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3299 patch removes a few documentation additions to RFC 3028, because the
3300 latest draft now contains them. It adds the new en;ascii-case comparator
3301 and a new error check for 8bit text in MIME parts. Comparator and
3302 require names are now matched exactly. I enabled the subaddress
3303 extension, but it is not well tested yet (read: it works for me)."
3305 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3306 rework some of the code of TK/09 above to avoid the hardwired use of
3307 "%lld" and "long long". Replaced the call to snprintf() with a call to
3310 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3311 records point to non-existent hosts", "retry timeout exceeded", and
3312 "retry time not reached for any host after a long failure period".
3314 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3315 experimental DomainKeys support:
3317 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3318 (2) On an error such as an illegally used "control", the wrong name for
3319 the control was given.
3321 These problems did NOT occur unless DomainKeys support was compiled.
3323 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3325 PH/24 Added ${if match_ip condition.
3327 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3328 databases so that it will be absolutely obvious if a crash occurs in the
3329 DB library. This is a regular occurrence (often caused by mis-matched
3332 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3333 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3334 picked up on a user's system that detects such things. There doesn't seem
3335 to be a gcc warning option for this - only an attribute that has to be
3336 put on the function's prototype. It seems that in Fedora Core 4 they have
3337 set this on a number of new functions. No doubt there will be more in due
3340 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3341 specify a transport. However, if an address that was verified by such a
3342 router was the subject of a callout, Exim crashed because it tried to
3343 read the rcpt_include_affixes from the non-existent transport. Now it
3344 just assumes that the setting of that option is false. This bug was
3345 introduced by 4.51/PH/31.
3347 PH/28 Changed -d+all to exclude +memory, because that information is very
3348 rarely of interest, but it makes the output a lot bigger. People tend to
3349 do -d+all out of habit.
3351 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3352 code in os-type was giving problems when libc.so lives in lib64, like on
3355 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3356 aren't the modern standard, and it seems that some systems' include files
3357 don't always have them. Exim was already checking for some of the newer
3358 ones like T_AAAA, and defining it itself. I've added checks for all the
3359 record types that Exim uses.
3361 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3362 not automatically generating a new one, as it is supposed to. This
3363 prevented TLS from working. If the file did exist, but contained invalid
3364 data, a new version was generated, as expected. It was only the case of a
3365 non-existent file that was broken.
3367 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3368 with a change in libdomainkeys > 0.64.
3370 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3371 from DNS. If the selector record carries the flag, it now has
3372 precedence over the domain-wide flag.
3374 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3376 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3377 the use of a port name that isn't defined in /etc/services) occurred, the
3378 message was deferred as in a normal delivery, and thus remained on the
3379 spool, instead of being failed because of the mua_wrapper setting. This
3380 is now fixed, and I tidied up some of the mua_wrapper messages at the
3383 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3384 the messages in a hash of arrays rather than using individual hashes.
3385 This is a bit cleaner and results in dramatic memory savings, albeit
3386 at a slight CPU cost.
3388 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3389 as requested by Marc Sherman.
3391 SC/10 Eximstats: added histograms for user specified patterns as requested
3394 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3396 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3397 fopen() in the content-scanning modules that did not already have it.
3403 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3404 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3406 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3408 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3410 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3411 to test Sieve filters that use "vacation".
3413 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3414 that changes the way the GnuTLS parameters are stored in the cache file.
3415 The new format can be generated externally. For backward compatibility,
3416 if the data in the cache doesn't make sense, Exim assumes it has read an
3417 old-format file, and it generates new data and writes a new file. This
3418 means that you can't go back to an older release without removing the
3421 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3422 work if there are any delivery delays because "one_time" forces the
3423 parent to be marked "delivered", so its unseen clone is never tried
3424 again. For this reason, Exim now forbids the simultaneous setting of
3427 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3428 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3429 entry if you want to know the details. The fix, however, neglected to
3430 consider the case when local delivery batching is involved. The test for
3431 "previously delivered" was not happening when checking to see if an
3432 address could be batched with a previous (undelivered) one; under
3433 certain circumstances this could lead to multiple deliveries to the same
3436 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3437 in its include files, and this causes problems building Exim.
3439 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3440 header_syntax) but Exim was just ignoring anything given after a slash.
3441 In particular, this caused confusion with an attempt to use "verify =
3442 reverse_host_lookup/defer_ok". An error is now given when options are
3443 supplied for verify items that do not have them. (Maybe reverse_host_
3444 lookup should have a defer_ok option, but that's a different point.)
3446 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3447 defined by RFC 821) to 2048, because there were problems with some AUTH
3448 commands, and RFC 1869 says the size should be increased for extended
3449 SMTP commands that take arguments.
3451 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3454 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3455 "unknown" error; now it says that the functionality isn't in the binary.
3457 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3458 an address' error message when a string expansion fails (syntax or
3459 whatever). Otherwise the password may appear in the log. Following change
3460 PH/42 below, there is no longer a chance of it appearing in a bounce
3463 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3465 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3466 of its addresses was ever tried. (Bugzilla bug #2.)
3468 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3469 the result incorrectly in the debug output. (It correctly added a newline
3470 to what was transported.)
3472 TF/01 Added $received_time.
3474 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3475 commented out examples of how to interface to a virus scanner and to
3476 SpamAssassin. Also added commented examples of av_scanner and
3477 spamd_address settings.
3479 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3480 and controls are allowed in which ACLs. There were a couple of minor
3481 errors. Some of the entries in the conditions table (which is a table of
3482 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3483 negation of where the condition IS allowed.
3485 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3487 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3488 header file does not have a version number, so I've had to invent a new
3489 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3490 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3491 radiusclient), but it was contributed by a Radius user.
3493 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3494 files or directories, for OpenSSL.
3496 PH/19 When an Exim process that is running as root has to create an Exim log
3497 file, it does so in a subprocess that runs as exim:exim so as to get the
3498 ownership right at creation (otherwise, other Exim processes might see
3499 the file with the wrong ownership). There was no test for failure of this
3500 fork() call, which would lead to the process getting stuck as it waited
3501 for a non-existent subprocess. Forks do occasionally fail when resources
3502 run out. I reviewed all the other calls to fork(); they all seem to check
3505 PH/20 When checking for unexpected SMTP input at connect time (before writing
3506 the banner), Exim was not dealing correctly with a non-positive return
3507 from the read() function. If the client had disconnected by this time,
3508 the result was a log entry for a synchronization error with an empty
3509 string after "input=" when read() returned zero. If read() returned -1
3510 (an event I could not check), uninitialized data bytes were printed.
3511 There were reports of junk text (parts of files, etc) appearing after
3514 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3516 PH/22 Added support for macro redefinition, and (re)definition in between
3517 driver and ACL definitions.
3519 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3520 forgetting to use the resulting value; it was using the unexpanded value.
3522 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3523 hadn't been configured. The fix is from Juergen Kreileder, who
3524 understands it better than I do:
3526 "Here's what I see happening with three configured cyrus_sasl
3527 authenticators configured (plain, login, cram-md5):
3529 On startup auth_cyrus_sasl_init() gets called for each of these.
3530 This means three calls to sasl_listmech() without a specified mech_list.
3531 => SASL tests which mechs of all available mechs actually work
3532 => three warnings about OTP not working
3533 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3535 With the patch, sasl_listmech() also gets called three times. But now
3536 SASL's mech_list option is set to the server_mech specified in the the
3537 authenticator. Or in other words, the answer from sasl_listmech()
3538 gets limited to just the mech you're testing for (which is different
3540 => the return list contains just 'plain' or 'login', 'cram-md5' or
3541 nothing depending on the value of ob->server_mech.
3543 I've just tested the patch: Authentication still works fine,
3544 unavailable mechs specified in the exim configuration are still
3545 caught, and the auth.log warnings about OTP are gone."
3547 PH/25 When debugging is enabled, the contents of the command line are added
3548 to the debugging output, even when log_selector=+arguments is not
3551 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3552 answer is "GNU", and only if the return is "GNU/something" is the answer
3555 PH/27 $acl_verify_message is now set immediately after the failure of a
3556 verification in an ACL, and so is available in subsequent modifiers. In
3557 particular, the message can be preserved by coding like this:
3559 warn !verify = sender
3560 set acl_m0 = $acl_verify_message
3562 Previously, $acl_verify_message was set only while expanding "message"
3563 and "log_message" when a very denied access.
3565 PH/28 Modified OS/os.c-Linux with
3567 -#ifndef OS_LOAD_AVERAGE
3568 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3570 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3571 nomenclature these days.)
3573 PH/29 Installed patch from the Sieve maintainer that adds the options
3574 sieve_useraddress and sieve_subaddress to the redirect router.
3576 PH/30 In these circumstances:
3577 . Two addresses routed to the same list of hosts;
3578 . First host does not offer TLS;
3579 . First host accepts first address;
3580 . First host gives temporary error to second address;
3581 . Second host offers TLS and a TLS session is established;
3582 . Second host accepts second address.
3583 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3584 and peerdn, if requested) that were in fact used only for the second
3587 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3588 attention to any local part prefix or suffix that was matched by the
3589 router that accepted the address. It now behaves in the same way as it
3590 does for delivery: the affixes are removed from the local part unless
3591 rcpt_include_affixes is set on the transport.
3593 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3594 timeout during the DATA phase of an incoming message.
3596 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3597 applied a patch sanctioned by the Sieve maintainer.
3599 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3600 the uid or gid is negative. A case of a negative gid caused this to be
3601 noticed. The fix allows for either to be negative.
3603 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3604 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3607 PH/36 The change PH/12 above was broken. Fixed it.
3609 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3610 the grounds that routing the same address twice would always produce the
3611 same answer. This might have been true once, but it is certainly no
3612 longer true now. Routing a child address may depend on the previous
3613 routing that produced that child. Some complicated redirection strategies
3614 went wrong when messages had multiple recipients, and made Exim's
3615 behaviour dependent on the order in which the addresses were given.
3617 I have moved the duplicate checking until after the routing is complete.
3618 Exim scans the addresses that are assigned to local and remote
3619 transports, and removes any duplicates. This means that more work will be
3620 done, as duplicates will always all be routed, but duplicates are
3621 presumably rare, so I don't expect this is of any significance.
3623 For deliveries to pipes, files, and autoreplies, the duplicate checking
3624 still happens during the routing process, since they are not going to be
3627 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3628 It corrects a timeout issue with spamd. This is Ian's comment: "The
3629 background is that sometimes spamd either never reads data from a
3630 connection it has accepted, or it never writes response data. The exiscan
3631 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3632 blindly assumes that writes won't block so it may never time out."
3634 PH/39 Allow G after quota size as well as K and M.
3636 PH/40 The value set for $authenticated_id in an authenticator may not contain
3637 binary zeroes or newlines because the value is written to log lines and
3638 to spool files. There was no check on this. Now the value is run through
3639 the string_printing() function so that such characters are converted to
3640 printable escape sequences.
3642 PH/41 $message_linecount is a new variable that contains the total number of
3643 lines in the message. Compare $body_linecount, which is the count for the
3646 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3647 bounce and delay warning messages, except in certain special cases, which
3650 (a) An SMTP error message from a remote host;
3651 (b) A message specified in a :fail: redirection;
3652 (c) A message specified in a "fail" command in a system filter;
3653 (d) A message specified in a FAIL return from the queryprogram router;
3654 (e) A message specified by the cannot_route_message router option.
3656 In these cases only, Exim does include the error details in bounce and
3657 warning messages. There are also a few cases where bland messages such
3658 as "unrouteable address" or "local delivery error" are given.
3660 PH/43 $value is now also set for the "else" part of a ${run expansion.
3662 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3663 being worked on, but at least Exim now implements the latest version to
3666 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3667 process to complete was treated as a delivery failure, a timeout while
3668 writing the message to the pipe was logged, but erroneously treated as a
3669 successful delivery. Such timeouts include transport filter timeouts. For
3670 consistency with the overall process timeout, these timeouts are now
3671 treated as errors, giving rise to delivery failures by default. However,
3672 there is now a new Boolean option for the pipe transport called
3673 timeout_defer, which, if set TRUE, converts the failures into defers for
3674 both kinds of timeout. A transport filter timeout is now identified in
3677 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3678 systems where "make" and "gmake" are different, calling "gmake" at top
3679 level broke things. I've arranged for the value of $(MAKE) to be passed
3680 from the Makefile to this script so that it can call the same version of
3684 A note about Exim versions 4.44 and 4.50
3685 ----------------------------------------
3687 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3688 changes of various kinds. As a consequence, a big documentation update was
3689 needed. This delayed the release for rather longer than seemed good, especially
3690 in the light of a couple of (minor) security issues. Therefore, the changes
3691 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3692 release. So 4.44 and 4.50 are in effect two different branches that both start
3695 I have left the 4.50 change log unchanged; it contains all the changes since
3696 4.43. The change log for 4.44 is below; many of its items are identical to
3697 those for 4.50. This seems to be the most sensible way to preserve the
3698 historical information.
3704 1. Minor wording change to the doc/README.SIEVE file.
3706 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3707 computation of the current number of files was incorrect.
3709 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3710 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3713 4. Give more explanation in the error message when the command for a transport
3714 filter fails to execute.
3716 5. There are several places where Exim runs a non-Exim command in a
3717 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3718 was being done only for the command run by the queryprogram router. It is
3719 now done for all such subprocesses. The other cases are: ${run, transport
3720 filters, and the commands run by the lmtp and pipe transports.
3722 6. Added CONFIGURE_GROUP build-time option.
3724 7. Some older OS have a limit of 256 on the maximum number of file
3725 descriptors. Exim was using setrlimit() to set 1000 as a large value
3726 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3727 systems. I've change it so that if it can't get 1000, it tries for 256.
3729 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3730 was an oversight, and furthermore, ever since the addition of extra
3731 controls (e.g. 4.43/32), the checks on when to allow different forms of
3732 "control" were broken. There should now be diagnostics for all cases when a
3733 control that does not make sense is encountered.
3735 9. Added the /retain_sender option to "control=submission".
3737 10. $recipients is now available in the predata ACL (oversight).
3739 11. Tidy the search cache before the fork to do a delivery from a message
3740 received from the command line. Otherwise the child will trigger a lookup
3741 failure and thereby defer the delivery if it tries to use (for example) a
3742 cached ldap connection that the parent has called unbind on.
3744 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3745 of $address_data from the recipient verification was clobbered by the
3746 sender verification.
3748 13. The value of address_data from a sender verification is now available in
3749 $sender_address_data in subsequent conditions in the ACL statement.
3751 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3753 15. Added a new option "connect=<time>" to callout options, to set a different
3756 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3757 was its contents. (It was OK if the option was not defined at all.)
3759 17. A "Completed" log line is now written for messages that are removed from
3760 the spool by the -Mrm option.
3762 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3763 information about exactly what failed.
3765 19. Added -dd to debug only the daemon process.
3767 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3768 handles timeouts, both on the server side and network timeouts. Renamed the
3769 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3771 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3772 It is now set to "smtps".
3774 22. $host_address is now set to the target address during the checking of
3775 ignore_target_hosts.
3777 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3778 being passed; this would have caused $sender_host_name to have been used if
3779 matching the list had actually called for a host name (not very likely,
3780 since this list is usually IP addresses). A host name is now passed as
3783 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3784 code that specifies a non-restarting handler (typically sigaction() in
3785 modern systems) in an attempt to fix a rare and obscure crash bug.
3787 25. Narrowed the window for a race in the daemon that could cause it to ignore
3788 SIGCHLD signals. This is not a major problem, because they are used only to
3789 wake it up if nothing else does.
3791 26. A malformed maildirsize file could cause Exim to calculate negative values
3792 for the mailbox size or file count. Odd effects could occur as a result.
3793 The maildirsize information is now recalculated if the size or filecount
3796 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3797 support for a long time. Removed HAVE_SYS_VFS_H.
3799 28. Installed the latest version of exipick from John Jetmore.
3801 29. In an address list, if the pattern was not a regular expression, an empty
3802 subject address (from a bounce message) matched only if the pattern was an
3803 empty string. Non-empty patterns were not even tested. This was the wrong
3804 because it is perfectly reasonable to use an empty address as part of a
3805 database query. An empty address is now tested by patterns that are
3806 lookups. However, all the other forms of pattern expect the subject to
3807 contain a local part and a domain, and therefore, for them, an empty
3808 address still always fails if the pattern is not itself empty.
3810 30. Exim went into a mad DNS loop when attempting to do a callout where the
3811 host was specified on an smtp transport, and looking it up yielded more
3812 than one IP address.
3814 31. Re-factored the code for checking spool and log partition space into a
3815 function that finds that data and another that does the check. The former
3816 is then used to implement four new variables: $spool_space, $log_space,
3817 $spool_inodes, and $log_inodes.
3819 32. The RFC2047 encoding function was originally intended for short strings
3820 such as real names; it was not keeping to the 75-character limit for
3821 encoded words that the RFC imposes. It now respects the limit, and
3822 generates multiple encoded words if necessary. To be on the safe side, I
3823 have increased the buffer size for the ${rfc2047: expansion operator from
3826 33. It is now permitted to omit both strings after an "if" condition; if the
3827 condition is true, the result is "true". As before, when the second string
3828 is omitted, a false condition yields an empty string. This makes it less
3829 cumbersome to write custom ACL and router conditions.
3831 34. Failure to deliver a bounce message always caused it to be frozen, even if
3832 there was an errors_to setting on the router. The errors_to setting is now
3835 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3836 canonical form (fully expanded) before being placed in
3837 $sender_host_address.
3839 36. The table in the code that translates DNS record types into text (T_A to
3840 "A" for instance) was missing entries for NS and CNAME. It is just possible
3841 that this could have caused confusion if both these types were looked up
3842 for the same domain, because the text type is used as part of Exim's
3843 per-process caching. But the chance of anyone hitting this buglet seems
3846 37. The dnsdb lookup has been extended in a number of ways.
3848 (1) There is a new type, "zns", which walks up the domain tree until it
3849 finds some nameserver records. It should be used with care.
3851 (2) There is a new type, "mxh", which is like "mx" except that it returns
3852 just the host names, not the priorities.
3854 (3) It is now possible to give a list of domains (or IP addresses) to be
3855 looked up. The behaviour when one of the lookups defers can be
3856 controlled by a keyword.
3858 (4) It is now possible to specify the separator character for use when
3859 multiple records are returned.
3861 38. The dnslists ACL condition has been extended: it is now possible to supply
3862 a list of IP addresses and/or domains to be looked up in a particular DNS
3865 39. Added log_selector=+queue_time_overall.
3867 40. When running the queue in the test harness, wait just a tad after forking a
3868 delivery process, to get repeatability of debugging output.
3870 41. Include certificate and key file names in error message when GnuTLS fails
3871 to set them up, because the GnuTLS error message doesn't include the name
3872 of the failing file when there is a problem reading it.
3874 42. Allow both -bf and -bF in the same test run.
3876 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3878 44. The "Exiscan patch" is now merged into the mainline Exim source.
3880 45. Sometimes the final signoff response after QUIT could fail to get
3881 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3882 < 0 before doing a fflush(). This bug looks as though it goes back to the
3883 introduction of TLS in release 3.20, but "sometimes" must have been rare
3884 because the tests only now provoked it.
3886 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3887 (this can affect the format of dates).
3889 47. exim_tidydb, when checking for the continued existence of a message for
3890 which it has found a message-specific retry record, was not finding
3891 messages that were in split spool directories. Consequently, it was
3892 deleting retry records that should have stayed in existence.
3894 48. Steve fixed some bugs in eximstats.
3896 49. The SPA authentication driver was not abandoning authentication and moving
3897 on to the next authenticator when an expansion was forced to fail,
3898 contradicting the general specification for all authenticators. Instead it
3899 was generating a temporary error. It now behaves as specified.
3901 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3902 (the order specifies the preference for clients). The order is now AES256,
3903 AES128, 3DES, ARCFOUR128.
3905 51. Small patch to Sieve code - explicitly set From: when generating an
3908 52. Exim crashed if a remote delivery caused a very long error message to be
3909 recorded - for instance if somebody sent an entire SpamAssassin report back
3910 as a large number of 550 error lines. This bug was coincidentally fixed by
3911 increasing the size of one of Exim's internal buffers (big_buffer) that
3912 happened as part of the Exiscan merge. However, to be on the safe side, I
3913 have made the code more robust (and fixed the comments that describe what
3916 53. Now that there can be additional text after "Completed" in log lines (if
3917 the queue_time_overall log selector is set), a one-byte patch to exigrep
3918 was needed to allow it to recognize "Completed" as not the last thing in
3921 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3922 patch that reportedly fixes this has been added. I am not expert enough to
3923 create a test for it. This is what the patch creator wrote:
3925 "I found a little strange behaviour of ldap code when working with
3926 Windows 2003 AD Domain, where users was placed in more than one
3927 Organization Units. When I tried to give exim partial DN, the exit code
3928 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3929 But simultaneously result of request was absolutely normal ldap result,
3930 so I produce this patch..."
3932 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3933 REFERENCE, so I have modified the code to exclude the patch when that macro
3936 55. Some experimental protocols are using DNS PTR records for new purposes. The
3937 keys for these records are domain names, not reversed IP addresses. The
3938 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3939 leaves it alone. Component reversal etc. now happens only for IP addresses.
3942 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3944 57. Double the size of the debug message buffer (to 2048) so that more of very
3945 long debug lines gets shown.
3947 58. The exicyclog utility now does better if the number of log files to keep
3948 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3950 59. Two changes related to the smtp_active_hostname option:
3952 (1) $smtp_active_hostname is now available as a variable.
3953 (2) The default for smtp_banner uses $smtp_active_hostname instead
3954 of $primary_hostname.
3956 60. The host_aton() function is supposed to be passed a string that is known
3957 to be a valid IP address. However, in the case of IPv6 addresses, it was
3958 not checking this. This is a hostage to fortune. Exim now panics and dies
3959 if the condition is not met. A case was found where this could be provoked
3960 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3961 components; fortuitously, this particular loophole had already been fixed
3962 by change 4.50/55 above.
3964 If there are any other similar loopholes, the new check in host_aton()
3965 itself should stop them being exploited. The report I received stated that
3966 data on the command line could provoke the exploit when Exim was running as
3967 exim, but did not say which command line option was involved. All I could
3968 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3969 running as the user.
3972 61. There was a buffer overflow vulnerability in the SPA authentication code
3973 (which came originally from the Samba project). I have added a test to the
3974 spa_base64_to_bits() function which I hope fixes it.
3977 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3978 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3980 63. The daemon start-up calls getloadavg() while still root for those OS that
3981 need the first call to be done as root, but it missed one case: when
3982 deliver_queue_load_max is set with deliver_drop_privilege. This is
3983 necessary for the benefit of the queue runner, because there is no re-exec
3984 when deliver_drop_privilege is set.
3986 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3987 This has been fixed.
3989 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3990 was in use, was not putting the data itself into the right store pool;
3991 consequently, it could be overwritten for a subsequent message in the same
3992 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3995 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3997 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3998 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3999 the function were treating the return as a boolean value, which happened to
4000 work because 0=false and not-0=true, but is not correct code.
4002 68. The host_aton() function was not handling scoped IPv6 addresses (those
4003 with, for example, "%eth0" on the end) correctly.
4005 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4006 negated items (that is, ~something) in unsigned ints. Some compilers
4007 apparently mutter when there is no cast.
4009 70. If an address verification called from an ACL failed, and did not produce a
4010 user-specific message (i.e. there was only a "system" message), nothing was
4011 put in $acl_verify_message. In this situation, it now puts the system
4014 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4015 session; change 4.31/43 added the unwanted input to the log line - except
4016 that it did not do this in the start of session case. It now does.
4018 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4019 This is wrong and can cause the other end to generate a synchronization
4020 error if it is another Exim or anything else that does the synchronization
4021 check. A QUIT command is no longer sent after a timeout.
4023 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4024 during host lookups.
4026 74. The defer_ok option of callout verification was not working if it was used
4027 when verifying addresses in header lines, that is, for this case:
4029 verify = header_sender/callout=defer_ok
4031 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4032 those file descriptors could be used for SMTP connections. If anything
4033 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4034 could be sent to the SMTP client, causing chaos. The daemon now opens
4035 stdin, stdout, and stderr to /dev/null when it puts itself into the
4038 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4039 log by default. The user can override this with suitable Perl magic.
4041 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4042 the log message when discard triggers, was not working for the DATA ACL or
4043 for the non-SMTP ACL.
4045 78. Error message wording change in sieve.c.
4047 79. If smtp_accept_max_per_host was set, the number of connections could be
4048 restricted to fewer than expected, because the daemon was trying to set up
4049 a new connection before checking whether the processes handling previous
4050 connections had finished. The check for completed processes is now done
4051 earlier. On busy systems, this bug wouldn't be noticed because something
4052 else would have woken the daemon, and it would have reaped the completed
4055 80. If a message was submitted locally by a user whose login name contained one
4056 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4057 It caused a spool format error. I have fixed the spool reading code. A
4058 related problem was that the "from" clause in the Received: line became
4059 illegal because of the space(s). It is now covered by ${quote_local_part.
4061 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4064 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4066 83. Patch from Sieve maintainer to fix unterminated string problem in
4067 "vacation" handling.
4069 84. Some minor changes to the Linux configuration files to help with other
4070 OS variants using glibc.
4072 85. One more patch for Sieve to update vacation handling to latest spec.
4075 ----------------------------------------------------
4076 See the note above about the 4.44 and 4.50 releases.
4077 ----------------------------------------------------
4083 1. Change 4.43/35 introduced a bug that caused file counts to be
4084 incorrectly computed when quota_filecount was set in an appendfile
4087 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4088 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4091 3. Give more explanation in the error message when the command for a transport
4092 filter fails to execute.
4094 4. There are several places where Exim runs a non-Exim command in a
4095 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4096 was being done only for the command run by the queryprogram router. It is
4097 now done for all such subprocesses. The other cases are: ${run, transport
4098 filters, and the commands run by the lmtp and pipe transports.
4100 5. Some older OS have a limit of 256 on the maximum number of file
4101 descriptors. Exim was using setrlimit() to set 1000 as a large value
4102 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4103 systems. I've change it so that if it can't get 1000, it tries for 256.
4105 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4106 was an oversight, and furthermore, ever since the addition of extra
4107 controls (e.g. 4.43/32), the checks on when to allow different forms of
4108 "control" were broken. There should now be diagnostics for all cases when a
4109 control that does not make sense is encountered.
4111 7. $recipients is now available in the predata ACL (oversight).
4113 8. Tidy the search cache before the fork to do a delivery from a message
4114 received from the command line. Otherwise the child will trigger a lookup
4115 failure and thereby defer the delivery if it tries to use (for example) a
4116 cached ldap connection that the parent has called unbind on.
4118 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4119 of $address_data from the recipient verification was clobbered by the
4120 sender verification.
4122 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4123 was its contents. (It was OK if the option was not defined at all.)
4125 11. A "Completed" log line is now written for messages that are removed from
4126 the spool by the -Mrm option.
4128 12. $host_address is now set to the target address during the checking of
4129 ignore_target_hosts.
4131 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4132 being passed; this would have caused $sender_host_name to have been used if
4133 matching the list had actually called for a host name (not very likely,
4134 since this list is usually IP addresses). A host name is now passed as
4137 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4138 code that specifies a non-restarting handler (typically sigaction() in
4139 modern systems) in an attempt to fix a rare and obscure crash bug.
4141 15. Narrowed the window for a race in the daemon that could cause it to ignore
4142 SIGCHLD signals. This is not a major problem, because they are used only to
4143 wake it up if nothing else does.
4145 16. A malformed maildirsize file could cause Exim to calculate negative values
4146 for the mailbox size or file count. Odd effects could occur as a result.
4147 The maildirsize information is now recalculated if the size or filecount
4150 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4151 support for a long time. Removed HAVE_SYS_VFS_H.
4153 18. Updated exipick to current release from John Jetmore.
4155 19. Allow an empty sender to be matched against a lookup in an address list.
4156 Previously the only cases considered were a regular expression, or an
4159 20. Exim went into a mad DNS lookup loop when doing a callout where the
4160 host was specified on the transport, if the DNS lookup yielded more than
4163 21. The RFC2047 encoding function was originally intended for short strings
4164 such as real names; it was not keeping to the 75-character limit for
4165 encoded words that the RFC imposes. It now respects the limit, and
4166 generates multiple encoded words if necessary. To be on the safe side, I
4167 have increased the buffer size for the ${rfc2047: expansion operator from
4170 22. Failure to deliver a bounce message always caused it to be frozen, even if
4171 there was an errors_to setting on the router. The errors_to setting is now
4174 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4175 canonical form (fully expanded) before being placed in
4176 $sender_host_address.
4178 24. Updated eximstats to version 1.33
4180 25. Include certificate and key file names in error message when GnuTLS fails
4181 to set them up, because the GnuTLS error message doesn't include the name
4182 of the failing file when there is a problem reading it.
4184 26. Expand error message when OpenSSL has problems setting up cert/key files.
4187 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4188 (this can affect the format of dates).
4190 28. exim_tidydb, when checking for the continued existence of a message for
4191 which it has found a message-specific retry record, was not finding
4192 messages that were in split spool directories. Consequently, it was
4193 deleting retry records that should have stayed in existence.
4195 29. eximstats updated to version 1.35
4196 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4197 1.35 - bugfix such that pie charts by volume are generated correctly
4199 30. The SPA authentication driver was not abandoning authentication and moving
4200 on to the next authenticator when an expansion was forced to fail,
4201 contradicting the general specification for all authenticators. Instead it
4202 was generating a temporary error. It now behaves as specified.
4204 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4205 (the order specifies the preference for clients). The order is now AES256,
4206 AES128, 3DES, ARCFOUR128.
4208 31. Small patch to Sieve code - explicitly set From: when generating an
4211 32. Exim crashed if a remote delivery caused a very long error message to be
4212 recorded - for instance if somebody sent an entire SpamAssassin report back
4213 as a large number of 550 error lines. This bug was coincidentally fixed by
4214 increasing the size of one of Exim's internal buffers (big_buffer) that
4215 happened as part of the Exiscan merge. However, to be on the safe side, I
4216 have made the code more robust (and fixed the comments that describe what
4219 33. Some experimental protocols are using DNS PTR records for new purposes. The
4220 keys for these records are domain names, not reversed IP addresses. The
4221 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4222 leaves it alone. Component reversal etc. now happens only for IP addresses.
4225 34. The host_aton() function is supposed to be passed a string that is known
4226 to be a valid IP address. However, in the case of IPv6 addresses, it was
4227 not checking this. This is a hostage to fortune. Exim now panics and dies
4228 if the condition is not met. A case was found where this could be provoked
4229 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4230 components; fortuitously, this particular loophole had already been fixed
4231 by change 4.50/55 or 4.44/33 above.
4233 If there are any other similar loopholes, the new check in host_aton()
4234 itself should stop them being exploited. The report I received stated that
4235 data on the command line could provoke the exploit when Exim was running as
4236 exim, but did not say which command line option was involved. All I could
4237 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4238 running as the user.
4241 35. There was a buffer overflow vulnerability in the SPA authentication code
4242 (which came originally from the Samba project). I have added a test to the
4243 spa_base64_to_bits() function which I hope fixes it.
4246 36. The daemon start-up calls getloadavg() while still root for those OS that
4247 need the first call to be done as root, but it missed one case: when
4248 deliver_queue_load_max is set with deliver_drop_privilege. This is
4249 necessary for the benefit of the queue runner, because there is no re-exec
4250 when deliver_drop_privilege is set.
4252 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4253 was in use, was not putting the data itself into the right store pool;
4254 consequently, it could be overwritten for a subsequent message in the same
4255 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4258 38. Sometimes the final signoff response after QUIT could fail to get
4259 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4260 < 0 before doing a fflush(). This bug looks as though it goes back to the
4261 introduction of TLS in release 3.20, but "sometimes" must have been rare
4262 because the tests only now provoked it.
4268 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4269 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4270 Now it returns an integer. A number of calls were still expecting a T/F
4271 return. Fortuitously, in all cases, the tests worked in OK situations,
4272 which is the norm. However, things would have gone wrong on any write
4273 failures on the smtp file descriptor. This function is used when sending
4274 messages over SMTP and also when doing verify callouts.
4276 2. When Exim is called to do synchronous delivery of a locally submitted
4277 message (the -odf or -odi options), it no longer closes stderr before doing
4280 3. Implemented the mua_wrapper option.
4282 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4284 5. Implemented the functions header_remove(), header_testname(),
4285 header_add_at_position(), and receive_remove_recipient(), and exported them
4288 6. If an ACL "warn" statement specified the addition of headers, Exim already
4289 inserted X-ACL-Warn: at the start if there was no header name. However, it
4290 was not making this test for the second and subsequent header lines if
4291 there were newlines in the string. This meant that an invalid header could
4292 be inserted if Exim was badly configured.
4294 7. Allow an ACL "warn" statement to add header lines at the start or after all
4295 the Received: headers, as well as at the end.
4297 8. Added the rcpt_4xx retry error code.
4299 9. Added postmaster_mailfrom=xxx to callout verification option.
4301 10. Added mailfrom=xxxx to the callout verification option, for verify=
4304 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4305 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4307 12. Inserted some casts to stop certain compilers warning when using pointer
4308 differences as field lengths or precisions in printf-type calls (mostly
4309 affecting debugging statements).
4311 13. Added optional readline() support for -be (dynamically loaded).
4313 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4314 same clock tick as a message's arrival, so that its received time was the
4315 same as the "first fail" time on the retry record, and that message
4316 remained on the queue past the ultimate address timeout, every queue runner
4317 would try a delivery (because it was past the ultimate address timeout) but
4318 after another failure, the ultimate address timeout, which should have then
4319 bounced the address, did not kick in. This was a "< instead of <=" error;
4320 in most cases the first failure would have been in the next clock tick
4321 after the received time, and all would be well.
4323 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4324 being recognized when the domain list was tested by the match_domain
4325 condition in an expansion string.
4327 16. Added the ${str2b64: operator.
4329 17. Exim was always calling setrlimit() to set a large limit for the number of
4330 processes, without checking whether the existing limit was already
4331 adequate. (It did check for the limit on file descriptors.) Furthermore,
4332 errors from getrlimit() and setrlimit() were being ignored. Now they are
4333 logged to the main and panic logs, but Exim does carry on, to try to do its
4334 job under whatever limits there are.
4336 18. Imported PCRE 5.0.
4338 19. Trivial typo in log message " temporarily refused connection" (the leading
4341 20. If the log selector return_path_on_delivery was set and an address was
4342 redirected to /dev/null, the delivery process crashed because it assumed
4343 that a return path would always be set for a "successful" delivery. In this
4344 case, the whole delivery is bypassed as an optimization, and therefore no
4347 21. Internal re-arrangement: the function for sending a challenge and reading
4348 a response while authentication was assuming a zero-terminated challenge
4349 string. It's now changed to take a pointer and a length, to allow for
4350 binary data in such strings.
4352 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4354 23. Exim was not respecting finduser_retries when seeking the login of the
4355 uid under which it was called; it was always trying 10 times. (The default
4356 setting of finduser_retries is zero.) Also, it was sleeping after the final
4357 failure, which is pointless.
4359 24. Implemented tls_on_connect_ports.
4361 25. Implemented acl_smtp_predata.
4363 26. If the domain in control=submission is set empty, Exim assumes that the
4364 authenticated id is a complete email address when it generates From: or
4365 Sender: header lines.
4367 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4368 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4369 chown and chgrp in /bin and hostname in /usr/bin.
4371 28. Exim was keeping the "process log" file open after each use, just as it
4372 does for the main log. This opens the possibility of it remaining open for
4373 long periods when the USR1 signal hits a daemon. Occasional processlog
4374 errors were reported, that could have been caused by this. Anyway, it seems
4375 much more sensible not to leave this file open at all, so that is what now
4378 29. The long-running daemon process does not normally write to the log once it
4379 has entered its main loop, and it closes the log before doing so. This is
4380 so that log files can straightforwardly be renamed and moved. However,
4381 there are a couple of unusual error situations where the daemon does write
4382 log entries, and I had neglected to close the log afterwards.
4384 30. The text of an SMTP error response that was received during a remote
4385 delivery was being truncated at 512 bytes. This is too short for some of
4386 the long messages that one sometimes sees. I've increased the limit to
4389 31. It is now possible to make retry rules that apply only when a message has a
4390 specific sender, in particular, an empty sender.
4392 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4393 it possible to be selective about when SMTP synchronization is enforced.
4395 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4397 32. Implemented hosts_connection_nolog.
4399 33. Added an ACL for QUIT.
4401 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4404 35. Added mailbox_size and mailbox_filecount to appendfile.
4406 36. Added control = no_multiline_responses to ACLs.
4408 37. There was a bug in the logic of the code that waits for the clock to tick
4409 in the case where the clock went backwards by a substantial amount such
4410 that the microsecond fraction of "now" was more than the microsecond
4411 fraction of "then" (but the whole seconds number was less).
4413 38. Added support for the libradius Radius client library this is found on
4414 FreeBSD (previously only the radiusclient library was supported).
4420 1. When certain lookups returned multiple values in the form name=value, the
4421 quoting of the values was not always being done properly. Specifically:
4422 (a) If the value started with a double quote, but contained no whitespace,
4424 (b) If the value contained whitespace other than a space character (i.e.
4425 tabs or newlines or carriage returns) it was not quoted.
4426 This fix has been applied to the mysql and pgsql lookups by writing a
4427 separate quoting function and calling it from the lookup code. The fix
4428 should probably also be applied to nisplus, ibase and oracle lookups, but
4429 since I cannot test any of those, I have not disturbed their existing code.
4431 2. A hit in the callout cache for a specific address caused a log line with no
4432 reason for rejecting RCPT. Now it says "Previous (cached) callout
4433 verification failure".
4435 3. There was an off-by-one bug in the queryprogram router. An over-long
4436 return line was truncated at 256 instead of 255 characters, thereby
4437 overflowing its buffer with the terminating zero. As well as fixing this, I
4438 have increased the buffer size to 1024 (and made a note to document this).
4440 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4441 when Exim is waiting for an SMTP response from a remote server, Exim
4442 restarts its select() call on the socket, thereby resetting its timeout.
4443 This is not a problem when such interrupts are rare. Somebody set up a cron
4444 job to run exiwhat every 2 minutes, which is less than the normal select()
4445 timeout (5 or 10 minutes). This meant that the select() timeout never
4446 kicked in because it was always reset. I have fixed this by comparing the
4447 time when an interrupt arrives with the time at the start of the first call
4448 to select(). If more time than the timeout has elapsed, the interrupt is
4449 treated as a timeout.
4451 5. Some internal re-factoring in preparation for the addition of Sieve
4452 extensions (by MH). In particular, the "personal" test is moved to a
4453 separate function, and given an option for scanning Cc: and Bcc: (which is
4454 not set for Exim filters).
4456 6. When Exim created an email address using the login of the caller as the
4457 local part (e.g. when creating a From: or Sender: header line), it was not
4458 quoting the local part when it contained special characters such as @.
4460 7. Installed new OpenBSD configuration files.
4462 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4463 try to make them clearer.
4465 9. Callout options, other than the timeout value, were being ignored when
4466 verifying sender addresses in header lines. For example, when using
4468 verify = header_sender/callout=no_cache
4470 the cache was (incorrectly) being used.
4472 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4473 only the Cygwin environment.
4475 11. When return_path_on_delivery was set as a log selector, if different remote
4476 addresses in the same message used different return paths and parallel
4477 remote delivery occurred, the wrong values would sometimes be logged.
4478 (Whenever a remote delivery process finished, the return path value from
4479 the most recently started remote delivery process was logged.)
4481 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4482 lines when AUTH and/or TLS are in use. This is the "received protocol"
4483 field. Exim used to use "asmtp" for authenticated SMTP, without any
4484 indication (in the protocol name) for TLS use. Now it follows the RFC and
4485 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4486 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4487 names appear in log lines as well as in Received: header lines.
4489 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4490 extensions, and comparison tests, and to fix some bugs.
4492 14. Changes to the "personal" filter test:
4494 (1) The test was buggy in that it was just doing the equivalent of
4495 "contains" tests on header lines. For example, if a user's address was
4496 anne@some.where, the "personal" test would incorrectly be true for
4498 To: susanne@some.where
4500 This test is now done by extracting each address from the header in turn,
4501 and checking the entire address. Other tests that are part of "personal"
4502 are now done using regular expressions (for example, to check local parts
4503 of addresses in From: header lines).
4505 (2) The list of non-personal local parts in From: addresses has been
4506 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4507 taken from the Sieve specification recommendations.
4509 (3) If the message contains any header line starting with "List-" it is
4510 treated as non-personal.
4512 (4) The test for "circular" in the Subject: header line has been removed
4513 because it now seems ill-conceived.
4515 15. Minor typos in src/EDITME comments corrected.
4517 16. Installed latest exipick from John Jetmore.
4519 17. If headers_add on a router specified a text string that was too long for
4520 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4521 of string_sprintf() is now avoided.
4523 18. $message_body_size was not set (it was always zero) when running the DATA
4524 ACL and the local_scan() function.
4526 19. For the "mail" command in an Exim filter, no default was being set for
4527 the once_repeat time, causing a random time value to be used if "once" was
4528 specified. (If the value happened to be <= 0, no repeat happened.) The
4529 default is now 0s, meaning "never repeat". The "vacation" command was OK
4530 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4531 (I found it when inspecting the code).
4533 20. There is now an overall timeout for performing a callout verification. It
4534 defaults to 4 times the callout timeout, which applies to individual SMTP
4535 commands during the callout. The overall timeout applies when there is more
4536 than one host that can be tried. The timeout is checked before trying the
4537 next host. This prevents very long delays if there are a large number of
4538 hosts and all are timing out (e.g. when the network connections are timing
4539 out). The value of the overall timeout can be changed by specifying an
4540 additional sub-option for "callout", called "maxwait". For example:
4542 verify = sender/callout=5s,maxwait=20s
4544 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4545 to the end before writing, but this should make it even safer).
4547 22. Exim was forgetting that it had advertised PIPELINING for the second and
4548 subsequent messages on an SMTP connection. It was also not resetting its
4549 memory on STARTTLS and an internal HELO.
4551 23. When Exim logs an SMTP synchronization error within a session, it now
4552 records whether PIPELINING has been advertised or not.
4554 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4555 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4556 rather than long int.
4558 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4560 26. Added the never_mail option to autoreply.
4566 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4567 crash if the getsockname() call failed; this can happen if a connection is
4568 closed very soon after it is established. The problem was simply in the
4569 order in which certain operations were done, causing Exim to try to write
4570 to the SMTP stream before it had set up the file descriptor. The bug has
4571 been fixed by making things happen in the correct order.
4577 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4578 before the connection was closed, thus losing the rejection response.
4580 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4581 some early Solaris releases, but causes trouble in current releases where
4582 socklen_t is defined.
4584 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4587 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4590 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4591 that releasing the top of it at the end releases what was used for sub-
4592 expansions (unless the block got too big). However, discard this block if
4593 the first thing is a variable or header, so that we can use its block when
4594 it is dynamic (useful for very large $message_headers, for example).
4596 6. Lookups now cache *every* query, not just the most recent. A new, separate
4597 store pool is used for this. It can be recovered when all lookup caches are
4598 flushed. Lookups now release memory at the end of their result strings.
4599 This has involved some general refactoring of the lookup sources.
4601 7. Some code has been added to the store_xxx() functions to reduce the amount
4602 of flapping under certain conditions.
4604 8. log_incoming_interface used to affect only the <= reception log lines. Now
4605 it causes the local interface and port to be added to several more SMTP log
4606 lines, for example "SMTP connection from", and rejection lines.
4608 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4610 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4612 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4613 could be overwritten at the end of the current message (or the start of a
4614 new message if it was set in a HELO ACL). The value is now preserved for
4615 the duration of the SMTP connection.
4617 12. If a transport had a headers_rewrite setting, and a matching header line
4618 contained an unqualified address, that address was qualified, even if it
4619 did not match any rewriting rules. The underlying bug was that the values
4620 of the flags that permit the existence of unqualified sender and recipient
4621 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4622 non-local messages, and by -bnq for local messages) were not being
4623 preserved with the message after it was received.
4625 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4626 "next input=" as part of the text comprising the host identity instead of
4627 the correct text. The code was using the same buffer for two different
4628 strings. However, depending on which order the printing function evaluated
4629 its arguments, the bug did not always show up. Under Linux, for example, my
4630 test suite worked just fine.
4632 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4633 doesn't work with some older versions of Perl. It has been changed to "my",
4634 which in any case is probably the better facility to use.
4636 15. A really picky compiler found some instances of statements for creating
4637 error messages that either had too many or two few arguments for the format
4640 16. The size of the buffer for calls to the DNS resolver has been increased
4641 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4642 for addresses that have a lot of PTR records. This alleviates a problem; it
4643 does not fully solve it.
4645 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4646 buffer now truncates the list and logs the incident, which is the same
4647 action as happens when Exim is looking up a host name and its aliases.
4648 Previously in this situation something unpredictable would happen;
4649 sometimes it was "internal error: store_reset failed".
4651 18. If a server dropped the connection unexpectedly when an Exim client was
4652 using GnuTLS and trying to read a response, the client delivery process
4653 crashed while trying to generate an error log message.
4655 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4656 string, for example:
4658 warn message = H1: something\nH2: something
4660 the text was added as a single header line from Exim's point of view
4661 though it ended up OK in the delivered message. However, searching for the
4662 second and subsequent header lines using $h_h2: did not work. This has been
4663 fixed. Similarly, if a system filter added multiple headers in this way,
4664 the routers could not see them.
4666 20. Expanded the error message when iplsearch is called with an invalid key to
4667 suggest using net-iplsearch in a host list.
4669 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4670 ACLs are no longer actually imposed (and a message to that effect is
4673 22. If a "gecos" field in a passwd entry contained escaped characters, in
4674 particular, if it contained a \" sequence, Exim got it wrong when building
4675 a From: or a Sender: header from that name. A second bug also caused
4676 incorrect handling when an unquoted " was present following a character
4677 that needed quoting.
4679 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4680 was not being matched caselessly.
4682 24. Arranged for all hyphens in the exim.8 source to be escaped with
4685 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4686 properly. Recipient callout cache records were still being keyed to include
4687 the sender, even when use_sender was set false. This led to far more
4688 callouts that were necessary. The sender is no longer included in the key
4689 when use_sender is false.
4691 26. Added "control = submission" modifier to ACLs.
4693 27. Added the ${base62d: operator to decode base 62 numbers.
4695 28. dnsdb lookups can now access SRV records.
4697 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4698 the configuration file.
4700 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4701 (-v) mode. This makes the output for a verbose queue run more intelligible.
4703 31. Added a use_postmaster feature to recipient callouts.
4705 32. Added the $body_zerocount variable, containing the number of binary zero
4706 bytes in the message body.
4708 33. The time of last modification of the "new" subdirectory is now used as the
4709 "mailbox time last read" when there is a quota error for a maildir
4712 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4714 35. Added +ignore_unknown as a special item in host lists.
4716 36. Code for decoding IPv6 addresses in host lists is now included, even if
4717 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4718 address was recognized as an IP address, but was then not correctly decoded
4719 into binary, causing unexpected and incorrect effects when compared with
4726 1. Very minor rewording of debugging text in manualroute to say "list of
4727 hosts" instead of "hostlist".
4729 2. If verify=header_syntax was set, and a header line with an unqualified
4730 address (no domain) and a large number of spaces between the end of the
4731 name and the colon was received, the reception process suffered a buffer
4732 overflow, and (when I tested it) crashed. This was caused by some obsolete
4733 code that should have been removed. The fix is to remove it!
4735 3. When running in the test harness, delay a bit after writing a bounce
4736 message to get a bit more predictability in the log output.
4738 4. Added a call to search_tidyup() just before forking a reception process. In
4739 theory, someone could use a lookup in the expansion of smtp_accept_max_
4740 per_host which, without the tidyup, could leave open a database connection.
4742 5. Added the variables $recipient_data and $sender_data which get set from a
4743 lookup success in an ACL "recipients" or "senders" condition, or a router
4744 "senders" option, similar to $domain_data and $local_part_data.
4746 6. Moved the writing of debug_print from before to after the "senders" test
4749 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4750 problems for message scanning, either using a data ACL, or using
4751 local_scan() because the Received: header was not generated till after they
4752 were called (in order to set the time as the time of reception completion).
4753 I have revised the way this works. The header is now generated after the
4754 body is received, but before the ACL or local_scan() are called. After they
4755 are run, the timestamp in the header is updated.
4761 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4762 before starting a queue runner without re-exec. This happened only when
4763 deliver_drop_privilege was set or when the Exim user was set to root. The
4764 effect of the bug was that timeouts during subsequent deliveries caused
4765 crashes instead of being properly handled. The handler is now left at its
4766 default (and expected) setting.
4768 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4769 message, again when deliver_drop_privilege is set or Exim is run as root.
4770 The bug described in (1) was not present in this case, but the tidying up
4771 of the other signals was missing. I have made the two cases consistent.
4773 3. The ignore_target_hosts setting on a manualroute router was being ignored
4774 for hosts that were looked up using the /MX notation.
4776 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4779 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4780 operated on the sender address. After changing the $sender_address to <>
4781 for the sender address verify, Exim was re-instated it as the original
4782 (before rewriting) address, but remembering that it had rewritten it, so it
4783 wasn't rewriting it again. This bug also had the effect of breaking the
4784 sender address verification caching when the sender address was rewritten.
4786 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4787 This has been changed so that if the ip literal address matches
4788 ignore_target_hosts, the router declines.
4790 7. Added expansion conditions match_domain, match_address, and match_local_
4791 part (NOT match_host).
4793 8. The placeholder for the Received: header didn't have a length field set.
4795 9. Added code to Exim itself and to exim_lock to test for a specific race
4796 condition that could lead to file corruption when using MBX delivery. The
4797 issue is with the lockfile that is created in /tmp. If this file is removed
4798 after a process has opened it but before that process has acquired a lock,
4799 there is the potential for a second process to recreate the file and also
4800 acquire a lock. This could lead to two Exim processes writing to the file
4801 at the same time. The added code performs the same test as UW imapd; it
4802 checks after acquiring the lock that its file descriptor still refers to
4803 the same named file.
4805 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4806 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4809 11. Added the smtp_active_hostname option. If used, this will typically be made
4810 to depend on the incoming interface address. Because $interface_address is
4811 not set up until the daemon has forked a reception process, error responses
4812 that can happen earlier (such as "too many connections") no longer contain
4815 12. If an expansion in a condition on a "warn" statement fails because a lookup
4816 defers, the "warn" statement is abandoned, and the next ACL statement is
4817 processed. Previously this caused the whole ACL to be aborted.
4819 13. Added the iplsearch lookup type.
4821 14. Added ident_timeout as a log selector.
4823 15. Added tls_certificate_verified as a log selector.
4825 16. Added a global option tls_require_ciphers (compare the smtp transport
4826 option of the same name). This controls incoming TLS connections.
4828 17. I finally figured out how to make tls_require_ciphers do a similar thing
4829 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4830 before starting the TLS session.
4832 18. Tabs are now shown as \t in -bP output.
4834 19. If the log selector return_path_on_delivery was set, Exim crashed when
4835 bouncing a message because it had too many Received: header lines.
4837 20. If two routers both had headers_remove settings, and the first one included
4838 a superfluous trailing colon, the final name in the first list and the
4839 first name in the second list were incorrectly joined into one item (with a
4840 colon in the middle).
4846 1. Added -C and -D options to the exinext utility, mainly to make it easier
4847 to include in the automated testing, but these could be helpful when
4848 multiple configurations are in use.
4850 2. The exinext utility was not formatting the output nicely when there was
4851 an alternate port involved in the retry record key, nor when there was a
4852 message id as well (for retries that were specific to a specific message
4853 and a specific host). It was also confused by IPv6 addresses, because of
4854 the additional colons they contain. I have fixed the IPv4 problem, and
4855 patched it up to do a reasonable job for IPv6.
4857 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4858 delivery, the log line now contains "pipelined" if PIPELINING was used.
4860 4. An SMTP transport process used to panic and die if the bind() call to set
4861 an explicit outgoing interface failed. This has been changed; it is now
4862 treated in the same way as a connect() failure.
4864 5. A reference to $sender_host_name in the part of a conditional expansion
4865 that was being skipped was still causing a DNS lookup. This no longer
4868 6. The def: expansion condition was not recognizing references to header lines
4869 that used bh_ and bheader_.
4871 7. Added the _cache feature to named lists.
4873 8. The code for checking quota_filecount in the appendfile transport was
4874 allowing one more file than it should have been.
4876 9. For compatibility with Sendmail, the command line option
4884 and sets the incoming protocol and host name (for trusted callers). The
4885 host name and its colon can be omitted when only the protocol is to be set.
4886 Note the Exim already has two private options, -pd and -ps, that refer to
4887 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4888 "s", but I don't think that's a major issue.
4890 10. A number of refactoring changes to the code, none of which should affect
4893 (a) The number of logging options was getting close to filling up the
4894 32-bit word that was used as a bit map. I have split them into two classes:
4895 those that are passed in the argument to log_write(), and those that are
4896 only ever tested independently outside of that function. These are now in
4897 separate 32-bit words, so there is plenty of room for expansion again.
4898 There is no change in the user interface or the logging behaviour.
4900 (b) When building, for example, log lines, the code previously used a
4901 macro that called string_cat() twice, in order to add two strings. This is
4902 not really sufficiently general. Furthermore, there was one instance where
4903 it was actually wrong because one of the argument was used twice, and in
4904 one call a function was used. (As it happened, calling the function twice
4905 did not affect the overall behaviour.) The macro has been replaced by a
4906 function that can join an arbitrary number of extra strings onto a growing
4909 (c) The code for expansion conditions now uses a table and a binary chop
4910 instead of a serial search (which was left over from when there were very
4911 few conditions). Also, it now recognizes conditions like "pam" even when
4912 the relevant support is not compiled in: a suitably worded error message is
4913 given if an attempt is made to use such a condition.
4915 11. Added ${time_interval:xxxxx}.
4917 12. A bug was causing one of the ddress fields not to be passed back correctly
4918 from remote delivery subprocesses. The field in question was not being
4919 subsequently used, so this caused to problems in practice.
4921 13. Added new log selectors queue_time and deliver_time.
4923 14. Might have fixed a bug in maildirsizefile handling that threw up
4924 "unexpected character" debug warnings, and recalculated the data
4925 unnecessarily. In any case, I expanded the warning message to give more
4928 15. Added the message "Restricted characters in address" to the statements in
4929 the default ACL that block characters like @ and % in local parts.
4931 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4932 Three changes have been made:
4934 (a) There was a serious bug; a negative response to MAIL caused the whole
4935 recipient domain to be cached as invalid, thereby blocking all messages
4936 to all local parts at the same domain, from all senders. This bug has
4937 been fixed. The domain is no longer cached after a negative response to
4938 MAIL if the sender used is not empty.
4940 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4943 (c) A new callout option, "use_sender" has been added for people who want
4944 the modified behaviour.
4950 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4953 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4954 indeed breaks things for older releases.
4956 3. Added additional logging to the case where there is a problem reading data
4957 from a filter that is running in a subprocess using a pipe, in order to
4958 try to track down a specific problem.
4960 4. Testing facility fudge: when running in the test harness and attempting
4961 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4962 getting "No route to host". Convert this to a timeout.
4964 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4967 6. Some OS don't have socklen_t but use size_t instead. This affects the
4968 fifth argument of getsockopt() amongst other things. This is now
4969 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4970 can be set for individual OS. I have set it for SunOS5, OSF1, and
4971 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4972 some earlier ones do not.
4974 7. Change 4.30/15 was not doing the test caselessly.
4976 8. The standard form for an IPv6 address literal was being rejected by address
4977 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4978 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4979 this, as well as the form without the "IPv6" on the front (but only when
4980 address literals are enabled, of course).
4982 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4984 10. Exim crashed if a message with an empty sender address specified by -f
4985 encountered a router with an errors_to setting. This could be provoked only
4986 by a command such as
4990 where an empty string was supplied; "<>" did not hit this bug.
4992 11. Installed PCRE release 4.5.
4994 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4995 remained set. It is now erased.
4997 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4998 times from message ids (which are base 36 rather than the normal 62).
5000 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5001 were being counted as actual protocol errors, and logged if the log
5002 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5003 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5004 and DATA following a set of rejected RCPTs do not count as protocol errors.
5005 In other words, Exim assumes they were pipelined, though this may not
5006 actually be the case. Of course, in all cases the client gets an
5007 appropriate error code.
5009 15. If a lookup fails in an ACL condition, a message about the failure may
5010 be available; it is used if testing the ACL cannot continue, because most
5011 such messages specify what the cause of the deferral is. However, some
5012 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5013 that caused an old message to be retained and used if a later statement
5014 caused a defer, replacing the real cause of the deferral.
5016 16. If an IP address had so many PTR records that the DNS lookup buffer
5017 was not large enough to hold them, Exim could crash while trying to process
5018 the truncated data. It now detects and logs this case.
5020 17. Further to 4.21/58, another change has been made: if (and only if) the
5021 first line of a message (the first header line) ends with CRLF, a bare LF
5022 in a subsequent header line has a space inserted after it, so as not to
5023 terminate the header.
5025 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5026 unnecessarily, used atoi() instead of strtol(), and didn't check the
5027 termination when getting file sizes from file names by regex.
5029 19. Completely re-implemented the support for maildirsize files, in the light
5030 of a number of problems with the previous contributed implementation
5031 (4.30/29). In particular:
5033 . If the quota is zero, the maildirsize file is maintained, but no quota is
5036 . If the maildir directory does not exist, it is created before any attempt
5037 to write a maildirsize file.
5039 . The quota value in the file is just a cache; if the quota is changed in
5040 the transport, the new value overrides.
5042 . A regular expression is available for excluding directories from the
5045 20. The autoreply transport checks the characters in options that define the
5046 message's headers; it allows continued headers, but it was checking with
5047 isspace() after an embedded newline instead of explicitly looking for a
5050 21. If all the "regular" hosts to which an address was routed had passed their
5051 expiry times, and had not reached their retry times, the address was
5052 bounced, even if fallback hosts were defined. Now Exim should go on to try
5055 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5056 equivalent code in the SMTP transport. Some hosts send humungous responses
5057 to HELO/EHLO, more than 1024 it seems.
5059 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5060 gives compiler warnings in some environments. I've now done it "properly",
5063 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5064 (because of problems with the built-in one) was declared to return uschar *
5065 instead of char *, causing compiler failure.
5067 25. Fixed a file descriptor leak when processing alias/forward files.
5069 26. Fixed a minor format string issue in dbfn.c.
5071 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5073 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5074 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5077 29. When a sender address is verified, it is cached, to save repeating the test
5078 when there is more than one recipient in a message. However, when the
5079 verification involves a callout, it is possible for different callout
5080 options to be set for different recipients. It is too complicated to keep
5081 track of this in the cache, so now Exim always runs a verification when a
5082 callout is required, relying on the callout cache for the optimization.
5083 The overhead is duplication of the address routing, but this should not be
5086 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5087 to be verified with callout=postmaster, and the main callout worked but the
5088 postmaster check failed, the verification correctly failed. However, if a
5089 subsequent RCPT command asked for sender verification *without* the
5090 postmaster check, incorrect caching caused this verification also to fail,
5093 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5094 it was not caching the DNS options (qualify_single, search_parents) that
5095 were used when the lookup failed. A subsequent lookup with different
5096 options therefore always gave the same answer, though there were cases
5097 where it should not have. (Example: a "domains = !$mx_any" option on a
5098 dnslookup router: the "domains" option is always processed without any
5099 widening, but the router might have qualify_single set.) Now Exim uses the
5100 cached value only when the same options are set.
5102 32. Added John Jetmore's "exipick" utility to the distribution.
5104 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5105 than a timeout (e.g. a certificate is required, and is not provided), an
5106 Exim server now closes the connection immediately. Previously it waited for
5107 the client to close - but if the client is SSL, it seems that they each
5108 wait for each other, leading to a delay before one of them times out.
5110 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5111 maintained 0.8.x compatibility because I don't think many are using it, and
5112 it is clearly obsolete.
5114 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5117 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5118 client certificate was expired. A simple patch fixes this, though I don't
5119 understand the full logic of why the verify callback is called multiple
5122 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5123 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5124 which causes problems with some clients (such as the Certicom SSL Plus
5125 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5126 disables the coutermeasure allowing Eudora to connect."
5128 38. Exim was not checking that a write() to a log file succeeded. This could
5129 lead to Bad Things if a log got too big, in particular if it hit a file
5130 size limit. Exim now panics and dies if it cannot write to a log file, just
5131 as it does if it cannot open a log file.
5133 39. Modified OS/Makefile-Linux so that it now contains
5135 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5137 The two -D definitions ensure that Exim is compiled with large file
5138 support, which makes it possible to handle log files that are bigger than
5141 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5142 instance) a domain was checked against a named list that involved a lookup,
5143 causing $domain_data to be set, then another domain was checked against the
5144 same list, then the first domain was re-checked, the value of $domain_data
5145 after the final check could be wrong. In particular, if the second check
5146 failed, it could be set empty. This bug probably also applied to
5149 41. The strip_trailing_dot option was not being applied to the address given
5150 with the -f command-line option.
5152 42. The code for reading a message's header from the spool was incrementing
5153 $received_count, but never initializing it. This meant that the value was
5154 incorrect (doubled) while delivering a message in the same process in which
5155 it was received. In the most common configuration of Exim, this never
5156 happens - a fresh exec is done - but it can happen when
5157 deliver_drop_privilege is set.
5159 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5160 it now includes up to 150 characters of the unexpected data in the log
5163 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5164 and building data strings. The size of both of these buffers was 10 000
5165 bytes - far larger than anybody would *ever* want, thought I. Needless to
5166 say, somebody hit the limit. I have increased the maximum line length to
5167 20 000 and the maximum data length of concatenated lines to 100 000. I have
5168 also fixed two bugs, because there was no checking on these buffers. Tsk,
5169 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5170 buffer is too small.
5172 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5173 lsearch lookups. Now it does.
5175 46. When parsing a route_list item in a manualroute router, a fixed-length
5176 buffer was used for the list of hosts. I made this 1024 bytes long,
5177 thinking that nobody would ever have a list of hosts that long. Wrong.
5178 Somebody had a whole pile of complicated expansion conditions, and the
5179 string was silently truncated, leading to an expansion error. It turns out
5180 that it is easier to change to an unlimited length (owing to other changes
5181 that have happened since this code was originally written) than to build
5182 structure for giving a limitation error. The length of the item that
5183 expands into the list of hosts is now unlimited.
5185 47. The lsearch lookup could not handle data where the length of text line was
5186 more than 4095 characters. Such lines were truncated, leading to shortened
5187 data being returned. It should now handle lines of any length.
5189 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5190 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5193 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5194 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5195 no longer happens in comment lines. A list of replacements is now placed
5196 at the head of all of the source files, except those whose only change is
5197 to replace PERL_COMMAND in the very first #! line.
5199 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5200 messages on the queue, with a bottom-up merge sort, using code contributed
5201 by Michael Haardt. This should make operations like -bp somewhat faster on
5202 large queues. It won't affect queue runners, except when queue_run_in_order
5205 51. Installed eximstats 1.31 in the distribution.
5207 52. Added support for SRV lookups to the dnslookup router.
5209 53. If an ACL referred to $message_body or $message_body_end, the value was not
5210 reset for any messages that followed in the same SMTP session.
5212 54. The store-handling optimization for building very long strings was not
5213 differentiating between the different store pools. I don't think this
5214 actually made any difference in practice, but I've tidied it.
5216 55. While running the routers to verify a sender address, $sender_address
5217 was still set to the sender address. This is wrong, because when routing to
5218 send a bounce to the sender, it would be empty. Therefore, I have changed
5219 it so that, while verifying a sender address, $sender_address is set to <>.
5220 (There is no change to what happens when verifying a recipient address.)
5222 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5223 target A or AAAA records (if not already returned) without resetting the
5224 qualify_single or search_parents options of the DNS resolver. These are
5225 inappropriate in this case because the targets of MX and SRV records must
5226 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5227 target that, when qualified, matched something in the local domain. These
5228 two options are now turned off when doing these lookups.
5230 57. It seems that at least some releases of Reiserfs (which does not have the
5231 concept of a fixed number of inodes) returns zero and not -1 for the
5232 number of available inodes. This interacted badly with check_spool_inodes,
5233 which assumed that -1 was the "no such thing" setting. What I have done is
5234 to check that the total number of inodes is greater than zero before doing
5235 the test of how many are available.
5237 58. When a "warn" ACL statement has a log_message modifier, the message is
5238 remembered, and not repeated. This is to avoid a lot of repetition when a
5239 message has many recipients that cause the same warning to be written.
5240 Howewer, Exim was preserving the list of already written lines for an
5241 entire SMTP session, which doesn't seem right. The memory is now reset if a
5242 new message is started.
5244 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5245 debugging output unless log_rewrite was also set.
5247 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5248 of (int)(handle) when we know that handle contains (void *)(-1).
5250 61. The Exim daemon panic-logs an error return when it closes the incoming
5251 connection. However "connection reset by peer" seems to be common, and
5252 isn't really an error worthy of noting specially, so that particular error
5255 62. When Exim is trying to find all the local interfaces, it used to panic and
5256 die if the ioctl to get the interface flags failed. However, it seems that
5257 on at least one OS (Solaris 9) it is possible to have an interface that is
5258 included in the list of interfaces, but for which you get a failure error
5259 for this call. This happens when the interface is not "plumbed" into a
5260 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5261 failure of the "get flags" call assumes that the interface is down.
5263 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5264 makes life easier for people who are doing arithmetic on fields extracted
5265 from dates, where you often get leading zeros that should not be
5266 interpreted as octal.
5268 64. Added qualify_domain to the redirect router, to override the global
5271 65. If a pathologically long header line contained very many addresses (the
5272 report of this problem mentioned 10 000) and each of them was rewritten,
5273 Exim could use up a very large amount of memory. (It kept on making new
5274 copies of the header line as it rewrote, and never released the old ones.)
5275 At the expense of a bit more processing, the header rewriting function has
5276 been changed so that it no longer eats memory in this way.
5278 66. The generation of the Received: header has been moved from the time that a
5279 message starts to be received, to the time that it finishes. The timestamp
5280 in the Received: header should now be very close to that of the <= log
5281 line. There are two side-effects of this change:
5283 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5284 logged header lines no longer include the local Received: line, because
5285 it has not yet been created. The same applies to a copy of the message
5286 that is returned to a non-SMTP sender when a message is rejected.
5288 (b) When a filter file is tested using -bf, no additional Received: header
5289 is added to the test message. After some thought, I decided that this
5292 This change does not affect the value of $received_for. It is still set
5293 after address rewriting, but before local_scan() is called.
5295 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5297 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5298 gave an unhelpful panic error message, and a defer error. I have managed to
5299 change this behaviour so that it now rejects any supplied certificate,
5300 which seems right, as the list of acceptable certificates is empty.
5302 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5303 gave an unhelpful defer error. I have not managed to make this reject any
5304 supplied certificates, but the error message it gives is "no certificate
5305 supplied", which is not helpful.
5307 70. exigrep's output now also includes lines that are not associated with any
5308 message, but which match the given pattern. Implemented by a patch from
5309 Martin Sluka, which also tidied up the Perl a bit.
5311 71. Recipient callout verification, like sender verification, was using <> in
5312 the MAIL FROM command. This isn't really the right thing, since the actual
5313 sender may affect whether the remote host accepts the recipient or not. I
5314 have changed it to use the actual sender in the callout; this means that
5315 the cache record is now keyed on a recipient/sender pair, not just the
5316 recipient address. There doesn't seem to be a real danger of callout loops,
5317 since a callout by the remote host to check the sender would use <>.
5318 [SEE ABOVE: changed after hitting problems.]
5320 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5321 temporary errors. However, in the case of such a code being given after
5322 the end of a data transmission (i.e. after ".") Exim was failing to write
5323 a retry record for the message. (Yes, there was some broken host that was
5324 actually sending 8xx at this point.)
5326 73. An unknown lookup type in a host list could cause Exim to panic-die when
5327 the list was checked. (An example that provoked this was putting <; in the
5328 middle of a list instead of at the start.) If this happened during a DATA
5329 ACL check, a -D file could be left lying around. This kind of configuration
5330 error no longer causes Exim to die; instead it causes a defer errror. The
5331 incident is still logged to the main and panic logs.
5333 74. Buglet left over from Exim 3 conversion. The message "too many messages
5334 in one connection" was written to the rejectlog but not the mainlog, except
5335 when address rewriting (yes!) was being logged.
5337 75. Added write_rejectlog option.
5339 76. When a system filter was run not as root (that is, when system_filter_user
5340 was set), the values of the $n variables were not being returned to the
5341 main process; thus, they were not subsequently available in the $sn
5344 77. Added +return_path_on_delivery log selector.
5346 78. A connection timeout was being treated differently from recipients deferred
5347 when testing hosts_max_try with a message that was older than the host's
5348 retry timeout. (The host should not be counted, thus allowing all hosts to
5349 be tried at least once before bouncing.) This may have been the cause of an
5350 occasionally reported bug whereby a message would remain on the queue
5351 longer than the retry timeout, but would be bounced if a delivery was
5352 forced. I say "may" because I never totally pinned down the problem;
5353 setting up timeout/retry tests is difficult. See also the next item.
5355 79. The ultimate address timeout was not being applied to errors that involved
5356 a combination of host plus message (for example, a timeout on a MAIL
5357 command). When an address resolved to a number of possible hosts, and they
5358 were not all tried for each delivery (e.g. because of hosts_max_try), a
5359 message could remain on the queue longer than the retry timeout.
5361 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5364 81. Fixed an obscure SMTP outgoing bug which required at least the following
5365 conditions: (a) there was another message waiting for the same server;
5366 (b) the server returned 5xx to all RCPT commands in the first message so
5367 that the message was not completed; (c) the server dropped the connection
5368 or gave a negative response to the RSET that Exim sends to abort the
5369 transaction. The observed case was a dropped connection after DATA that had
5370 been sent in pipelining mode. That is, the server had advertised PIPELINING
5371 but was not implementing it correctly. The effect of the bug was incorrect
5372 behaviour, such as trying another host, and this could lead to a crash.
5378 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5379 and daemon.c were passed as pointers to ints; they should have been
5380 pointers to socklen_t variables (which are typically unsigned ints).
5382 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5385 3. Fixed a really odd bug that affected only the testing scheme; patching a
5386 certain fixed string in the binary changed the value of another string that
5387 happened to be identical to the end of the original first string.
5389 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5390 name", it returns that address as the IP address. On some operating
5391 systems (e.g. Solaris), it also passes back the IP address string as the
5392 "host name". However, on others (e.g. Linux), it passes back an empty
5393 string. Exim wasn't checking for this, and was changing the host name to an
5394 empty string, assuming it had been canonicized.
5396 5. Although rare, it is permitted to have more than one PTR record for a given
5397 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5398 all the names associated with an address, because they do in Solaris.
5399 However, it seems that they do not in Linux for data that comes from the
5400 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5401 I found this out when I moved to a new Linux workstation and tried to run
5402 the Exim test suite.
5404 To get round this problem I have changed the code so that it now does its
5405 own call to the DNS to look up PTR records when searching for a host name.
5406 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5407 addresses that are only in /etc/hosts are still found.
5409 This behaviour is, however, controlled by an option called host_lookup_
5410 order, which defaults to "bydns:byaddr". If people want to use the other
5411 order, or indeed, just use one or the other means of lookup, they can
5412 specify it in this variable.
5414 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5415 some operating systems, this comes back from gethostbyaddr() as an empty
5416 string, and this is what Exim used to test for. However, it seems that in
5417 other systems, "." is yielded. Exim now tests for this case too.
5419 7. The values of check_spool_space and check_log_space are now held internally
5420 as a number of kilobytes instead of an absolute number of bytes. If a
5421 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5422 kilobyte. This means that much larger values can be stored.
5424 8. Exim monitor: an attempt to get the action menu when not actually pointing
5425 at a message produces an empty menu entitled "No message selected". This
5426 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5427 no entries in it ("Shell widget menu has zero width and/or height"). So I
5428 have added a single, blank menu entry in this case.
5430 9. Added ${quote_local_part.
5432 10. MIME decoding is now applied to the contents of Subject: header lines when
5435 11. Now that a reference to $sender_host_address automatically causes a reverse
5436 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5437 host lookup before query-style lookups in lists that might use this
5438 variable. This has therefore been abolished, and the "net-" prefix is no
5439 longer necessary for query-style lookups.
5441 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5442 have been a typo for LFLAGS, so it has been changed.
5444 13. The install script calls Exim with "-C /dev/null" in order to find the
5445 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5446 to be output. Howeve, since Exim outputs its version number before the
5447 error, it didn't break the script. It just looked ugly. I fixed this by
5448 always allowing "-C /dev/null" if the caller is root.
5450 14. Ignore overlarge ACL variable number when reading spool file - insurance
5451 against a later release with more variables having written the file.
5453 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5454 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5455 this, as well as the form without the "IPv6" on the front.
5457 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5458 OS/Makefile-Darwin file.
5460 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5461 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5463 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5464 variable. However, some OS use ints for inodes. I've added cast to long int
5465 to get rid of the compiler warning.
5467 19. I had forgotten to lock out "/../" in configuration file names when
5468 ALT_CONFIG_PREFIX was set.
5470 20. Routers used for verification do not need to specify transports. However,
5471 if such a router generated a host list, and callout was configured, Exim
5472 crashed, because it could not find a port number from the (non-existent)
5473 transport. It now assumes port 25 in this circumstance.
5475 21. Added the -t option to exigrep.
5477 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5478 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5479 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5480 them is set, LOOKUP_LSEARCH is forced.
5482 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5484 24. Added sender and host information to the "rejected by local_scan()" log
5485 line; previously there was no indication of these.
5487 25. Added .include_if_exists.
5489 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5490 Linux. It turns out that not all file systems support this. Apparently some
5491 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5492 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5493 not supported on the file descriptor, is now ignored when Exim is trying to
5494 sync a directory. This applies only to Linux.
5496 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5498 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5499 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5502 29. Added support for maildirsize files from supplied patch (modified a bit).
5504 30. The use of :fail: followed by an empty string could lead Exim to respond to
5505 sender verification failures with (e.g.):
5507 550 Verification failed for <xxx>
5508 550 Sender verify failed
5510 where the first response line was missing the '-' that indicates it is not
5511 the final line of the response.
5513 31. The loop for finding the name of the user that called Exim had a hardwired
5514 limit of 10; it now uses the value of finduser_retries, which is used for
5515 all other user lookups.
5517 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5520 33. Exim was neglecting to zero errno before one call of strtol() when
5521 expanding a string and expecting an integer value. On some systems this
5522 resulted in spurious "integer overflow" errors. Also, it was casting the
5523 result into an int without checking.
5525 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5526 did not work. The code looks as if it has *never* worked, though it appears
5527 to have been documented since at least releast 1.62. I have made it work.
5529 35. The "timeout_DNS" error in retry rules, also documented since at least
5530 1.62, also never worked. As it isn't clear exactly what this means, and
5531 clearly it isn't a major issue, I have abolished the feature by treating it
5532 as "timeout", and writing a warning to the main and panic logs.
5534 36. The display of retry rules for -brt wasn't always showing the error code
5537 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5538 timeout_connect_A, timeout_connect_MX.
5540 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5541 to the empty sender.
5543 39. The daemon was not analysing the content of -oX till after it had closed
5544 stderr and disconnected from the controlling terminal. This meant that any
5545 syntax errors were only noted on the panic log, and the return code from
5546 the command was 0. By re-arranging the code a little, I've made the
5547 decoding happen first, so such errors now appear on stderr, and the return
5548 code is 1. However, the actual setting up of the sockets still happens in
5549 the disconnected process, so errors there are still only recorded on the
5552 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5553 connections (as happens on some IP stacks) was logged at start up time as
5554 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5555 it from "IPv6 and IPv4", which means that two separate sockets are being
5558 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5559 says whether AF_INET or AF_INET6 was passed as an argument.
5561 42. Exiwhat output was messed up when time zones were included in log
5564 43. Exiwhat now gives more information about the daemon's listening ports,
5565 and whether -tls-on-connect was used.
5567 44. The "port" option of the smtp transport is now expanded.
5569 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5570 silently ignored. Now an error message is written to the main and panic
5573 46. There's a new ACL modifier called "logwrite" which writes to a log file
5574 as soon as it is encountered.
5576 47. Added $local_user_uid and $local_user_gid at routing time.
5578 48. Exim crashed when trying to verify a sender address that was being
5581 49. Exim was recognizing only a space character after ".include". It now also
5582 recognizes a tab character.
5584 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5585 extracting the relevant information from the specification. The man page no
5586 longer contains scrambled data for the -d option, and I've added a section
5587 at the front about calling Exim under different names.
5589 51. Added "extra_headers" argument to the "mail" command in filter files.
5591 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5594 53. Installed eximstats 1.29.
5596 54. Added transport_filter_timeout as a generic transport option.
5598 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5599 Cc: header lines. This was required by RFC 822, but it not required by RFC
5602 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5603 incoming messages that did not have them. Now it does so only if the
5604 message originates locally, that is, if there is no associated remote host
5605 address. When Resent- header lines are present, this applies to the Resent-
5606 lines rather than the non-Resent- lines.
5608 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5609 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5611 58. Messages for configuration errors now include the name of the main
5612 configuration file - useful now that there may be more than one file in a
5613 list (.included file names were always shown).
5615 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5616 for those rare installations that do not start the daemon as root or run it
5617 setuid root. I've cut out the call to initgroups() if the daemon is not
5620 60. The Exim user and group can now be bound into the binary as text strings
5621 that are looked up at the start of Exim's processing.
5623 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5625 62. Added $mailstore_basename variable.
5627 63. Installed patch to sieve.c from Michael Haardt.
5629 64. When Exim failed to open the panic log after failing to open the main log,
5630 the original message it was trying to log was written to stderr and debug
5631 output, but if they were not available (the usual case in production), it
5632 was lost. Now it is written to syslog before the two lines that record the
5633 failures to open the logs.
5635 65. Users' Exim filters run in subprocesses under the user's uid. It is
5636 possible for a "deliver" command or an alias in a "personal" command to
5637 provoke an address rewrite. If logging of address rewriting is configured,
5638 this fails because the process is not running as root or exim. There may be
5639 a better way of dealing with this, but for the moment (because 4.30 needs
5640 to be released), I have disabled address rewrite logging when running a
5641 filter in a non-root, non-exim process.
5647 1. The buildconfig auxiliary program wasn't quoting the value set for
5648 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5649 not defined. This bug was masked in 4.22 by the effect that was fixed in
5652 2. Some messages that were rejected after a message id was allocated were
5653 shown as "incomplete" by exigrep. It no longer does this for messages that
5654 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5656 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5657 have allow_domain_literals set, the ID did not get logged in the <= line.
5658 Domain literals are now always recognized in Message-ID: header lines.
5660 4. The first argument for a ${extract expansion item is the key name or field
5661 number. Leading and trailing spaces in this item were not being ignored,
5662 causing some misleading effects.
5664 5. When deliver_drop_privilege was set, single queue runner processes started
5665 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5666 same command in the process it spins off) were not dropping privilege.
5668 6. When the daemon running as "exim" started a queue runner, it always
5669 re-executed Exim in the spun-off process. This is a waste of effort when
5670 deliver_drop_privilege is set. The new process now just calls the
5671 queue-runner function directly.
5677 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5680 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5681 ignored. Though the use of -oP was forcing the writing of a pid file, it
5682 was always written to the default place.
5684 3. If the message "no IP address found for host xxxx" is generated during
5685 incoming verification, it is now followed by identification of the incoming
5686 connection (so you can more easily find what provoked it).
5688 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5690 5. Added some features to "harden" Exim a bit more against certain attacks:
5692 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5693 be put in Local/Makefile. This is like the never_users runtime option,
5694 but it cannot be overridden. The default setting is "root".
5696 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5697 prefix string with which any file named in a -C command line option
5700 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5701 is retained for -C and -D only if the caller of Exim is root. Without
5702 it, the exim user may also use -C and -D and retain privilege.
5704 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5705 command line option is disabled.
5707 6. Macro names set by the -D option must start with an upper case letter, just
5708 like macro names defined in the configuration file.
5710 7. Added "dereference=" facility to LDAP.
5712 8. Two instances of the typo "uknown" in the source files are fixed.
5714 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5715 the Configure-Makefile script screwed up while processing it.
5717 10. Incorporated PCRE 4.4.
5719 11. The SMTP synchronization check was not operating right at the start of an
5720 SMTP session. For example, it could not catch a HELO sent before the client
5721 waited for the greeting. There is now a check for outstanding input at the
5722 point when the greeting is written. Because of the duplex, asynchronous
5723 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5724 way, but not yet received, when the check is performed.
5726 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5727 on TCP/IP sockets, because this apparently causes some broken clients to
5730 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5731 unchanged) from the Cygwin maintainer.
5733 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5734 was supported instead of testing for mbx. Effectively a typo.
5736 15. The spa authenticator server code was not checking that the input it
5737 received was valid base64.
5739 16. The debug output line for the "set" modifier in ACLs was not showing the
5740 name of the variable that was being set.
5742 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5744 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5745 lookup on its own. Something else was needed to trigger the lookup. For
5746 example, a match in host_lookup or the need for a host name in a host list.
5747 Now, if $sender_host_name is referenced and the host name has not yet been
5748 looked up, a lookup is performed. If the lookup fails, the variable remains
5749 empty, and $host_lookup_failed is set to "1".
5751 19. Added "eqi" as a case-independent comparison operator.
5753 20. The saslauthd authentication condition could segfault if neither service
5754 nor realm was specified.
5756 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5757 error message that was logged was misleading, and incoming SMTP
5758 connections were dropped. The message is now more accurate, and temporary
5759 errors are given to SMTP connections.
5761 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5762 (except RSET) with a 421 error, until QUIT is received. However, it was
5763 failing to send a response to QUIT.
5765 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5766 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5767 ACL is now run after the helo_try_verify_hosts code.
5769 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5770 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5771 case-independent, so other case variants are also recognized). Apparently
5772 some systems use these upper case variants.
5774 25. If more than two messages were waiting for the same host, and a transport
5775 filter was specified for the transport, Exim sent two messages over the
5776 same TCP/IP connection, and then failed with "socket operation on non-
5777 socket" when it tried to send the third.
5779 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5781 27. The extern definition of crypt16() in expand.c was not being excluded when
5782 the OS had its own crypt16() function.
5784 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5785 as a preferred synonym for return_size_limit, both as an option and as an
5788 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5790 30. Changed the default configuration ACL to relax the local part checking rule
5791 for addresses that are not in any local domains. For these addresses,
5792 slashes and pipe symbols are allowed within local parts, but the sequence
5793 /../ is explicitly forbidden.
5795 31. SPA server authentication was not clearing the challenge buffer before
5798 32. log_message in a "warn" ACL statement was writing to the reject log as
5799 well as to the main log, which contradicts the documentation and doesn't
5800 seem right (because no rejection is happening). So I have stopped it.
5802 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5803 However, I am unable to do any testing of this.
5805 34. Fixed an infelicity in the appendfile transport. When checking directories
5806 for a mailbox, to see if any needed to be created, it was accidentally
5807 using path names with one or more superfluous leading slashes; tracing
5808 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5810 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5811 contents are added to the log line that is written for every discarded
5812 recipient. (Previously a log_message setting was ignored.)
5814 36. The ${quote: operator now quotes the string if it is empty.
5816 37. The install script runs exim in order to find its version number. If for
5817 some reason other than non-existence or emptiness, which it checks, it
5818 could not run './exim', it was installing it with an empty version number,
5819 i.e. as "exim-". This error state is now caught, and the installation is
5822 38. An argument was missing from the function that creates an error message
5823 when Exim fails to connect to the socket for saslauthd authentication.
5824 This could cause Exim to crash, or give a corrupted message.
5826 39. Added isip, isip4, and isip6 to ${if conditions.
5828 40. The ACL variables $acl_xx are now saved with the message, and can be
5829 accessed later in routers, transports, and filters.
5831 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5832 strings in the file are not string-expanded.
5834 42. If a MAIL command specified a SIZE value that was too large to fit into an
5835 int variable, the check against message_size_limit failed. Such values are
5836 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5837 day this will have to be increased, but I don't think I want to be around
5838 when emails are that large.
5845 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5846 iconv() is not standard in FreeBSD.
5848 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5849 IPv6 enabled. The observed symptom was a segmentation fault on return from
5850 the function os_common_find_running_interfaces() in src/os.c.
5852 3. In the check_special_case() function in daemon.c I had used "errno" as an
5853 argument name, which causes warnings on some systems. This was basically a
5854 typo, since it was named "eno" in the comments!
5856 4. The code that waits for the clock to tick (at a resolution of some fraction
5857 of a second) so as to ensure message-id uniqueness was always waiting for
5858 at least one whole tick, when it could have waited for less. [This is
5859 almost certainly not relevant at current processor speeds, where it is
5860 unlikely to ever wait at all. But we try to future-proof.]
5862 5. The function that sleeps for a time interval that includes fractions of a
5863 second contained a race. It did not block SIGALRM between setting the
5864 timer, and suspending (a couple of lines later). If the interval was short
5865 and the sigsuspend() was delayed until after it had expired, the suspension
5866 never ended. On busy systems this could lead to processes getting stuck for
5869 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5870 process, before it forks any delivery processes. The open lookup caching
5871 mechanism meant that the open file or database connection was passed into
5872 the delivery process. The problem was that delivery processes always tidy
5873 up cached lookup data. This could cause a problem for the next delivery
5874 process started by the queue runner, because the external queue runner
5875 process does not know about the closure. So the next delivery process
5876 still has data in the lookup cache. In the case of a file lookup, there was
5877 no problem because closing a file descriptor in a subprocess doesn't affect
5878 the parent. However, if the lookup was caching a connection to a database,
5879 the connection was closed, and the second delivery process was likely to
5880 see errors such as "PGSQL: query failed: server closed the connection
5881 unexpectedly". The problem has been fixed by closing all cached lookups
5882 in a queue runner before running a delivery process.
5884 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5885 seem to have the "const" qualifier which it has on other OS. I've
5888 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5889 *of the same type* (client or server) with the same public name that an
5890 error should be diagnosed.
5892 9. When Exim looked up a host name for an IP address, but failed to find the
5893 original IP address when looking up the host name (a safety check), it
5894 output the message "<ip address> does not match any IP for NULL", which was
5895 confusing, to say the least. The bug was that the host name should have
5896 appeared instead of "NULL".
5898 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5899 user that is built into the binary, and the -C or -D options is used, root
5900 privilege is dropped before the configuration file is read. In addition,
5901 logging is switched to stderr instead of the normal log files. If the
5902 configuration then re-defines the Exim user, the unprivileged environment
5903 is probably not what is expected, so Exim logs a panic warning message (but
5906 However, if deliver_drop_privilege is set, the unprivileged state may well
5907 be exactly what is intended, so the warning has been cut out in that case,
5908 and Exim is allowed to try to write to its normal log files.
5914 1. smtp_return_error_details was not giving details for temporary sender
5915 or receiver verification errors.
5917 2. Diagnose a configuration error if two authenticators have the same public
5920 3. Exim used not to create the message log file for a message until the first
5921 delivery attempt. This could be confusing when incoming messages were held
5922 for policy or load reasons. The message log file is now created at the time
5923 the message is received, and an initial "Received" line is written to it.
5925 4. The automatically generated man page for command line options had a minor
5926 bug that caused no ill effects; however, a more serious problem was that
5927 the procedure for building the man page automatically didn't always
5928 operate. Consequently, release 4.20 contains an out-of-date version. This
5929 shouldn't happen again.
5931 5. When building Exim with embedded Perl support, the script that builds the
5932 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5933 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5935 6. The freeze_tell option was not being used for messages that were frozen on
5936 arrival, either by an ACL or by local_scan().
5938 7. Added the smtp_incomplete_transaction log selector.
5940 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5941 was accepting AUTH without a new EHLO.
5943 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5944 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5947 10. Logging of TCP/IP connections (when configured) now happens in the main
5948 daemon process instead of the child process, so that the TCP/IP connection
5949 count is more accurate (but it can never be perfect).
5951 11. The use of "drop" in a nested ACL was not being handled correctly in the
5952 outer ACL. Now, if condition failure induced by the nested "drop" causes
5953 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5954 or "require"), the connection is dropped.
5956 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5957 that yield "discard" can now be used with an "accept" or a "discard" verb,
5958 but an error is generated for any others (because I can't see a useful way
5959 to define what should happen).
5961 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5962 are now processed in the same way as lines in the Exim configuration file.
5963 In particular, continuation lines are supported.
5965 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5967 15. Added -ti meaning -t -i.
5969 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5970 domains, and warn by logging if others are found.
5972 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5973 returned from GIFCONF when Exim is trying to find the list of interfaces on
5974 a host. The code in os.c has been modified to copy each ifreq to an aligned
5975 structure in all cases.
5977 Also, in some cases, the returned ifreq's were being copied to a 'struct
5978 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5979 means the last couple of bytes of an IPv6 address could be chopped if the
5980 ifreq contained only a normal sockaddr (14 bytes storage).
5982 18. Named domain lists were not supported in the hosts_treat_as_local option.
5983 An entry such as +xxxx was not recognized, and was treated as a literal
5986 19. Ensure that header lines added by a DATA ACL are included in the reject log
5987 if the ACL subsequently rejects the message.
5989 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5990 MD5 (which is deprecated).
5992 21. When testing a filter file using -bf, Exim was writing a message when it
5993 took the sender from a "From " line in the message, but it was not doing so
5994 when it took $return_path from a Return-Path: header line. It now does.
5996 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5997 with a valid header line field name (a series of printing characters
5998 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6000 23. Changed "disc" in the source to "disk" to conform to the documentation and
6001 the book and for uniformity.
6003 24. Ignore Sendmail's -Ooption=value command line item.
6005 25. When execve() failed while trying to run a command in a pipe transport,
6006 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6007 could be confused with a return value of 69 from the command itself. This
6008 has been changed to 127, the value the shell returns if it is asked to run
6009 a non-existent command. The wording for the related log line suggests a
6010 non-existent command as the problem.
6012 26. If received_header_text expands to an empty string, do not add a Received:
6013 header line to the message. (Well, it adds a token one on the spool, but
6014 marks it "old" so that it doesn't get used or transmitted.)
6016 27. Installed eximstats 1.28 (addition of -nt option).
6018 28. There was no check for failure on the call to getsockname() in the daemon
6019 code. This can fail if there is a shortage of resources on the system, with
6020 ENOMEM, for example. A temporary error is now given on failure.
6022 29. Contrary to the C standard, it seems that in some environments, the
6023 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6024 program. Exim now does this explicitly; it affects the formatting of
6025 timestamps using strftime().
6027 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6028 complaints. I've now initialized all the variables, to avoid this.
6030 32. Header lines added by a system filter were not being "seen" during
6031 transport-time rewrites.
6033 33. The info_callback() function passed to OpenSSL is set up with type void
6034 (*)(SSL *, int, int), as described somewhere. However, when calling the
6035 function (actually a macro) that sets it up, the type void(*)() is
6036 expected. I've put in a cast to prevent warnings from picky compilers.
6038 34. If a DNS black list lookup found a CNAME record, but there were no A
6039 records associated with the domain it pointed at, Exim crashed.
6041 35. If a DNS black list lookup returned more than one A record, Exim ignored
6042 all but the first. It now scans all returned addresses if a particular IP
6043 value is being sought. In this situation, the contents of the
6044 $dnslist_value variable are a list of all the addresses, separated by a
6047 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6048 to accept a host name and all its aliases if the forward lookup for any of
6049 them yielded the IP address of the incoming connection. Now it accepts only
6050 those names whose forward lookup yields the correct IP address. Any other
6051 names are discarded. This closes a loophole whereby a rogue DNS
6052 administrator could create reverse DNS records to break through a
6053 wildcarded host restriction in an ACL.
6055 37. If a user filter or a system filter that ran in a subprocess used any of
6056 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6057 the wrong values were passed to the pipe command ($thisaddress had the
6058 value of $0, $0 had the value of $1, etc). This bug was introduced by
6059 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6061 38. Improved the line breaking for long SMTP error messages from ACLs.
6062 Previously, if there was no break point between 40 and 75 characters, Exim
6063 left the rest of the message alone. Two changes have been made: (a) I've
6064 reduced the minimum length to 35 characters; (b) if it can't find a break
6065 point between 35 and 75 characters, it looks ahead and uses the first one
6066 that it finds. This may give the occasional overlong line, but at least the
6067 remaining text gets split now.
6069 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6070 file descriptors might be low, and that setting 1000 would always raise it.
6071 It turns out that in some environments, the limit is already over 1000 and
6072 that lowering it causes trouble. So now Exim takes care not to decrease it.
6074 40. When delivering a message, the value of $return_path is set to $sender_
6075 address at the start of routing (routers may change the value). By an
6076 oversight, this default was not being set up when an address was tested by
6077 -bt or -bv, which affected the outcome if any router or filter referred to
6080 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6081 log only when "message" or "log_message" are set. However, if one of the
6082 conditions was an address verification, or a call to a nested ACL, the
6083 messages generated by the underlying test were being passed through. This
6084 no longer happens. The underlying message is available in $acl_verify_
6085 message for both "message" and "log_message" expansions, so it can be
6086 passed through if needed.
6088 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6089 new expansion $bh_ to give the encoded byte string without charset
6090 translation. Translation happens only if iconv() is available; HAVE_ICONV
6091 indicates this at build time. HEADERS_CHARSET gives the charset to
6092 translate to; headers_charset can change it in the configuration, and
6093 "headers charset" can change it in an individual filter file.
6095 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6096 that creates RFC 2047 encoded "words" labels them as that charset instead
6097 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6098 expansion operator; (ii) when Exim creates a From: line for a local
6099 message; (iii) when a header line is rewritten to include a "phrase" part.
6101 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6102 buggy, causing it to skip the first lines of messages whose message ID
6103 ended in 'D'. This would not have bitten before Exim release 4.14, because
6104 message IDs were unlikely to end in 'D' before then. The effect was to have
6105 incorrect size information for certain domains.
6107 45. #include "config.h" was missing at the start of the crypt16.c module. This
6108 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6111 46. If there was a timeout during a "random" callout check, Exim treated it as
6112 a failure of the random address, and carried on sending RSET and the real
6113 address. If the delay was just some slowness somewhere, the response to the
6114 original RCPT would be taken as a response to RSET and so on, causing
6115 mayhem of various kinds.
6117 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6118 when I implemented it. It didn't allow for the fact that some option values
6119 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6120 the right test for positive values.
6122 48. Domain names in DNS records are case-independent. Exim always looks them up
6123 in lower case. Some resolvers return domain names in exactly the case they
6124 appear in the zone file, that is, they may contain uppercase letters. Not
6125 all resolvers do this - some return always lower case. Exim was treating a
6126 change of case by a resolver as a change of domain, similar to a widening
6127 of a domain abbreviation. This triggered its re-routing code and so it was
6128 trying to route what was effectively the same domain again. This normally
6129 caused routing to fail (because the router wouldn't handle the domain
6130 twice). Now Exim checks for this case specially, and just changes the
6131 casing of the domain that it ultimately uses when it transmits the message
6134 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6137 50. If a filter generated a file delivery with a non-absolute name (possible if
6138 no home directory exists for the router), the forbid_file option was not
6141 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6142 the existing equality matching.
6144 52. Exim was using ints instead of ino_t variables in some places where it was
6145 dealing with inode numbers.
6147 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6148 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6149 called TMPDIR, and if it finds it is different, it changes its value.
6151 54. The smtp_printf() function is now made available to local_scan() so
6152 additional output lines can be written before returning. There is also an
6153 smtp_fflush() function to enable the detection of a dropped connection.
6154 The variables smtp_input and smtp_batched_input are exported to
6157 55. Changed the default runtime configuration: the message "Unknown user"
6158 has been removed from the ACL, and instead placed on the localuser router,
6159 using the cannot_route_message feature. This means that any verification
6160 failures that generate their own messages won't get overridden. Similarly,
6161 the "Unrouteable address" message that was in the ACL for unverifiable
6162 relay addresses has also been removed.
6164 56. Added hosts_avoid_esmtp to the smtp transport.
6166 57. The exicyclog script was not checking for the esoteric option
6167 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6168 will work only if exicyclog is run under the appropriate euid.
6170 58. Following a discussion on the list, the rules by which Exim recognises line
6171 endings on incoming messages have been changed. The -dropcr and drop_cr
6172 options are now no-ops, retained only for backwards compatibility. The
6173 following line terminators are recognized: LF CRLF CR. However, special
6174 processing applies to CR:
6176 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6177 nor a local message in the state where . is a terminator.
6179 (ii) If a bare CR is encountered in a header line, an extra space is added
6180 after the line terminator so as not to end the header. The reasoning
6181 behind this is that bare CRs in header lines are most likely either
6182 to be mistakes, or people trying to play silly games.
6184 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6185 was being incorrectly given as 18 bytes larger than it should have been.
6186 This is a VOB (very old bug).
6188 60. This may never have affected anything current, but just in case it has:
6189 When the local host is found other than at the start of a list of hosts,
6190 the local host, those with the same MX, and any that follow, are discarded.
6191 When the list in question was part of a longer list of hosts, the following
6192 hosts (not currently being processed) were also being discarded. This no
6193 longer happens. I'm not sure if this situation could ever has previously
6196 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6199 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6200 header. This is something that is recommended in a new Internet Draft, and
6201 is something that is documented as being done by Sendmail. There are two
6202 possible values. For messages generated by the autoreply transport, Exim
6205 Auto-Submitted: auto-replied
6207 whereas for all other generated messages (e.g. bounces) it adds
6209 Auto-Submitted: auto-generated
6211 63. The "personal" condition in filters now includes a test for the
6212 Auto-Submitted: header. If it contains the string "auto-" the message it
6213 not considered personal.
6215 64. Added rcpt_include_affixes as a generic transport option.
6217 65. Added queue_only_override (default true).
6219 66. Added the syslog_duplication option.
6221 67. If what should have been the first header line of a message consisted of
6222 a space followed by a colon, Exim was mis-interpreting it as a header line.
6223 It isn't of course - it is syntactically invalid and should therefore be
6224 treated as the start of the message body. The misbehaviour could have
6225 caused a number of strange effects, including loss of data in subsequent
6226 header lines, and spool format errors.
6228 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6229 client host had authenticated. This control can now be exercised by an ACL
6230 for more flexibility.
6232 69. By default, callouts do not happen when testing with -bh. There is now a
6233 variant, -bhc, which does actually run the callout code, including
6234 consulting and updating the callout cache.
6236 70. Added support for saslauthd authentication, courtesy of Alexander
6239 71. If statvfs() failed on the spool or log directories while checking their
6240 size for availability, Exim confusingly gave the error "space shortage".
6241 Furthermore, in debugging mode it crashed with a floating point exception.
6242 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6243 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6244 serious problem, Exim now writes to the main and panic logs when this
6245 happens, with details of the failure. It then refuses to accept the
6246 incoming message, giving the message "spool directory problem" or "log
6247 directory problem" with a 421 code for SMTP messages.
6249 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6250 0, 1, and 2 exist, because some OS complain for execs without them (see
6251 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6252 descriptors. However, the code omitted to check that the open succeeded,
6253 causing mysterious errors if for some reason the permissions on /dev/null
6254 got screwed. Now Exim writes a message to the main and panic logs, and
6255 bombs out if it can't open /dev/null.
6257 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6258 interact so that it is all more flexible. It is supposed to remain
6259 backwards compatible. Also added extra_local_interfaces.
6261 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6262 to bomb out with an assertion failure - to the client this appears as a
6263 connection drop. This problem occurs in the part of the code that was taken
6264 from the Samba project. Fortunately, the assertion is in a very simple
6265 function, so I have fixed this by reproducing the function inline in the
6266 one place where it is called, and arranging for authentication to fail
6267 instead of killing the process with assert().
6269 75. The SPA client code was not working when the server requested OEM rather
6270 than Unicode encoding.
6272 76. Added code to make require_files with a specific uid setting more usable in
6273 the case where statting the file as root fails - usually a non-root-mounted
6274 NFS file system. When this happens and the failure is EACCES, Exim now
6275 forks a subprocess and does the per-uid checking as the relevant uid.
6277 77. Added process_log_path.
6279 78. If log_file_path was not explicitly set, a setting of check_log_space or
6280 check_log_inodes was ignored.
6282 79. If a space check for the spool or log partitions fails, the incident is now
6283 logged. Of course, in the latter case the data may get lost...
6285 80. Added the %p formatting code to string_format() so that it can be used to
6286 print addresses in debug_print(). Adjusted all the address printing in the
6287 debugging in store.c to use %p rather than %d.
6289 81. There was a concern that a line of code in smtp_in.c could overflow a
6290 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6291 initially expressed, the concern was not well-founded, because trailing
6292 spaces are removed early. However, if the trailing spaces were followed by
6293 a NULL, they did not get removed, so the overflow was possible. Two fixes
6296 (a) I re-wrote the offending code in a cleaner fashion.
6297 (b) If an incoming SMTP command contains a NULL character, it is rejected
6300 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6301 runs initgroups(), so that if the Exim user is in any additional groups,
6302 they will be used during message reception.
6308 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob