1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostnames
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way. Bug 68.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and priority
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
61 and remote servers. Remote servers can be IPv6 addresses, and
64 JH/17 Bug 68: The spamd_address main option now supports an optional
65 timeout value per server.
67 JH/18 Bug 1581: Router and transport options headers_add/remove can
68 now have the list separator specified.
70 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
73 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
76 JH/21 Support for the A6 type of dns record is withdrawn.
78 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
79 rather than the verbs used.
81 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
82 from 255 to 1024 chars.
84 JH/24 Verification callouts now attempt to use TLS by default.
86 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
87 are generic router options now. The defaults didn't change.
89 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
90 Original patch from Alexander Shikoff, worked over by JH.
92 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
93 files if ZCAT_COMMAND is not executable.
95 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
97 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
99 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
100 Normally benign, it bites when the pair was led to by a CNAME;
101 modern usage is to not canoicalize the domain to a CNAME target
102 (and we were inconsistent anyway for A-only vs AAAA+A).
104 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
106 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
107 when evaluating $sender_host_dnssec.
109 JH/31 Check the HELO verification lookup for DNSSEC, adding new
110 $sender_helo_dnssec variable.
112 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
114 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
116 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
121 TL/01 When running the test suite, the README says that variables such as
122 no_msglog_check are global and can be placed anywhere in a specific
123 test's script, however it was observed that placement needed to be near
124 the beginning for it to behave that way. Changed the runtest perl
125 script to read through the entire script once to detect and set these
126 variables, reset to the beginning of the script, and then run through
127 the script parsing/test process like normal.
129 TL/02 The BSD's have an arc4random API. One of the functions to induce
130 adding randomness was arc4random_stir(), but it has been removed in
131 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
132 function when detected.
134 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
135 cause callback expansion.
137 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
138 syntax errors in an expansion can be treated as a string instead of
139 logging or causing an error, due to the internal use of bool_lax
140 instead of bool when processing it.
142 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
143 server certificates when making smtp deliveries.
145 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
147 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
149 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
151 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
152 Merged patch from Sebastian Wiedenroth.
154 JH/05 Fix results-pipe from transport process. Several recipients, combined
155 with certificate use, exposed issues where response data items split
156 over buffer boundaries were not parsed properly. This eventually
157 resulted in duplicates being sent. This issue only became common enough
158 to notice due to the introduction of conection certificate information,
159 the item size being so much larger. Found and fixed by Wolfgang Breyha.
161 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
162 size buffer was used, resulting in syntax errors when an expansion
165 JH/07 Add support for directories of certificates when compiled with a GnuTLS
166 version 3.3.6 or later.
168 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
169 is EXPERIMENTAL_EVENT, the main-configuration and transport options
170 both become "event_action", the variables become $event_name, $event_data
171 and $event_defer_errno. There is a new variable $verify_mode, usable in
172 routers, transports and related events. The tls:cert event is now also
173 raised for inbound connections, if the main configuration event_action
176 TL/06 In test suite, disable OCSP for old versions of openssl which contained
177 early OCSP support, but no stapling (appears to be less than 1.0.0).
179 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
180 server certificate names available under the smtp transport option
181 "tls_verify_cert_hostname" now do not permit multi-component wildcard
184 JH/10 Time-related extraction expansions from certificates now use the main
185 option "timezone" setting for output formatting, and are consistent
186 between OpenSSL and GnuTLS compilations. Bug 1541.
188 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
189 encoded parameter in the incoming message. Bug 1558.
191 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
192 include certificate info, eximon was claiming there were spoolfile
195 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
197 JH/14 Log delivery-related information more consistently, using the sequence
198 "H=<name> [<ip>]" wherever possible.
200 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
201 are problematic for Debian distribution, omit them from the release
204 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
206 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
209 JH/17 Fix a null-indirection in certextract expansions when a nondefault
210 output list separator was used.
215 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
216 checkers that were complaining about end of non-void function with no
219 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
220 This was a regression intruduced in 4.83 by another bugfix.
222 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
224 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
225 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
231 TF/01 Correctly close the server side of TLS when forking for delivery.
233 When a message was received over SMTP with TLS, Exim failed to clear up
234 the incoming connection properly after forking off the child process to
235 deliver the message. In some situations the subsequent outgoing
236 delivery connection happened to have the same fd number as the incoming
237 connection previously had. Exim would try to use TLS and fail, logging
238 a "Bad file descriptor" error.
240 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
241 utilities have not been installed.
243 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
244 temporary space as the ACL may create new global variables.
246 TL/01 LDAP support uses per connection or global context settings, depending
247 upon the detected version of the libraries at build time.
249 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
250 to extract and use the src ip:port in logging and expansions as if it
251 were a direct connection from the outside internet. PPv2 support was
252 updated based on HAProxy spec change in May 2014.
254 JH/02 Add ${listextract {number}{list}{success}{fail}}.
256 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
257 Properly escape header and check for NULL return.
259 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
262 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
264 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
265 characters in header names, implemented as a verify condition.
266 Contributed by Michael Fischer v. Mollard.
268 TL/05 Rename SPF condition results err_perm and err_temp to standardized
269 results permerror and temperror. Previous values are deprecated but
270 still accepted. In a future release, err_perm and err_temp will be
271 completely removed, which will be a backward incompatibility if the
272 ACL tests for either of these two old results. Patch contributed by
273 user bes-internal on the mailing list.
275 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
277 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
278 selectors, in both main and reject logs.
280 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
283 JH/07 Add malware type "sock" for talking to simple daemon.
285 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
287 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
288 routers/transports under cutthrough routing.
290 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
291 numbers. Touch up "bool" conditional to keep the same definition.
293 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
295 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
297 JH/12 Expand items in router/transport headers_add or headers_remove lists
298 individually rather than the list as a whole. Bug 1452.
300 Required for reasonable handling of multiple headers_ options when
301 they may be empty; requires that headers_remove items with embedded
302 colons must have them doubled (or the list-separator changed).
304 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
305 view the policy declared in the DMARC record. Currently, $dmarc_status
306 is a combined value of both the record presence and the result of the
309 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
311 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
312 dnslookup router and the smtp transport (applying to the forward
315 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
316 of ldap servers used for a specific lookup. Patch provided by Heiko
319 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
320 New variable $lookup_dnssec_authenticated for observability.
322 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
323 Patch submitted by Lars Timman.
325 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
327 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
328 Requires trusted mode and valid format message id, aborts otherwise.
329 Patch contributed by Heiko Schlichting.
331 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
332 certextract with support for various fields. Bug 1358.
334 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
335 is requested by default, modifiable by smtp transport option
338 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
339 operate on certificate variables to give certificate fingerprints
340 Also new ${sha256:cert_variable}.
342 JH/23 The PRDR feature is moved from being Experimental into the mainline.
344 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
345 Christian Aistleitner.
347 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
349 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
350 file. Patch from Wolfgang Breyha.
352 JH/25 Expand the coverage of the delivery $host and $host_address to
353 client authenticators run in verify callout. Bug 1476.
355 JH/26 Port service names are now accepted for tls_on_connect_ports, to
356 align with daemon_smtp_ports. Bug 72.
358 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
359 support and error reporting did not work properly.
361 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
362 and is readable. Patch from Andrew Colin Kissa.
364 TL/14 Enhance documentation of ${run expansion and how it parses the
365 commandline after expansion, particularly in the case when an
366 unquoted variable expansion results in an empty value.
368 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
370 PP/02 Fix internal collision of T_APL on systems which support RFC3123
371 by renaming away from it. Addresses GH issue 15, reported by
374 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
376 TL/15 SECURITY: prevent double expansion in math comparison functions
377 (can expand unsanitized data). Not remotely exploitable.
384 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
386 PP/02 Make -n do something, by making it not do something.
387 When combined with -bP, the name of an option is not output.
389 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
392 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
393 $sender_host_name and config options to manage this, and basic check
396 PP/05 DSCP support for outbound connections and control modifier for inbound.
398 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
399 (Only plugin which currently uses this is kerberos4, which nobody should
400 be using, but we should make it available and other future plugins might
401 conceivably use it, even though it would break NAT; stuff *should* be
402 using channel bindings instead).
404 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
405 name; added for Sendmail compatibility; requires admin caller.
406 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
407 just ignore it); requires trusted caller.
408 Also parse but ignore: -Ac -Am -X<logfile>
411 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
413 TL/02 Add +smtp_confirmation as a default logging option.
415 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
416 Patch by Magnus Holmgren from 2007-02-20.
418 TL/04 Bugzilla 1281 - Spec typo.
419 Bugzilla 1283 - Spec typo.
420 Bugzilla 1290 - Spec grammar fixes.
422 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
424 TL/06 Add Experimental DMARC support using libopendmarc libraries.
426 TL/07 Fix an out of order global option causing a segfault. Reported to dev
427 mailing list by by Dmitry Isaikin.
429 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
431 JH/02 Support "G" suffix to numbers in ${if comparisons.
433 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
435 NM/01 Bugzilla 1197 - Spec typo
436 Bugzilla 1196 - Spec examples corrections
438 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
440 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
441 gnutls_enable_pkcs11, but renamed to more accurately indicate its
444 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
445 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
447 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
448 "acl {{name}{arg}...}", and optional args on acl condition
451 JH/05 Permit multiple router/transport headers_add/remove lines.
453 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
455 JH/07 Avoid using a waiting database for a single-message-only transport.
456 Performance patch from Paul Fisher. Bugzilla 1262.
458 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
461 JH/09 Add $headers_added variable, with content from use of ACL modifier
462 add_header (but not yet added to the message). Bugzilla 199.
464 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
465 Pulled from Bugzilla 817 by Wolfgang Breyha.
467 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
469 (nb: this is the same fix as in Exim 4.80.1)
471 JH/11 Add A= logging on delivery lines, and a client_set_id option on
474 JH/12 Add optional authenticated_sender logging to A= and a log_selector
477 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
479 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
480 advertise SMTP AUTH mechanism to us, instead of a generic
481 protocol violation error. Also, make Exim more robust to bad
482 data from the Dovecot auth socket.
484 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
486 When a queue runner is handling a message, Exim first routes the
487 recipient addresses, during which it prunes them based on the retry
488 hints database. After that it attempts to deliver the message to
489 any remaining recipients. It then updates the hints database using
492 So if a recipient address works intermittently, it can get repeatedly
493 deferred at routing time. The retry hints record remains fresh so the
494 address never reaches the final cutoff time.
496 This is a fairly common occurrence when a user is bumping up against
497 their storage quota. Exim had some logic in its local delivery code
498 to deal with this. However it did not apply to per-recipient defers
499 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
501 This change adds a proper retry rule check during routing so that the
502 final cutoff time is checked against the message's age. We only do
503 this check if there is an address retry record and there is not a
504 domain retry record; this implies that previous attempts to handle
505 the address had the retry_use_local_parts option turned on. We use
506 this as an approximation for the destination being like a local
507 delivery, as in LMTP.
509 I suspect this new check makes the old local delivery cutoff check
510 redundant, but I have not verified this so I left the code in place.
512 TF/02 Correct gecos expansion when From: is a prefix of the username.
514 Test 0254 submits a message to Exim with the header
518 When I ran the test suite under the user fanf2, Exim expanded
519 the header to contain my full name, whereas it should have added
520 a Resent-Sender: header. It erroneously treats any prefix of the
521 username as equal to the username.
523 This change corrects that bug.
525 GF/01 DCC debug and logging tidyup
526 Error conditions log to paniclog rather than rejectlog.
527 Debug lines prefixed by "DCC: " to remove any ambiguity.
529 TF/03 Avoid unnecessary rebuilds of lookup-related code.
531 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
532 Bug spotted by Jeremy Harris; was flawed since initial commit.
533 Would have resulted in OCSP responses post-SNI triggering an Exim
534 NULL dereference and crash.
536 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
538 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
539 Bug detection, analysis and fix by Samuel Thibault.
540 Bugzilla 1331, Debian bug #698092.
542 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
544 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
545 Server implementation by Todd Lyons, client by JH.
546 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
547 config variable "prdr_enable" controls whether the server
548 advertises the facility. If the client requests PRDR a new
549 acl_data_smtp_prdr ACL is called once for each recipient, after
550 the body content is received and before the acl_smtp_data ACL.
551 The client is controlled by bolth of: a hosts_try_prdr option
552 on the smtp transport, and the server advertisement.
553 Default client logging of deliveries and rejections involving
554 PRDR are flagged with the string "PRDR".
556 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
557 fclose(). Diagnosis by Todd Lyons.
559 PP/17 Update configure.default to handle IPv6 localhost better.
560 Patch by Alain Williams (plus minor tweaks).
563 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
564 This is now consistent with GnuTLS, and is now documented: the
565 previous undocumented portable approach to treating the option as
566 unset was to force an expansion failure. That still works, and
567 an empty string is now equivalent.
569 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
570 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
571 not performing validation itself.
573 PP/20 Added force_command boolean option to pipe transport.
574 Patch from Nick Koston, of cPanel Inc.
576 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
579 TF/04 Added udpsend ACL modifer and hexquote expansion operator
581 PP/21 Fix eximon continuous updating with timestamped log-files.
582 Broken in a format-string cleanup in 4.80, missed when I repaired the
583 other false fix of the same issue.
584 Report and fix from Heiko Schlichting.
587 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
588 Report from Prashanth Katuri.
590 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
591 It's SecureTransport, so affects any MacOS clients which use the
592 system-integrated TLS libraries, including email clients.
594 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
595 using a MIME ACL for non-SMTP local injection.
596 Report and assistance in diagnosis by Warren Baker.
598 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
600 JH/16 Fix comparisons for 64b. Bugzilla 1385.
602 TL/09 Add expansion variable $authenticated_fail_id to keep track of
603 last id that failed so it may be referenced in subsequent ACL's.
605 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
608 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
609 ldap library initialization, allowing self-signed CA's to be
610 used. Also properly sets require_cert option later in code by
611 using NULL (global ldap config) instead of ldap handle (per
612 session). Bug diagnosis and testing by alxgomz.
614 TL/12 Enhanced documentation in the ratelimit.pl script provided in
615 the src/util/ subdirectory.
617 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
618 renamed to Transport Post Delivery Action by Jeremy Harris, as
621 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
622 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
623 redis_servers = needs to be configured which will be used by the redis
624 lookup. Patch from Warren Baker, of The Packet Hub.
626 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
628 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
629 hostname or reverse DNS when processing a host list. Used suggestions
630 from multiple comments on this bug.
632 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
634 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
635 Missed a few lines, added it to make the runtest require no keyboard
638 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
639 contains upper case chars. Make router use caseful_local_part.
641 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
642 support when GnuTLS has been built with p11-kit.
648 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
650 This, or similar/improved, will also be change PP/11 of 4.82.
656 PP/01 Handle short writes when writing local log-files.
657 In practice, only affects FreeBSD (8 onwards).
658 Bugzilla 1053, with thanks to Dmitry Isaikin.
660 NM/01 Bugzilla 949 - Documentation tweak
662 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
665 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
667 PP/02 Implemented gsasl authenticator.
669 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
671 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
672 `pkg-config foo` for cflags/libs.
674 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
675 with rest of GSASL and with heimdal_gssapi.
677 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
678 `pkg-config foo` for cflags/libs for the TLS implementation.
680 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
681 properties get this fed in as external SSF. A number of robustness
682 and debugging improvements to the cyrus_sasl authenticator.
684 PP/08 cyrus_sasl server now expands the server_realm option.
686 PP/09 Bugzilla 1214 - Log authentication information in reject log.
687 Patch by Jeremy Harris.
689 PP/10 Added dbmjz lookup type.
691 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
693 PP/12 MAIL args handles TAB as well as SP, for better interop with
694 non-compliant senders.
695 Analysis and variant patch by Todd Lyons.
697 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
698 Bug report from Lars Müller <lars@samba.org> (via SUSE),
699 Patch from Dirk Mueller <dmueller@suse.com>
701 PP/13 tls_peerdn now print-escaped for spool files.
702 Observed some $tls_peerdn in wild which contained \n, which resulted
703 in spool file corruption.
705 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
706 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
707 or write after TLS renegotiation, which otherwise led to messages
710 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
711 as a tracking header (ie: a signed header comes before the signature).
712 Patch from Wolfgang Breyha.
714 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
715 comma-sep list; embedded commas doubled.
717 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
719 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
721 Report and patch from Dmitry Banschikov.
723 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
724 Removed SSL_clear() after SSL_new() which led to protocol negotiation
725 failures. We appear to now support TLS1.1+ with Exim.
727 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
728 lets Exim select keys and certificates based upon TLS SNI from client.
729 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
730 before an outbound SMTP session. New log_selector, +tls_sni.
732 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
733 NULL dereference. Report and patch from Alun Jones.
735 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
736 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
737 Not seeing resolver debug output on NetBSD, but suspect this is a
738 resolver implementation change.
740 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
741 Left warnings. Added "eximon gdb" invocation mode.
743 PP/21 Defaulting "accept_8bitmime" to true, not false.
745 PP/22 Added -bw for inetd wait mode support.
747 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
748 locate the relevant includes and libraries. Made this the default.
750 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
751 Bugzilla 1246, report and most of solution from Tomasz Kusy.
753 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
754 This may cause build issues on older platforms.
756 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
757 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
758 gnutls_require_mac & gnutls_require_protocols (no longer supported).
759 Added SNI support via GnuTLS too.
760 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
762 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
764 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
765 Applied second patch from Janne, implementing suggestion to default
766 multiple-strings-in-record handling to match SPF spec.
768 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
770 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
771 read-only, out of scope).
772 Patch from Wolfgang Breyha, report from Stuart Northfield.
774 PP/29 Fix three issues highlighted by clang analyser static analysis.
775 Only crash-plausible issue would require the Cambridge-specific
776 iplookup router and a misconfiguration.
777 Report from Marcin Mirosław.
779 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
781 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
782 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
783 As part of this, removing so much warning spew let me fix some minor
784 real issues in debug logging.
786 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
787 assignment on my part. Fixed.
789 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
790 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
791 Janne Snabb (who went above and beyond: thank you).
793 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
794 string otherwise requires a connection and a bunch more work and it's
795 relatively easy to get wrong. Should also expose TLS library linkage
798 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
799 64-bit ${eval} (JH/03).
801 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
802 GNU libc to support some of the 64-bit stuff, should not lead to
803 conflicts. Defined before os.h is pulled in, so if a given platform
804 needs to override this, it can.
806 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
807 protection layer was required, which is not implemented.
808 Bugzilla 1254, patch from Wolfgang Breyha.
810 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
811 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
812 tls_dhparam take prime identifiers. Also unbreak combination of
813 OpenSSL+DH_params+TLSSNI.
815 PP/39 Disable SSLv2 by default in OpenSSL support.
821 PP/01 Solaris build fix for Oracle's LDAP libraries.
822 Bugzilla 1109, patch from Stephen Usher.
824 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
826 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
829 TF/02 Fix a couple more cases where we did not log the error message
830 when unlink() failed. See also change 4.74-TF/03.
832 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
833 lock up or crash if it happened to be inside a call to libc when it
834 got a SIGUSR1 from exiwhat.
836 The SIGUSR1 handler appends the current process status to the process
837 log which is later printed by exiwhat. It used to use the general
838 purpose logging code to do this, but several functions it calls are
839 not safe for signals.
841 The new output code in the SIGUSR1 handler is specific to the process
842 log, and simple enough that it's easy to inspect for signal safety.
843 Removing some special cases also simplifies the general logging code.
844 Removing the spurious timestamps from the process log simplifies
847 TF/04 Improved ratelimit ACL condition.
849 The /noupdate option has been deprecated in favour of /readonly which
850 has clearer semantics. The /leaky, /strict, and /readonly update modes
851 are mutually exclusive. The update mode is no longer included in the
852 database key; it just determines when the database is updated. (This
853 means that when you upgrde Exim will forget old rate measurements.)
855 Exim now checks that the per_* options are used with an update mode that
856 makes sense for the current ACL. For example, when Exim is processing a
857 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
858 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
859 must specify per_mail/readonly. If you omit the update mode it defaults to
860 /leaky where that makes sense (as before) or /readonly where required.
862 The /noupdate option is now undocumented but still supported for
863 backwards compatibility. It is equivalent to /readonly except that in
864 ACLs where /readonly is required you may specify /leaky/noupdate or
865 /strict/noupdate which are treated the same as /readonly.
867 A useful new feature is the /count= option. This is a generalization
868 of the per_byte option, so that you can measure the throughput of other
869 aggregate values. For example, the per_byte option is now equivalent
870 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
872 The per_rcpt option has been generalized using the /count= mechanism
873 (though it's more complicated than the per_byte equivalence). When it is
874 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
875 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
876 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
877 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
878 Note that using per_rcpt with a non-readonly update mode in more than
879 one ACL will cause the recipients to be double-counted. (The per_mail
880 and per_byte options don't have this problem.)
882 The handling of very low rates has changed slightly. If the computed rate
883 is less than the event's count (usually one) then this event is the first
884 after a long gap. In this case the rate is set to the same as this event's
885 count, so that the first message of a spam run is counted properly.
887 The major new feature is a mechanism for counting the rate of unique
888 events. The new per_addr option counts the number of different
889 recipients that someone has sent messages to in the last time period. It
890 behaves like per_rcpt if all the recipient addresses are different, but
891 duplicate recipient addresses do not increase the measured rate. Like
892 the /count= option this is a general mechanism, so the per_addr option
893 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
894 example, measure the rate that a client uses different sender addresses
895 with the options per_mail/unique=$sender_address. There are further
896 details in the main documentation.
898 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
900 TF/06 Removed a few PCRE remnants.
902 TF/07 Automatically extract Exim's version number from tags in the git
903 repository when doing development or release builds.
905 PP/02 Raise smtp_cmd_buffer_size to 16kB.
906 Bugzilla 879. Patch from Paul Fisher.
908 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
909 Heavily based on revision 40f9a89a from Simon Arlott's tree.
912 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
914 PP/05 Variable $av_failed, true if the AV scanner deferred.
915 Bugzilla 1078. Patch from John Horne.
917 PP/06 Stop make process more reliably on build failure.
918 Bugzilla 1087. Patch from Heiko Schlittermann.
920 PP/07 Make maildir_use_size_file an _expandable_ boolean.
921 Bugzilla 1089. Patch from Heiko Schlittermann.
923 PP/08 Handle ${run} returning more data than OS pipe buffer size.
924 Bugzilla 1131. Patch from Holger Weiß.
926 PP/09 Handle IPv6 addresses with SPF.
927 Bugzilla 860. Patch from Wolfgang Breyha.
929 PP/10 GnuTLS: support TLS 1.2 & 1.1.
931 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
934 PP/11 match_* no longer expand right-hand-side by default.
935 New compile-time build option, EXPAND_LISTMATCH_RHS.
936 New expansion conditions, "inlist", "inlisti".
938 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
940 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
942 PP/14 fix log_write() format string regression from TF/03.
943 Bugzilla 1152. Patch from Dmitry Isaikin.
949 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
951 PP/02 Harmonised TLS library version reporting; only show if debugging.
952 Layout now matches that introduced for other libraries in 4.74 PP/03.
954 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
956 PP/04 New "dns_use_edns0" global option.
958 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
961 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
962 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
964 TK/01 Updated PolarSSL code to 0.14.2.
965 Bugzilla 1097. Patch from Andreas Metzler.
967 PP/07 Catch divide-by-zero in ${eval:...}.
970 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
973 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
974 format-string attack -- SECURITY: remote arbitrary code execution.
976 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
977 time unintentionally subject to list matching rules, letting the header
978 cause arbitrary Exim lookups (of items which can occur in lists, *not*
979 arbitrary string expansion). This allowed for information disclosure.
981 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
982 INT_MIN/-1 -- value coerced to INT_MAX.
988 NM/01 Workround for PCRE version dependency in version reporting
991 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
992 This fixes portability to compilers other than gcc, notably
993 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
995 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
996 makefiles for portability to HP-UX and POSIX correctness.
998 PP/01 Permit LOOKUP_foo enabling on the make command-line.
999 Also via indented variable definition in the Makefile.
1000 (Debugging by Oliver Heesakkers).
1002 PP/02 Restore caching of spamd results with expanded spamd_address.
1003 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1005 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1006 Improves build reliability. Fix from: Frank Elsner
1008 NM/02 Fix wide character breakage in the rfc2047 coding
1009 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1011 NM/03 Allow underscore in dnslist lookups
1012 Fixes bug 1026. Patch from Graeme Fowler
1014 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1015 Code patches from Adam Ciarcinski of NetBSD.
1017 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1020 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1021 is logged, to avoid truncation. Patch from John Horne.
1023 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1024 Patch from Jakob Hirsch.
1026 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1027 SQL string expansion failure details.
1028 Patch from Andrey Oktyabrski.
1030 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1031 Patch from Simon Arlott.
1033 PP/09 New lookups functionality failed to compile on old gcc which rejects
1034 extern declarations in function scope.
1035 Patch from Oliver Fleischmann
1037 PP/10 Use sig_atomic_t for flags set from signal handlers.
1038 Check getgroups() return and improve debugging.
1039 Fixed developed for diagnosis in bug 927 (which turned out to be
1042 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1043 Patch from Mark Zealey.
1045 PP/12 Bugzilla 1056: Improved spamd server selection.
1046 Patch from Mark Zealey.
1048 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1049 Based on patch from Heiko Schlittermann.
1051 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1052 Patch from Uwe Doering, sign-off by Michael Haardt.
1054 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1055 variable declaration deep within a block. Bug and patch from
1058 PP/15 lookups-Makefile IRIX compatibilty coercion.
1060 PP/16 Make DISABLE_DKIM build knob functional.
1062 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1063 Patch by Simon Arlott
1065 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1066 variable argument macros. Our copy now differs from upstream.
1072 TF/01 Failure to get a lock on a hints database can have serious
1073 consequences so log it to the panic log.
1075 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1076 controlled using the smtp_confirmation log selector.
1078 TF/03 Include the error message when we fail to unlink a spool file.
1080 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1081 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1082 for maintaining out-of-tree patches for some time.
1084 PP/01 Bugzilla 139: Documentation and portability issues.
1085 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1086 Handle per-OS dynamic-module compilation flags.
1088 PP/02 Let /dev/null have normal permissions.
1089 The 4.73 fixes were a little too stringent and complained about the
1090 permissions on /dev/null. Exempt it from some checks.
1091 Reported by Andreas M. Kirchwitz.
1093 PP/03 Report version information for many libraries, including
1094 Exim version information for dynamically loaded libraries. Created
1095 version.h, now support a version extension string for distributors
1096 who patch heavily. Dynamic module ABI change.
1098 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1099 privilege escalation vulnerability whereby the Exim run-time user
1100 can cause root to append content of the attacker's choosing to
1103 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1106 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1107 If dropping privileges for untrusted macros, we disabled normal logging
1108 on the basis that it would fail; for the Exim run-time user, this is not
1109 the case, and it resulted in successful deliveries going unlogged.
1110 Fixed. Reported by Andreas Metzler.
1116 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1117 only prepend for the Resent-* case. Fixes regression introduced in
1118 Exim 4.70 by NM/22 for Bugzilla 607.
1120 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1121 increasing numbers of administrators be bitten by this.
1123 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1125 PP/03 Bugzilla 994: added openssl_options main configuration option.
1127 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1129 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1131 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1133 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1134 without a peer certificate, leading to a segfault because of an
1135 assumption that peers always have certificates. Be a little more
1136 paranoid. Problem reported by Martin Tscholak.
1138 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1139 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1140 NB: ClamAV planning to remove STREAM in "middle of 2010".
1141 CL also introduces -bmalware, various -d+acl logging additions and
1142 more caution in buffer sizes.
1144 PP/09 Implemented reverse_ip expansion operator.
1146 PP/10 Bugzilla 937: provide a "debug" ACL control.
1148 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1150 PP/12 Bugzilla 973: Implement --version.
1152 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1154 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1156 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1158 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1159 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1160 ignore trailing whitespace.
1162 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1164 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1167 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1168 Notification from Dr Andrew Aitchison.
1170 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1171 ExtendedDetectionInfo response format.
1172 Notification from John Horne.
1174 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1177 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1178 XSL and documented dependency on system catalogs, with examples of how
1181 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1184 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1185 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1186 configuration file which is writeable by the Exim user or group.
1188 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1189 of configuration files to cover files specified with the -C option if
1190 they are going to be used with root privileges, not just the default
1193 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1194 option (effectively making it always true).
1196 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1197 files to be used while preserving root privileges.
1199 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1200 that rogue child processes cannot use them.
1202 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1203 run-time user, instead of root.
1205 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1206 Exim run-time user without dropping privileges.
1208 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1209 result string, instead of calling string_vformat() twice with the same
1212 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1213 for other users. Others should always drop root privileges if they use
1214 -C on the command line, even for a whitelisted configure file.
1216 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1218 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1224 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1225 $data_path, and $header_path variables; fixed documentation bugs and
1228 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1229 exipick to access non-standard spools, including the "frozen" queue
1232 NM/01 Bugzilla 965: Support mysql stored procedures.
1233 Patch from Alain Williams
1235 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1237 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1238 Patch from Andreas Metzler
1240 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1241 Patch from Kirill Miazine
1243 NM/05 Bugzilla 671: Added umask to procmail example.
1245 JJ/03 installed exipick 20100323.0, fixing doc bug
1247 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1248 directory. Notification and patch from Dan Rosenberg.
1250 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1252 TK/02 Improve log output when DKIM signing operation fails.
1254 MH/01 Treat the transport option dkim_domain as a colon separated
1255 list, not as a single string, and sign the message with each element,
1256 omitting multiple occurences of the same signer.
1258 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1259 Bugzilla 985, 986. Patch by Simon Arlott
1261 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1262 Patch by Simon Arlott
1264 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1265 MBX locking. Notification from Dan Rosenberg.
1271 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1273 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1275 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1277 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1279 NM/04 Bugzilla 915: Flag broken perl installation during build.
1285 TK/01 Added patch by Johannes Berg that expands the main option
1286 "spamd_address" if it starts with a dollar sign.
1288 TK/02 Write list of recipients to X-Envelope-Sender header when building
1289 the mbox-format spool file for content scanning (suggested by Jakob
1292 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1293 (http://www.dcc-servers.net/) support via dccifd. Activated by
1294 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1296 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1297 by Mark Daniel Reidel <mr@df.eu>.
1299 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1300 When building exim an external PCRE library is now needed -
1301 PCRE is a system library on the majority of modern systems.
1302 See entry on PCRE_LIBS in EDITME file.
1304 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1305 conversation. Added nologin parameter to request.
1306 Patch contributed by Kirill Miazine.
1308 TF/01 Do not log submission mode rewrites if they do not change the address.
1310 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1312 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1313 log files in place. Contributed by Roberto Lima.
1315 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1317 TF/03 Bugzilla 615: When checking the local_parts router precondition
1318 after a local_part_suffix or local_part_prefix option, Exim now
1319 does not use the address's named list lookup cache, since this
1320 contains cached lookups for the whole local part.
1322 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1323 Robert Millan. Documentation is in experimental-spec.txt.
1325 TF/04 Bugzilla 668: Fix parallel build (make -j).
1327 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1329 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1330 Patch provided by Jan Srzednicki.
1332 TF/05 Leading white space used to be stripped from $spam_report which
1333 wrecked the formatting. Now it is preserved.
1335 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1336 that they are available at delivery time.
1338 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1340 TF/08 TLS error reporting now respects the incoming_interface and
1341 incoming_port log selectors.
1343 TF/09 Produce a more useful error message if an SMTP transport's hosts
1344 setting expands to an empty string.
1346 NM/06 Bugzilla 744: EXPN did not work under TLS.
1347 Patch provided by Phil Pennock.
1349 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1350 Patch provided by Richard Godbee.
1352 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1353 acl_smtp_notquit, added index entry.
1355 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1356 Patch provided by Eugene Bujak.
1358 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1359 accept(). Patch provided by Maxim Dounin.
1361 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1362 Patch provided by Phil Pennock.
1364 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1366 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1367 Patch provided by Brad "anomie" Jorsch.
1369 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1370 Patch provided by Dean Brooks.
1372 TK/05 Add native DKIM support (does not depend on external libraries).
1374 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1375 Patch provided by Graeme Fowler.
1377 NM/16 Bugzilla 851: Documentation example syntax fix.
1379 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1381 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1384 NM/19 Bugzilla 745: TLS version reporting.
1385 Patch provided by Phil Pennock.
1387 NM/20 Bugzilla 167: bool: condition support.
1388 Patch provided by Phil Pennock.
1390 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1391 clients. Patch provided by Phil Pennock.
1393 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1394 Patch provided by Brad "anomie" Jorsch.
1396 NM/23 Bugzilla 687: Fix misparses in eximstats.
1397 Patch provided by Heiko Schlittermann.
1399 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1400 Patch provided by Heiko Schlittermann.
1402 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1403 plus update to original patch.
1405 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1407 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1408 Patch provided by David Brownlee.
1410 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1412 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1414 NM/30 Bugzilla 888: TLS documentation bugfixes.
1416 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1418 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1419 Unlike the original bugzilla I have changed all shell scripts in src tree.
1421 NM/33 Bugzilla 898: Transport filter timeout fix.
1422 Patch by Todd Rinaldo.
1424 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1425 Patch by Serge Demonchaux.
1427 NM/35 Bugzilla 39: Base64 decode bug fixes.
1428 Patch by Jakob Hirsch.
1430 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1432 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1434 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1436 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1442 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1443 ALT-N's libdkim that I have put here:
1444 http://duncanthrax.net/exim-experimental/
1446 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1447 were called 'true' and it seems that C99 defines that as a reserved
1448 keyword to be used with 'bool' variable types. That means you could
1449 not include C99-style headers which use bools without triggering
1450 build errors in sieve.c.
1452 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1453 as mailq or other aliases. Changed the --help handling significantly
1454 to do whats expected. exim_usage() emits usage/help information.
1456 SC/01 Added the -bylocaldomain option to eximstats.
1458 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1460 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1462 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1468 PH/01 Another patch from the Sieve maintainer.
1470 PH/02 When an IPv6 address is converted to a string for single-key lookup
1471 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1472 dots are used instead of colons so that keys in lsearch files need not
1473 contain colons. This was done some time before quoting was made available
1474 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1475 (notated using the quote facility) so as to distinguish them from IPv4
1476 keys. This meant that lookups for IP addresses in host lists did not work
1477 for iplsearch lookups.
1479 This has been fixed by arranging for IPv6 addresses to be expressed with
1480 colons if the lookup type is iplsearch. This is not incompatible, because
1481 previously such lookups could never work.
1483 The situation is now rather anomolous, since one *can* have colons in
1484 ordinary lsearch keys. However, making the change in all cases is
1485 incompatible and would probably break a number of configurations.
1487 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1490 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1491 conversion specification without a maximum field width, thereby enabling
1492 a rogue spamd server to cause a buffer overflow. While nobody in their
1493 right mind would setup Exim to query an untrusted spamd server, an
1494 attacker that gains access to a server running spamd could potentially
1495 exploit this vulnerability to run arbitrary code as the Exim user.
1497 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1498 $primary_hostname instead of what libspf2 thinks the hosts name is.
1500 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1501 a directory entry by the name of the lookup key. Previously, if a
1502 symlink pointed to a non-existing file or a file in a directory that
1503 Exim lacked permissions to read, a lookup for a key matching that
1504 symlink would fail. Now it is enough that a matching directory entry
1505 exists, symlink or not. (Bugzilla 503.)
1507 PH/03 The body_linecount and body_zerocount variables are now exported in the
1510 PH/04 Added the $dnslist_matched variable.
1512 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1513 This means they are set thereafter only if the connection becomes
1516 PH/06 Added the client_condition to authenticators so that some can be skipped
1517 by clients under certain conditions.
1519 PH/07 The error message for a badly-placed control=no_multiline_responses left
1520 "_responses" off the end of the name.
1522 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1524 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1525 (without spaces) instead of just copying the configuration text.
1527 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1529 PH/11 Added $max_received_linelength.
1531 PH/12 Added +ignore_defer and +include_defer to host lists.
1533 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1534 way in which PCRE > 7.0 is built.
1536 PH/14 Implemented queue_only_load_latch.
1538 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1539 MAIL command. The effect was to mangle the value on 64-bit systems.
1541 PH/16 Another patch from the Sieve maintainer.
1543 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1545 PH/18 If a system quota error occurred while trying to create the file for
1546 a maildir delivery, the message "Mailbox is full" was not appended to the
1547 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1548 only to a quota excession during the actual writing of the file.
1550 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1551 characters?) which causes problems in log lines. The DN values are now
1552 passed through string_printing() before being added to log lines.
1554 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1555 and InterBase are left for another time.)
1557 PH/21 Added message_body_newlines option.
1559 PH/22 Guard against possible overflow in moan_check_errorcopy().
1561 PH/23 POSIX allows open() to be a macro; guard against that.
1563 PH/24 If the recipient of an error message contained an @ in the local part
1564 (suitably quoted, of course), incorrect values were put in $domain and
1565 $local_part during the evaluation of errors_copy.
1571 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1572 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1575 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1576 issue a MAIL command.
1578 PH/02 In an ACL statement such as
1580 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1582 if a client was not listed at all, or was listed with a value other than
1583 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1584 the condition was not true (as it should be), so access was not denied.
1585 The bug was that the ! inversion was incorrectly passed on to the second
1586 item. This has been fixed.
1588 PH/03 Added additional dnslists conditions == and =& which are different from
1589 = and & when the dns lookup returns more than one IP address.
1591 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1592 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1594 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1595 FSYNC, which compiles an option called disable_fsync that allows for
1596 bypassing fsync(). The documentation is heavily laced with warnings.
1598 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1600 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1601 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1602 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1603 including adding "make clean"; (3) Added -fPIC when compiling the test
1604 dynamically loaded module, to get rid of a warning.
1606 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1607 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1608 = 0s. The bug is otherwise harmless.
1610 PH/07 There was a bug in the dovecot authenticator such that the value of
1611 $auth1 could be overwritten, and so not correctly preserved, after a
1612 successful authentication. This usually meant that the value preserved by
1613 the server_setid option was incorrect.
1615 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1617 PH/09 Installed PCRE release 7.0.
1619 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1620 run for batched SMTP input. It is now run at the start of every message
1621 in the batch. While fixing this I discovered that the process information
1622 (output by running exiwhat) was not always getting set for -bs and -bS
1623 input. This is fixed, and it now also says "batched" for BSMTP.
1625 PH/11 Added control=no_pipelining.
1627 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1628 patch, slightly modified), and move the expansion of helo_data till after
1629 the connection is made in the smtp transport (so it can use these
1632 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1634 PH/14 Added log_selector = +pid.
1636 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1638 PH/16 Add ${if forany and ${if forall.
1640 PH/17 Added dsn_from option to vary the From: line in DSNs.
1642 PH/18 Flush SMTP output before performing a callout, unless control =
1643 no_callout_flush is set.
1645 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1646 was true (the default) a successful delivery failed to delete the retry
1647 item, thus causing premature timeout of the address. The bug is now
1650 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1652 PH/21 Long custom messages for fakedefer and fakereject are now split up
1653 into multiline reponses in the same way that messages for "deny" and
1654 other ACL rejections are.
1656 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1657 with slight modification.
1659 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1660 draft, changing the syntax and factoring some duplicate code".
1662 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1663 for deliveries of the second and subsequent messages over the same SMTP
1666 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1667 ${reduce, with only minor "tidies".
1669 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1671 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1672 expansion side effects.
1674 PH/27 When a message times out after an over-quota error from an Exim-imposed
1675 quota, the bounce message says "mailbox is full". This message was not
1676 being given when it was a system quota that was exceeded. It now should
1679 MH/03 Made $recipients available in local_scan(). local_scan() already has
1680 better access to the recipient list through recipients_list[], but
1681 $recipients can be useful in postmaster-provided expansion strings.
1683 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1684 in the case of a MAIL command with additional options following the
1685 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1686 were accidentally chopped off.
1688 PH/29 SMTP synchronization checks are implemented when a command is read -
1689 there is a check that no more input is waiting when there shouldn't be
1690 any. However, for some commands, a delay in an ACL can mean that it is
1691 some time before the response is written. In this time, more input might
1692 arrive, invalidly. So now there are extra checks after an ACL has run for
1693 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1694 pipelining has not been advertised.
1696 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1698 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1699 returned in the SMTP response when smtp_return_error_details was false.
1700 This has been fixed.
1702 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1703 instead of the C I/O that was originally supplied, because problems were
1704 reported on Solaris.
1706 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1707 Exim which did not show up earlier: it was assuming that a call to
1708 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1709 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1710 was a macro that became an assignment, so it seemed to work. This has
1711 changed to a proper function call with a void return, hence the compile
1712 error. Exim's code has been fixed.
1714 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1717 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1719 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1721 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1722 because it turns out to be too big or there is a timeout) there is no
1723 "Completed" line in the log. When some messages of this type were
1724 selected by exigrep, they were listed as "not completed". Others were
1725 picked up by some special patterns. I have improved the selection
1726 criteria to be more general.
1728 PH/38 The host_find_failed option in the manualroute router can now be set
1729 to "ignore", to completely ignore a host whose IP address cannot be
1730 found. If all hosts are ignored, the behaviour is controlled by the new
1731 host_all_ignored option.
1733 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1734 homing or because of multiple MX records with /mx) generated more than
1735 one IP address, and the following item turned out to be the local host,
1736 all the secondary addresses of the first item were incorrectly removed
1737 from the list, along with the local host and any following hosts (which
1738 is what is supposed to happen).
1740 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1741 whoever called Exim into the -H file. In the case of the daemon it was
1742 behaving confusingly. When first started, it used values for whoever
1743 started the daemon, but after a SIGHUP it used the Exim user (because it
1744 calls itself on a restart). I have changed the code so that it now always
1747 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1748 message are rejected with the same error (e.g. no authentication or bad
1749 sender address), and a DATA command is nevertheless sent (as can happen
1750 with PIPELINING or a stupid MUA), the error message that was given to the
1751 RCPT commands is included in the rejection of the DATA command. This is
1752 intended to be helpful for MUAs that show only the final error to their
1755 PH/42 Another patch from the Sieve maintainer.
1757 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1758 Eximstats - Fixed some broken HTML links and added missing column headers
1760 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1761 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1763 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1765 PH/43 Yet another patch from the Sieve maintainer.
1767 PH/44 I found a way to check for a TCP/IP connection going away before sending
1768 the response to the final '.' that terminates a message, but only in the
1769 case where the client has not sent further data following the '.'
1770 (unfortunately, this is allowed). However, in many cases there won't be
1771 any further data because there won't be any more messages to send. A call
1772 to select() can be used: if it shows that the input is "ready", there is
1773 either input waiting, or the socket has been closed. An attempt to read
1774 the next input character can distinguish the two cases. Previously, Exim
1775 would have sent an OK response which the client would never have see.
1776 This could lead to message repetition. This fix should cure that, at
1777 least in a lot of common cases.
1779 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1780 advertised in response to EHLO.
1786 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1787 fixed by 4.65/MH/01 (is this a record?) are fixed:
1789 (i) An empty string was always treated as zero by the numeric comparison
1790 operators. This behaviour has been restored.
1792 (ii) It is documented that the numeric comparison operators always treat
1793 their arguments as decimal numbers. This was broken in that numbers
1794 starting with 0 were being interpreted as octal.
1796 While fixing these problems I realized that there was another issue that
1797 hadn't been noticed. Values of message_size_limit (both the global option
1798 and the transport option) were treated as octal if they started with 0.
1799 The documentation was vague. These values are now always treated as
1800 decimal, and I will make that clear in the documentation.
1806 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1807 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1810 MH/01 Don't check that the operands of numeric comparison operators are
1811 integers when their expansion is in "skipping" mode (fixes bug
1812 introduced by 4.64-PH/07).
1814 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1815 child addresses, Exim now panics and dies. Previously, because the count
1816 is held in a short int, deliveries were likely to be lost. As such a
1817 large number of recipients for a single message is ridiculous
1818 (performance will be very, very poor), I have chosen to impose a limit
1819 rather than extend the field.
1825 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1826 leftover -K file (the existence of which was triggered by #402).
1827 While we were at it, introduced process PID as part of the -K
1828 filename. This should rule out race conditions when creating
1831 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1832 processing considerably. Previous code took too long for large mails,
1833 triggering a timeout which in turn triggers #401.
1835 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1836 in the DK code in transports.c. sendfile() is not really portable,
1837 hence the _LINUX specificness.
1839 TF/01 In the add_headers option to the mail command in an Exim filter,
1840 there was a bug that Exim would claim a syntax error in any
1841 header after the first one which had an odd number of characters
1844 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1845 callout verification, Exim cached a "reject" for the entire domain. This
1846 is correct for most verifications, but it is not correct for a recipient
1847 verification with use_sender or use_postmaster set, because in that case
1848 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1849 case of MAIL FROM:<> rejection from other early rejections (e.g.
1850 rejection of HELO). When verifying a recipient using a non-null MAIL
1851 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1852 Whatever the result of the callout, the value of the domain cache is
1853 left unchanged (for any other kind of callout, getting as far as trying
1854 RCPT means that the domain itself is ok).
1856 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1859 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1860 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1861 session, the daemon ignores ECONNECT errors and logs others; it now
1862 ignores EPIPE as well.
1864 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1865 (quoted-printable decoding).
1867 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1868 later the small subsequent patch to fix an introduced bug.
1870 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1872 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1874 PH/08 An error is now given if message_size_limit is specified negative.
1876 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1877 to be given (somewhat) arbitrary names.
1879 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1882 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1883 miscellaneous code fixes
1885 PH/10 Added the log_reject_target ACL modifier to specify where to log
1888 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1889 hostname. This is wrong, because it relates to the incoming message (and
1890 probably the interface on which it is arriving) and not to the outgoing
1891 callout (which could be using a different interface). This has been
1892 changed to use the value of the helo_data option from the smtp transport
1893 instead - this is what is used when a message is actually being sent. If
1894 there is no remote transport (possible with a router that sets up host
1895 addresses), $smtp_active_hostname is used.
1897 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1898 tweaks were necessary in order to get it to work (see also 21 below):
1899 (a) The code assumed that strncpy() returns a negative number on buffer
1900 overflow, which isn't the case. Replaced with Exim's string_format()
1902 (b) There were several signed/unsigned issues. I just did the minimum
1903 hacking in of casts. There is scope for a larger refactoring.
1904 (c) The code used strcasecmp() which is not a standard C function.
1905 Replaced with Exim's strcmpic() function.
1906 (d) The code set only $1; it now sets $auth1 as well.
1907 (e) A simple test gave the error "authentication client didn't specify
1908 service in request". It would seem that Dovecot has changed its
1909 interface. Fortunately there's a specification; I followed it and
1910 changed what the client sends and it appears to be working now.
1912 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1915 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1916 address A is aliased to B and C, where B exists and C does not. Without
1917 -v the output is "A verified" because verification stops after a
1918 successful redirection if more than one address is generated. However,
1919 with -v the child addresses are also verified. Exim was outputting "A
1920 failed to verify" and then showing the successful verification for C,
1921 with its parentage. It now outputs "B failed to verify", showing B's
1922 parentage before showing the successful verification of C.
1924 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1925 look up a TXT record in a specific list after matching in a combined
1928 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1929 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1930 they consult the DNS. I had assumed they would set it the way they
1931 wanted; and indeed my experiments on Linux seem to show that in some
1932 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1933 To be on the safe side, however, I have now made the interface to
1934 host_find_byname() similar to host_find_bydns(), with an argument
1935 containing the DNS resolver options. The host_find_byname() function now
1936 sets these options at its start, just as host_find_bydns() does. The smtp
1937 transport options dns_qualify_single and dns_search_parents are passed to
1938 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1939 of host_find_byname() use the default settings of RES_DEFNAMES
1940 (qualify_single) but not RES_DNSRCH (search_parents).
1942 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1943 spool_read_header() do less string testing, by means of a preliminary
1944 switch on the second character of optional "-foo" lines. (This is
1945 overdue, caused by the large number of possibilities that now exist.
1946 Originally there were few.) While I was there, I also converted the
1947 str(n)cmp tests so they don't re-test the leading "-" and the first
1948 character, in the hope this might squeeze out yet more improvement.
1950 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1951 flag allowing group syntax was set by the header_syntax check but not
1952 turned off, possible causing trouble later; (2) The flag was not being
1953 set at all for the header_verify test, causing "group"-style headers to
1954 be rejected. I have now set it in this case, and also caused header_
1955 verify to ignore an empty address taken from a group. While doing this, I
1956 came across some other cases where the code for allowing group syntax
1957 while scanning a header line wasn't quite right (mostly, not resetting
1958 the flag correctly in the right place). These bugs could have caused
1959 trouble for malformed header lines. I hope it is now all correct.
1961 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1962 with the "reply" argument non-NULL. The code, however (which originally
1963 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1964 but it didn't always do it. This confused somebody who was copying the
1965 code for some other use. I have removed all the tests.
1967 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1968 feature that was used to support insecure browsers during the U.S. crypto
1969 embargo. It requires special client support, and Exim is probably the
1970 only MTA that supported it -- and would never use it because real RSA is
1971 always available. This code has been removed, because it had the bad
1972 effect of slowing Exim down by computing (never used) parameters for the
1973 RSA_EXPORT functionality.
1975 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1976 authenticator to fail if there's a tab character in the incoming data
1977 (there should never be unless someone is messing about, as it's supposed
1978 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1979 if the connection is using TLS or if the remote IP is the same as the
1980 local IP, and the "valid-client-cert option" if a client certificate has
1983 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1984 authenticators. This can be used for authorization after authentication
1985 succeeds. (In the case of plaintext, it servers for both authentication
1988 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1989 if any retry times were supplied.
1991 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1992 connection, where there is no client IP address to check. In this
1993 situation, the verify now always succeeds.
1995 PH/25 Applied John Jetmore's -Mset patch.
1997 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1999 PH/27 In a string expansion for a processed (not raw) header when multiple
2000 headers of the same name were present, leading whitespace was being
2001 removed from all of them, but trailing whitespace was being removed only
2002 from the last one. Now trailing whitespace is removed from each header
2003 before concatenation. Completely empty headers in a concatenation (as
2004 before) are ignored.
2006 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2007 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2009 PH/29 [Removed. This was a change that I later backed out, and forgot to
2010 correct the ChangeLog entry (that I had efficiently created) before
2011 committing the later change.]
2013 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2014 address errors (4xx response to RCPT) over the same connection as other
2015 messages routed to the same hosts. Such deliveries are always "forced",
2016 so retry times are not inspected. This resulted in far too many retries
2017 for the affected addresses. The effect occurred only when there were more
2018 hosts than the hosts_max_try setting in the smtp transport when it had
2019 the 4xx errors. Those hosts that it had tried were not added to the list
2020 of hosts for which the message was waiting, so if all were tried, there
2021 was no problem. Two fixes have been applied:
2023 (i) If there are any address or message errors in an SMTP delivery, none
2024 of the hosts (tried or untried) are now added to the list of hosts
2025 for which the message is waiting, so the message should not be a
2026 candidate for sending over the same connection that was used for a
2027 successful delivery of some other message. This seems entirely
2028 reasonable: after all the message is NOT "waiting for some host".
2029 This is so "obvious" that I'm not sure why it wasn't done
2030 previously. Hope I haven't missed anything, but it can't do any
2031 harm, as the worst effect is to miss an optimization.
2033 (ii) If, despite (i), such a delivery is accidentally attempted, the
2034 routing retry time is respected, so at least it doesn't keep
2035 hammering the server.
2037 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2038 in ${readsocket because some servers need this prod.
2040 PH/32 Added some extra debug output when updating a wait-xxx database.
2042 PH/33 The hint "could be header name not terminated by colon", which has been
2043 given for certain expansion errors for a long time, was not being given
2044 for the ${if def:h_colon_omitted{... case.
2046 PH/34 The spec says: "With one important exception, whenever a domain list is
2047 being scanned, $domain contains the subject domain." There was at least
2048 one case where this was not true.
2050 PH/35 The error "getsockname() failed: connection reset by peer" was being
2051 written to the panic log as well as the main log, but it isn't really
2052 panic-worthy as it just means the connection died rather early on. I have
2053 removed the panic log writing for the ECONNRESET error when getsockname()
2056 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2057 runs only) independently of the message's sender address. This meant
2058 that, if the 4xx error was in fact related to the sender, a different
2059 message to the same recipient with a different sender could confuse
2060 things. In particualar, this can happen when sending to a greylisting
2061 server, but other circumstances could also provoke similar problems.
2062 I have changed the default so that the retry time for these errors is now
2063 based a combination of the sender and recipient addresses. This change
2064 can be overridden by setting address_retry_include_sender=false in the
2067 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2068 remote server are returned as part of bounce messages. This was not
2069 happening for LMTP over a pipe (the lmtp transport), but now it is the
2070 same for both kinds of LMTP.
2072 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2073 in header lines that were in fact CNAMEs. This is no longer the case.
2075 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2076 and queue runs started by the daemon processed all messages. This has
2077 been fixed so that -R and -S can now usefully be given with -q<time>.
2079 PH/40 Import PCRE release 6.7 (fixes some bugs).
2081 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2083 PH/42 Give an error if -q is specified more than once.
2085 PH/43 Renamed the variables $interface_address and $interface_port as
2086 $received_ip_address and $received_port, to make it clear that these
2087 values apply to message reception, and not to the outgoing interface when
2088 a message is delivered. (The old names remain recognized, of course.)
2090 PH/44 There was no timeout on the connect() call when using a Unix domain
2091 socket in the ${readsocket expansion. There now is.
2093 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2094 be meaningful with "accept".
2096 SC/01 Eximstats V1.43
2097 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2099 SC/02 Eximstats V1.44
2100 Use a glob alias rather than an array ref in the generated
2101 parser. This improves both readability and performance.
2103 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2104 Collect SpamAssassin and rejection statistics.
2105 Don't display local sender or destination tables unless
2106 there is data to show.
2107 Added average volumes into the top table text output.
2109 SC/04 Eximstats V1.46
2110 Collect data on the number of addresses (recipients)
2111 as well as the number of messages.
2113 SC/05 Eximstats V1.47
2114 Added 'Message too big' to the list of mail rejection
2115 reasons (thanks to Marco Gaiarin).
2117 SC/06 Eximstats V1.48
2118 Mainlog lines which have GMT offsets and are too short to
2119 have a flag are now skipped.
2121 SC/07 Eximstats V1.49 (Alain Williams)
2122 Added the -emptyok flag.
2124 SC/08 Eximstats V1.50
2125 Fixes for obtaining the IP address from reject messages.
2127 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2128 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2129 whitesspace changes from 4.64-PH/27
2131 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2134 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2135 are found, allow negative numbers in numeric criteria)
2137 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2139 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2140 to match changes made in 4.64-PH/43
2142 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2144 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2145 "domain" is actually an IP address, and force a failure. This locks out
2146 those revolvers/nameservers that support "A-for-A" lookups, in
2147 contravention of the specifications.
2149 PH/48 When a host name was looked up from an IP address, and the subsequent
2150 forward lookup of the name timed out, the host name was left in
2151 $sender_host_name, contrary to the specification.
2153 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2154 restricted to single-key lookups, Exim was not diagnosing an error if
2155 * or *@ was used with a query-style lookup.
2157 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2159 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2160 long ago, but noone interested enough thought of it. Let's just say that
2161 the "1.1" means that there are some new functions that weren't there at
2162 some point in the past.
2164 PH/51 Error processing for expansion failure of helo_data from an smtp
2165 transport during callout processing was broken.
2167 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2168 tested/used via the -bh/-bhc/-bs options.
2170 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2171 bug, fixed in subsequent PCRE releases).
2173 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2174 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2176 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2182 SC/01 Use a glob alias rather than an array ref in eximstats generated
2183 parser. This improves both readability and performance.
2185 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2186 Don't display local sender or destination tables in eximstats unless
2187 there is data to show.
2188 Added average volumes into the eximstats top table text output.
2190 SC/03 Collect data on the number of addresses (recipients) as well
2191 as the number of messages in eximstats.
2193 TF/01 Correct an error in the documentation for the redirect router. Exim
2194 does (usually) call initgroups() when daemonizing.
2196 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2197 with consistent privilege compared to when running as a daemon.
2199 TF/03 Note in the spec that $authenticated_id is not set for local
2200 submissions from trusted users.
2202 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2203 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2205 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2206 by adding some example configuration directives to the default
2207 configuration file. A little bit of work is required to uncomment the
2208 directives and define how usernames and passwords are checked, but
2209 there is now a framework to start from.
2211 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2212 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2213 without this. I don't know how relevant this is to other LDAP libraries.
2215 PH/02 Add the verb name to the "unknown ACL verb" error.
2217 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2219 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2221 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2222 directory not expanded when it should be if an expanded home directory
2223 was set for the address (which is overridden by the transport).
2225 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2228 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2229 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2230 because it is too late at that time, and has no effect.
2232 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2233 security issue with \' (bugzilla #107). I could not use the
2234 PQescapeStringConn() function, because it needs a PGconn value as one of
2237 PH/08 When testing addresses using -bt, indicate those final addresses that
2238 are duplicates that would not cause an additional delivery. At least one
2239 person was confused, thinking that -bt output corresponded to deliveries.
2240 (Suppressing duplicates isn't a good idea as you lose the information
2241 about possibly different redirections that led to the duplicates.)
2243 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2244 systems where poll() doesn't work, in particular OS X.
2246 PH/10 Added more information to debugging output for retry time not reached.
2248 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2249 operations in malware.c.
2251 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2254 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2255 syslog_duplication set false, log lines that would normally be written
2256 both the the main log and to the reject log were not written to syslog at
2259 PH/14 In the default configuration, change the use of "message" in ACL warn
2260 statements to "add_header".
2262 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2263 not followed by a command (e.g. "seen endif").
2265 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2266 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2269 PH/17 Added extra conditions to the default value of delay_warning_condition
2273 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2274 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2275 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2278 The Auto-Submitted: and various List- headers are standardised, whereas I
2279 don't think Precedence: ever was.
2281 PH/18 Refactored debugging code in route_finduser() to show more information,
2282 in particular, the error code if getpwnam() issues one.
2284 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2285 This is apparently needed in addition to the PH/07 change above to avoid
2286 any possible encoding problems.
2288 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2289 but not after initializing Perl.
2291 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2292 output them only if debugging. By default they are written stderr,
2293 apparently, which is not desirable.
2295 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2298 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2301 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2303 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2304 authenticated or an ident call has been made. Suppress the default
2305 values for $authenticated_id and $authenticated_sender (but permit -oMai
2306 and -oMas) when testing with -bh.
2308 PH/24 Re-jigged the order of the tests in the default configuration so that the
2309 tests for valid domains and recipients precede the DNS black list and CSA
2310 tests, on the grounds that those ones are more expensive.
2312 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2313 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2314 This bug exists in every version of Exim that I still have, right back to
2317 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2318 However, an attempt to turn on case-sensitivity in a regex key by
2319 including (?-i) didn't work because the subject string was already
2320 lowercased, and the effects were non-intuitive. It turns out that a
2321 one-line patch can be used to allow (?-i) to work as expected.
2327 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2328 other effects) broke the use of negated acl sub-conditions.
2330 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2333 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2334 "Deny" causes Exim to reject the incoming connection with a 554 error.
2335 Unfortunately, if there is a major crisis, such as a disk failure,
2336 tcp-wrappers gives "deny", whereas what one would like would be some
2337 kind of temporary error. A kludge has been added to help with this.
2338 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2339 554 error is used if errno is still zero or contains ENOENT (which occurs
2340 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2343 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2345 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2346 errors. Otherwise a message that provokes a temporary error (when other
2347 messages do not) can cause a whole host to time out.
2349 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2350 addresses were routed directly to files or pipes from a redirect router.
2351 File deliveries just didn't batch; pipe deliveries might have suffered
2354 PH/06 A failure to get a lock for a hints database would erroneously always say
2355 "Failed to get write lock", even when it was really a read lock.
2357 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2358 of 0600. This has been changed to use the value of the lockfile_mode
2359 option (which defaults to 0600).
2361 PH/08 Applied small patch from the Sieve maintainer.
2363 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2364 folder from quota calculations, a direct delivery into this folder messed
2365 up the contents of the maildirsize file. This was because the regex was
2366 used only to exclude .Trash (or whatever) when the size of the mailbox
2367 was calculated. There was no check that a delivery was happening into an
2368 excluded directory. This bug has been fixed by ignoring all quota
2369 processing for deliveries into excluded directories.
2371 PH/10 Added the maildirfolder_create_regex option to appendfile.
2377 PH/01 The code for finding all the local interface addresses on a FreeBSD
2378 system running IPv6 was broken. This may well have applied to all BSD
2379 systems, as well as to others that have similar system calls. The broken
2380 code found IPv4 interfaces correctly, but gave incorrect values for the
2381 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2382 that it would not match correctly against @[] and not recognize the IPv6
2385 PH/02 The ipliteral router was not recognizing addresses of the form user@
2386 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2388 PH/03 Added disable_ipv6.
2390 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2391 decoded form, because it is most often used to construct To: headers
2392 lines in autoreplies, and the decoded form may well be syntactically
2393 invalid. However, $reply_address has leading white space removed, and all
2394 newlines turned into spaces so that the autoreply transport does not
2397 PH/05 If group was specified without a user on a router, and no group or user
2398 was specified on a transport, the group from the router was ignored.
2400 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2401 for visible compile-time settings that can be used to change these
2402 numbers, for those that want even more. Backwards compatibility with old
2403 spool files has been maintained. However, going back to a previous Exim
2404 release will lost any variables that are in spool files.
2406 PH/07 Two small changes when running in the test harness: increase delay when
2407 passing a TCP/IP connection to a new process, in case the original
2408 process has to generate a bounce, and remove special handling of
2409 127.0.0.2 (sic), which is no longer necessary.
2411 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2412 be the same on different OS.
2414 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2417 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2418 whether --show-vars was specified or not
2420 JJ/02 exipick: Added support for new ACL variable spool format introduced
2423 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2424 syntactically invalid From: or Reply-to: line, and a filter used this to
2425 generate an autoreply, and therefore failed to obtain an address for the
2426 autoreply, Exim could try to deliver to a non-existent relative file
2427 name, causing unrelated and misleading errors. What now happens is that
2428 it logs this as a hard delivery error, but does not attempt to create a
2431 PH/11 The exinext utility has a -C option for testing purposes, but although
2432 the given file was scanned by exinext itself; it wasn't being passed on
2433 when Exim was called.
2435 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2436 an end-of-file indication when reading a command response.
2438 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2439 compiled. In many other places in Exim, IPv6 addresses are always
2440 recognized, so I have changed this. It also means that IPv4 domain
2441 literals of the form [IPV4:n.n.n.n] are now always recognized.
2443 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2444 used if the router is not running as root, for example, when verifying at
2445 ACL time, or when using -bh. The debugging output from this situation was
2446 non-existent - all you got was a failure to exec. I have made two
2449 (a) Failures to set uid/gid, the current directory, or a process leader
2450 in a subprocess such as that created by queryprogram now generate
2451 suitable debugging ouput when -d is set.
2453 (b) The queryprogram router detects when it is not running as root,
2454 outputs suitable debugging information if -d is set, and then runs
2455 the subprocess without attempting to change uid/gid.
2457 PH/15 Minor change to Makefile for building test_host (undocumented testing
2460 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2461 additional section of a DNS packet that returns MX or SRV records.
2462 Instead, it always explicitly searches for A/AAAA records. This avoids
2463 major problems that occur when a DNS server includes only records of one
2464 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2465 fixed another bug: if SRV records were looked up and the corresponding
2466 address records were *not* found in the additional section, the port
2467 values from the SRV records were lost.
2469 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2470 using the correct key (the original address) when searching the retry
2471 rules in order to find which one to use for generating the retry hint.
2473 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2474 adding the default one. Similarly, if it contains a Reply-To: header, the
2475 errors_reply_to option, if set, is not used.
2477 PH/19 When calculating a retry time, Exim used to measure the "time since
2478 failure" by looking at the "first failed" field in the retry record. Now
2479 it does not use this if it is later than than the arrival time of the
2480 message. Instead it uses the arrival time. This makes for better
2481 behaviour in cases where some deliveries succeed, thus re-setting the
2482 "first failed" field. An example is a quota failure for a huge message
2483 when small messages continue to be delivered. Without this change, the
2484 "time since failure" will always be short, possible causing more frequent
2485 delivery attempts for the huge message than are intended.
2486 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2488 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2489 $1, $2, $3) because the numerical variables can be reset during some
2490 expansion items (e.g. "match"), thereby losing the authentication data.
2492 PH/21 Make -bV show the size of off_t variables so that the test suite can
2493 decide whether to run tests for quotas > 2G.
2495 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2496 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2497 filecount value is greater than 2G or if a quota value is greater than 2G
2498 on a system where the size of off_t is not greater than 4, a panic error
2501 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2502 never match. The debug and -bh output now contains an explicit error
2503 message indicating a malformed IPv4 address or mask.
2505 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2506 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2507 PH/23 above applies.
2509 PH/25 Do not write to syslog when running in the test harness. The only
2510 occasion when this arises is a failure to open the main or panic logs
2511 (for which there is an explicit test).
2513 PH/26 Added the /no_tell option to "control=freeze".
2515 PH/27 If a host name lookup failed very early in a connection, for example, if
2516 the IP address matched host_lookup and the reverse lookup yielded a name
2517 that did not have a forward lookup, an error message of the form "no IP
2518 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2519 could be logged. Now it outputs the IP address instead of "NULL".
2521 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2522 allows the sender and the authenticated sender to be set when
2523 submitting a message from within Exim. Since child_open_exim() is
2524 documented for local_scan(), the new function should be too.
2526 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2527 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2528 results in an empty string is now treated as unset.
2530 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2532 PH/31 Added sender_verify_fail logging option.
2534 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2535 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2536 tidied the source and removed it altogether.
2538 PH/33 When a queue run was abandoned because the load average was too high, a
2539 log line was always written; now it is written only if the queue_run log
2540 selector is set. In addition, the log line for abandonment now contains
2541 information about the queue run such as the pid. This is always present
2542 in "start" and "stop" lines but was omitted from the "abandon" line.
2544 PH/34 Omit spaces between a header name and the colon in the error message that
2545 is given when verify = headers_syntax fails (if there are lots of them,
2546 the message gets confusing).
2548 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2549 names, as there are now some PTR records that contain slashes. This check
2550 is only to protect against broken name servers that fall over on strange
2551 characters, so the fact that it applies to all lookups doesn't matter.
2553 PH/36 Now that the new test suite is complete, we can remove some of the
2554 special code in Exim that was needed for the old test suite. For example,
2555 sorting DNS records because real resolvers return them in an arbitrary
2556 order. The new test suite's fake resolver always returns records in the
2559 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2560 bounces) except when queue_only is set, to avoid logging races between
2561 the different processes.
2563 PH/38 Panic-die if .include specifies a non-absolute path.
2565 PH/39 A tweak to the "H" retry rule from its user.
2567 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2568 a label. They prevented compilation on older perls.
2570 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2571 a warning to be raised on newish perls.
2573 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2574 on queue. Changes to match documented behaviour of showing count of
2575 messages matching specified criteria.
2577 PH/40 Changed the default ident timeout from 30s to 5s.
2579 PH/41 Added support for the use of login_cap features, on those BSD systems
2580 that have them, for controlling the resources used by pipe deliveries.
2582 PH/42 The content-scanning code uses fopen() to create files in which to put
2583 message data. Previously it was not paying any attention to the mode of
2584 the files. Exim runs with umask(0) because the rest of the code creates
2585 files with open(), and sets the required mode explicitly. Thus, these
2586 files were ending up world-writeable. This was not a big issue, because,
2587 being within the spool directory, they were not world-accessible. I have
2588 created a function called modefopen, which takes an additional mode
2589 argument. It sets umask(777), creates the file, chmods it to the required
2590 mode, then resets the umask. All the relevant calls to fopen() in the
2591 content scanning code have been changed to use this function.
2593 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2594 to 24 hours. This avoids potential overflow problems when processing G
2595 and H retry rules. I suspect nobody ever tinkers with this value.
2597 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2599 PH/45 When the plaintext authenticator is running as a client, the server's
2600 challenges are checked to ensure they are valid base64 strings. By
2601 default, the authentication attempt is cancelled if an invalid string is
2602 received. Setting client_ignore_invalid_base64 true ignores these errors.
2603 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2604 they are received. Thus, the responses can be made to depend on the
2605 challenges. If an invalid string is ignored, an empty string is placed in
2608 PH/46 Messages that are created by the autoreply transport now contains a
2609 References: header, in accordance with RFCs 2822 and 3834.
2611 PH/47 Added authenticated_sender_force to the smtp transport.
2613 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2615 PH/49 Installed latest patch from the Sieve maintainer.
2617 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2618 was also set, the appendfile transport was unnecessarily scanning a
2619 directory of message files (e.g. for maildir delivery) to find the count
2620 of files (along with the size), even though it did not need this
2621 information. It now does the scan only if it needs to find either the
2622 size of the count of files.
2624 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2626 PH/52 Two bugs concerned with error handling when the smtp transport is
2629 (i) Exim was not creating retry information for temporary errors given
2630 for individual recipients after the DATA command when the smtp transport
2631 was used in LMTP mode. This meant that they could be retried too
2632 frequently, and not timed out correctly.
2634 (ii) Exim was setting the flag that allows error details to be returned
2635 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2636 recipients that were returned after the DATA command.
2638 PH/53 This is related to PH/52, but is more general: for any failing address,
2639 when detailed error information was permitted to be returned to the
2640 sender, but the error was temporary, then after the final timeout, only
2641 "retry timeout exceeded" was returned. Now it returns the full error as
2642 well as "retry timeout exceeded".
2644 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2645 do this, and (what is worse) MTAs that accept it.
2647 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2648 will now be deprecated.
2650 PH/56 New os.c-cygwin from the Cygwin maintainer.
2652 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2653 formats (previously only available in exim formats via -bpr, -bpru,
2654 and -bpra. Now also available in native and exiqgrep formats)
2656 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2657 with very large, slow to parse queues
2659 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2661 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2663 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2664 responses to authentication challenges, though it was showing the
2665 challenges; (ii) I've removed the CR characters from the debug output for
2668 PH/58 Allow for the insertion of a newline as well as a space when a string
2669 is turned into more than one encoded-word during RFC 2047 encoding. The
2670 Sieve code now uses this.
2672 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2673 data_4xx, lost_connection, tls_required.
2675 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2676 message was being sent as an SMTP response.
2678 PH/61 Add -l and -k options to exicyclog.
2680 PH/62 When verifying, if an address was redirected to one new address, so that
2681 verification continued, and the new address failed or deferred after
2682 having set something in $address_data, the value of $address_data was not
2683 passed back to the ACL. This was different to the case when no
2684 redirection occurred. The value is now passed back in both cases.
2686 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2687 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2688 use login_cap.h, so on its own it isn't the distinguishing feature. The
2689 new name refers directly to the setclassresources() function.
2691 PH/65 Added configuration files for NetBSD3.
2693 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2695 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2696 is preferred over IPv4.
2698 PH/68 The bounce_return_message and bounce_return_body options were not being
2699 honoured for bounces generated during the reception of non-SMTP messages.
2700 In particular, this applied to messages rejected by the ACL. This bug has
2701 been fixed. However, if bounce_return_message is true and bounce_return_
2702 body is false, the headers that are returned for a non-SMTP message
2703 include only those that have been read before the error was detected.
2704 (In the case of an ACL rejection, they have all been read.)
2706 PH/69 The HTML version of the specification is now built in a directory called
2707 spec_html instead of spec.html, because the latter looks like a path with
2708 a MIME-type, and this confuses some software.
2710 PH/70 Catch two compiler warnings in sieve.c.
2712 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2713 function verify_get_ident() calls ip_connect() to connect a socket, but
2714 if the "connect()" function timed out, ip_connect() used to close the
2715 socket. However, verify_get_ident() also closes the socket later, and in
2716 between Exim writes to the log, which may get opened at this point. When
2717 the socket was closed in ip_connect(), the log could get the same file
2718 descriptor number as the socket. This naturally causes chaos. The fix is
2719 not to close the socket in ip_connect(); the socket should be closed by
2720 the function that creates it. There was only one place in the code where
2721 this was missing, in the iplookup router, which I don't think anybody now
2722 uses, but I've fixed it anyway.
2724 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2725 well as to direct DNS lookups. Otherwise the handling of names in host
2726 lists is inconsistent and therefore confusing.
2732 PH/01 Two changes to the default runtime configuration:
2734 (1) Move the checks for relay_from_hosts and authenticated clients from
2735 after to before the (commented out) DNS black list checks.
2737 (2) Add control=submission to the relay_from_hosts and authenticated
2738 clients checks, on the grounds that messages accepted by these
2739 statements are most likely to be submissions.
2741 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2743 (1) Generate an error if the third argument for the ${prvs expansion is
2746 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2749 (3) Reset the variables that are obtained from the first argument of
2750 ${prvscheck and used in the second argument before leaving the code,
2751 because their memory is reclaimed, so using them afterwards may do
2754 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2755 one (it's much easier than Tom thought :-).
2757 (5) Because of (4), we can now allow for the use of $prvscheck_result
2758 inside the third argument.
2760 PH/03 For some reason, the default setting of PATH when running a command from
2761 a pipe transport was just "/usr/bin". I have changed it to
2764 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2765 anything to be listed in the output from -bV.
2767 PH/05 When a filter generated an autoreply, the entire To: header line was
2768 quoted in the delivery log line, like this:
2770 => >A.N.Other <ano@some.domain> <original@ddress> ...
2772 This has been changed so that it extracts the operative address. There
2773 may be more than one such address. If so, they are comma-separated, like
2776 => >ano@some.domain,ona@other.domain <original@ddress> ...
2778 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2779 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2780 not being looked up in the rDNS to get a host name, Exim was showing the
2781 IP address twice in Received: lines, even though the IP addresses were
2782 identical. For example:
2784 Received: from [1.2.3.4] (helo=[1.2.3.4])
2786 However, if the real host name was known, it was omitting the HELO data
2787 if it matched the actual IP address. This has been tidied up so that it
2788 doesn't show the same IP address twice.
2790 PH/07 When both +timestamp and +memory debugging was on, the value given by
2791 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2792 called by the debug printing, thereby overwriting the timestamp buffer.
2793 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2795 PH/08 When the original message was included in an autoreply transport, it
2796 always said "this is a copy of the message, including all the headers",
2797 even if body_only or headers_only was set. It now gives an appropriate
2800 PH/09 Applied a patch from the Sieve maintainer which:
2802 o fixes some comments
2803 o adds the (disabled) notify extension core
2804 o adds some debug output for the result of if/elsif tests
2805 o points to the current vacation draft in the documentation
2806 and documents the missing references header update
2810 o fixes a bug in processing the envelope test (when testing
2811 multiple envelope elements, the last element determinted the
2814 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2815 Electronic Mail") by including:
2817 Auto-submitted: auto-generated
2819 in the messages that it generates (bounce messages and others, such as
2820 warnings). In the case of bounce messages for non-SMTP mesages, there was
2821 also a typo: it was using "Auto_submitted" (underscore instead of
2822 hyphen). Since every message generated by Exim is necessarily in response
2823 to another message, thes have all been changed to:
2825 Auto-Submitted: auto-replied
2827 in accordance with these statements in the RFC:
2829 The auto-replied keyword:
2831 - SHOULD be used on messages sent in direct response to another
2832 message by an automatic process,
2834 - MUST NOT be used on manually-generated messages,
2836 - MAY be used on Delivery Status Notifications (DSNs) and Message
2837 Disposition Notifications (MDNs),
2839 - MUST NOT be used on messages generated by automatic or periodic
2840 processes, except for messages which are automatic responses to
2843 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2844 to the default Received: header definition.
2846 PH/12 Added log selector acl_warn_skipped (default on).
2848 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2849 variables because (a) they are in the wrong storage pool and (b) even if
2850 they were copied, it wouldn't work properly because of the caching.
2852 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2853 checking when decoding. Apparently there are clients that generate
2854 overlong encoded strings. Why am I not surprised?
2856 PH/15 If the first argument of "${if match_address" was not empty, but did not
2857 contain an "@" character, Exim crashed. Now it writes a panic log message
2858 and treats the condition as false.
2860 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2862 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2863 extension "envelope-auth". The code is finished and in agreement with
2864 other implementations, but there is no documentation so far and in fact,
2865 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2866 not changing the active code.
2868 Print executed "if" and "elsif" statements when debugging is used. This
2869 helps a great deal to understand what a filter does.
2871 Document more things not specified clearly in RFC3028. I had all this
2872 sorted out, when out of a sudden new issues came to my mind. Oops."
2874 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2877 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2878 canonical form (as documented). However, after a host name lookup from
2879 the IP address, check_host() was doing a simple string comparison with
2880 addresses acquired from the DNS when checking that the found name did
2881 have the original IP as one of its addresses. Since any found IPv6
2882 addresses are likely to be in abbreviated form, the comparison could
2883 fail. Luckily, there already exists a function for doing the comparison
2884 by converting both addresses to binary, so now that is used instead of
2885 the text comparison.
2887 PH/20 There was another similar case to PH/19, when a complete host name was
2888 given in a host list; looking up its IP address could give an abbreviated
2889 form, whereas the current host's name might or might not be abbreviated.
2890 The same fix has been applied.
2896 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2897 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2900 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2901 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2903 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2905 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2906 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2907 Also, for "Auto-Submitted", treat anything other than "no" as
2908 non-personal, in accordance with RFC 3834. (Previously it treated
2909 anything starting "auto-" as non-personal.)
2911 TF/01 The control=submission/name=... option had a problem with syntax
2912 errors if the name included a slash character. The /name= option
2913 now slurps the rest of the string, so it can include any characters
2914 but it must come last in the list of options (after /sender_retain
2917 PH/05 Some modifications to the interface to the fake nameserver for the new
2925 TK/01 Added the "success_on_redirect" address verification option. See
2926 NewStuff for rationale and an example.
2928 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2930 PH/02 Patch to exigrep to allow it to work on syslog lines.
2932 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2933 fread() to skip over the body file's header line, because in Cygwin the
2934 header line is locked and is inaccessible.
2936 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2937 co-exist for some time) to make it clear that it is the Exim ID that is
2938 referenced, not the Message-ID: header line.
2940 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2941 string_format() function, because snprintf() does not exist on all
2944 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2945 ${stat: expansion item.
2947 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2948 protocol synchronization error", to keep the pedants happy.
2950 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2951 well as for IRIX systems, when gcc is being used. See the host.c source
2954 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2956 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2959 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2960 transport and to the smtp transport in LMTP mode.
2962 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2964 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2965 run a filter in a subprocess. This could lead to confusion in subsequent
2966 lookups in the parent process. There should also be a search_tidyup() at
2967 the end of the subprocess.
2969 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2970 only if the host matched helo_try_verify_hosts, which caused the
2971 verification to occur when the EHLO/HELO command was issued. The ACL just
2972 tested the remembered result. Now, if a previous verification attempt has
2973 not happened, "verify = helo" does it there and then.
2975 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2977 TK/03 Fix log output including CR from clamd.
2979 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2980 exist provoked a memory error which could cause a segfault.
2982 PH/15 Installed PCRE 6.2
2984 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2986 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2987 of the problem. Specifically, suggested +O2 rather than +O1 for the
2990 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2992 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2993 record, the port was not correctly logged when the outgoing_port log
2994 selector was set (it logged the transort's default port).
2996 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2997 fallback_hosts, and "hosts" in the smtp transport.
2999 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3000 host errors such as "Connection refused".
3002 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3003 authentication with radiusclient 0.4.9:
3005 - Error returned from rc_read_config was caught wrongly
3006 - Username/password not passed on to radius server due to wrong length.
3008 The presumption is that some radiusclient API changes for 4.51/PH/17
3009 were not taken care of correctly. The code is still untested by me (my
3010 Linux distribution still has 0.3.2 of radiusclient), but it was
3011 contributed by a Radius user.
3013 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3014 expanding the "port" option of the smtp transport.
3016 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3017 while reading a MIME header. Thanks to Tom Hughes for a patch.
3019 PH/24 Include config.h inside local_scan.h so that configuration settings are
3022 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3023 that in an ACL for RCPT (for example), you can examine exactly what was
3026 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3027 commands, but it was not correctly comparing the address with the actual
3028 client host address. Thus, it would show the EHLO address in Received:
3029 header lines when this was not necessary.
3031 PH/27 Added the % operator to ${eval:}.
3033 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3034 it should be ignoring failures (because with -C, for example, it has lost
3035 privilege). It wasn't ignoring creation failures other than "already
3038 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3041 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3042 because an input line was too long, either on its own, or by virtue of
3043 too many continuations, the temporary file was not being removed, and the
3044 return code was incorrect.
3046 PH/31 Missing "BOOL" in function definition in filtertest.c.
3048 PH/32 Applied Sieve patches from the maintainer.
3050 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3052 PH/33 Added "verify = not_blind".
3054 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3055 Local/Makefile (with some defaults set). These are used in built scripts
3056 such as exicyclog, but they have never been used in the exim_install
3057 script (though there are many overriding facilities there). I have
3058 arranged that the exim_install script now takes note of these two
3061 PH/35 Installed configuration files for Dragonfly.
3063 PH/36 When a locally submitted message by a trusted user did not contain a
3064 From: header, and the sender address was obtained from -f or from an SMTP
3065 MAIL command, and the trusted user did not use -F to supply a sender
3066 name, $originator_name was incorrectly used when constructing a From:
3067 header. Furthermore, $originator_name was used for submission mode
3068 messages from external hosts without From: headers in a similar way,
3069 which is clearly wrong.
3071 PH/37 Added control=suppress_local_fixups.
3073 PH/38 When log_selector = +received_sender was set, and the addition of the
3074 sender made the log line's construction buffer exactly full, or one byte
3075 less than full, an overflow happened when the terminating "\n" was
3078 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3079 when the result of a list match is failure because a DNS lookup failed.
3081 PH/40 RM_COMMAND is now used in the building process.
3083 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3084 the "build-* directories that it finds.
3086 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3087 address in a domain literal was a prefix of an interface address.
3089 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3090 when verifying a sender address, unless rewrite_headers is false.
3092 PH/44 Wrote a long comment about why errors_to addresses are verified as
3093 recipients, not senders.
3095 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3096 the ratelimit ACL was added.
3098 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3100 PH/46 Added extra information about PostgreSQL errors to the error string.
3102 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3103 avoiding the need to install special zones in a real server. This is
3104 backwards compatible; if it can't find the fake resolver, it drops back.
3105 Thus, both old and new test suites can be run.
3107 TF/02 Added util/ratelimit.pl
3109 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3110 clock is set back in time.
3112 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3113 Candler <B.Candler@pobox.com>.
3115 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3116 OK for addresses that are the result of redirections.
3118 PH/48 A number of further additions for the benefit of the new test suite,
3119 including a fake gethostbyname() that interfaces to the fake DNS resolver
3122 TF/06 The fix for widen_domains has also been applied to qualify_single and
3123 search_parents which are the other dnslookup options that can cause
3126 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3129 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3131 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3132 service" if it encountered an MX record with an empty target hostname.
3133 The message is now "an MX or SRV record indicated no SMTP service".
3135 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3136 if the DNS of the sending site is misconfigured. This is quite a
3137 common situation. This change restores the behaviour of treating a
3138 helo verification defer as a failure.
3140 PH/51 If self=fail was set on a router, the bounce message did not include the
3141 actual error message.
3147 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3149 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3150 command itself ended in error, the underlying message about the transport
3151 filter timeout was being overwritten with the pipe command error. Now the
3152 underlying error message should be appended to the second error message.
3154 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3156 PH/02 Reduce the amount of output that "make" produces by default. Full output
3157 can still be requested.
3159 PH/03 The warning log line about a condition test deferring for a "warn" verb
3160 was being output only once per connection, rather than after each
3161 occurrence (because it was using the same function as for successful
3162 "warn" verbs). This seems wrong, so I have changed it.
3164 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3165 it should not have, which might have caused a crash in the right
3166 circumstances, but probably never did.
3168 PH/04 Installed a modified version of Tony Finch's patch to make submission
3169 mode fix the return path as well as the Sender: header line, and to
3170 add a /name= option so that you can make the user's friendly name appear
3173 TF/03 Added the control = fakedefer ACL modifier.
3175 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3176 Mark Lowes for thorough testing.
3178 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3180 TK/03 Merged latest SRS patch from Miles Wilton.
3182 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3183 with the definition in sysexits.h (which is #included earlier).
3184 Fortunately, Exim does not actually use EX_OK. The code used to try to
3185 preserve the sysexits.h value, by assumimg that macro definitions were
3186 scanned for macro replacements. I have been disabused of this notion,
3187 so now the code just undefines EX_OK before #including unistd.h.
3189 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3190 in the smtp transport. When a block could not be written in a single
3191 write() function, the timeout was being re-applied to each part-write.
3192 This seems wrong - if the receiver was accepting one byte at a time it
3193 would take for ever. The timeout is now adjusted when this happens. It
3194 doesn't have to be particularly precise.
3196 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3197 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3199 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3200 without a domain if the check to <postmaster@domain> fails.
3202 SC/01 Eximstats: added -xls and the ability to specify output files
3203 (patch written by Frank Heydlauf).
3205 SC/02 Eximstats: use FileHandles for outputing results.
3207 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3209 SC/04 Eximstats: fixed display of large numbers with -nvr option
3211 SC/05 Eximstats: fixed merging of reports with empty tables.
3213 SC/06 Eximstats: added the -include_original_destination flag
3215 SC/07 Eximstats: removed tabs and trailing whitespace.
3217 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3219 TK/06 MBOX spool code: Add real "From " MBOX separator line
3220 so the .eml file is really in mbox format (even though
3221 most programs do not really care). Patch from Alex Miller.
3223 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3224 The latter is generated from $received_to and is only set if the
3225 message has one envelope recipient. SA can use these headers,
3226 obviously out-of-the-box. Patch from Alex Miller.
3228 PH/08 The ${def test on a variable was returning false if the variable's
3229 value was "0", contrary to what the specification has always said!
3230 The result should be true unless the variable is empty.
3232 PH/09 The syntax error of a character other than { following "${if
3233 def:variable_name" (after optional whitespace) was not being diagnosed.
3234 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3235 accidental colon was present, for example, could give incorrect results.
3237 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3238 result is used (not including appendfile, where other changes are about
3241 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3242 This involved changing a lot of size variables from int to off_t. It
3243 should work with maildirs and everything.
3245 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3246 spamd dying while we are connected to it.
3248 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3251 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3252 function for BDB 4.3.
3254 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3256 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3257 the sources for PCRE have been re-arranged and more files are now
3260 PH/15 The code I had for printing potentially long long variables in PH/11
3261 above was not the best (it lost precision). The length of off_t variables
3262 is now inspected at build time, and an appropriate printing format (%ld
3263 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3264 to be "long long int" or "long int". This is needed for the internal
3265 formatting function string_vformat().
3267 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3268 the configuration file to be ":syslog", then the script "guesses" where
3269 the logs files are, rather than using the compiled in default. In our
3270 case the guess is not the same as the compiled default, so the script
3271 suddenly stopped working when I started to use syslog. The patch checks
3272 to see if log_file_path is "". If so, it attempts to read it from exim
3273 with no configuration file to get the compiled in version, before it
3274 falls back to the previous guessing code."
3276 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3277 implementing BATV in an Exim configuration. See NewStuff for the gory
3280 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3281 Makefile that are specific to HP-UX.
3283 PH/18 If the "use_postmaster" option was set for a recipient callout together
3284 with the "random" option, the postmaster address was used as the MAIL
3285 FROM address for the random test, but not for the subsequent recipient
3286 test. It is now used for both.
3288 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3289 patch removes a few documentation additions to RFC 3028, because the
3290 latest draft now contains them. It adds the new en;ascii-case comparator
3291 and a new error check for 8bit text in MIME parts. Comparator and
3292 require names are now matched exactly. I enabled the subaddress
3293 extension, but it is not well tested yet (read: it works for me)."
3295 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3296 rework some of the code of TK/09 above to avoid the hardwired use of
3297 "%lld" and "long long". Replaced the call to snprintf() with a call to
3300 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3301 records point to non-existent hosts", "retry timeout exceeded", and
3302 "retry time not reached for any host after a long failure period".
3304 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3305 experimental DomainKeys support:
3307 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3308 (2) On an error such as an illegally used "control", the wrong name for
3309 the control was given.
3311 These problems did NOT occur unless DomainKeys support was compiled.
3313 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3315 PH/24 Added ${if match_ip condition.
3317 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3318 databases so that it will be absolutely obvious if a crash occurs in the
3319 DB library. This is a regular occurrence (often caused by mis-matched
3322 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3323 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3324 picked up on a user's system that detects such things. There doesn't seem
3325 to be a gcc warning option for this - only an attribute that has to be
3326 put on the function's prototype. It seems that in Fedora Core 4 they have
3327 set this on a number of new functions. No doubt there will be more in due
3330 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3331 specify a transport. However, if an address that was verified by such a
3332 router was the subject of a callout, Exim crashed because it tried to
3333 read the rcpt_include_affixes from the non-existent transport. Now it
3334 just assumes that the setting of that option is false. This bug was
3335 introduced by 4.51/PH/31.
3337 PH/28 Changed -d+all to exclude +memory, because that information is very
3338 rarely of interest, but it makes the output a lot bigger. People tend to
3339 do -d+all out of habit.
3341 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3342 code in os-type was giving problems when libc.so lives in lib64, like on
3345 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3346 aren't the modern standard, and it seems that some systems' include files
3347 don't always have them. Exim was already checking for some of the newer
3348 ones like T_AAAA, and defining it itself. I've added checks for all the
3349 record types that Exim uses.
3351 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3352 not automatically generating a new one, as it is supposed to. This
3353 prevented TLS from working. If the file did exist, but contained invalid
3354 data, a new version was generated, as expected. It was only the case of a
3355 non-existent file that was broken.
3357 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3358 with a change in libdomainkeys > 0.64.
3360 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3361 from DNS. If the selector record carries the flag, it now has
3362 precedence over the domain-wide flag.
3364 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3366 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3367 the use of a port name that isn't defined in /etc/services) occurred, the
3368 message was deferred as in a normal delivery, and thus remained on the
3369 spool, instead of being failed because of the mua_wrapper setting. This
3370 is now fixed, and I tidied up some of the mua_wrapper messages at the
3373 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3374 the messages in a hash of arrays rather than using individual hashes.
3375 This is a bit cleaner and results in dramatic memory savings, albeit
3376 at a slight CPU cost.
3378 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3379 as requested by Marc Sherman.
3381 SC/10 Eximstats: added histograms for user specified patterns as requested
3384 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3386 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3387 fopen() in the content-scanning modules that did not already have it.
3393 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3394 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3396 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3398 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3400 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3401 to test Sieve filters that use "vacation".
3403 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3404 that changes the way the GnuTLS parameters are stored in the cache file.
3405 The new format can be generated externally. For backward compatibility,
3406 if the data in the cache doesn't make sense, Exim assumes it has read an
3407 old-format file, and it generates new data and writes a new file. This
3408 means that you can't go back to an older release without removing the
3411 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3412 work if there are any delivery delays because "one_time" forces the
3413 parent to be marked "delivered", so its unseen clone is never tried
3414 again. For this reason, Exim now forbids the simultaneous setting of
3417 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3418 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3419 entry if you want to know the details. The fix, however, neglected to
3420 consider the case when local delivery batching is involved. The test for
3421 "previously delivered" was not happening when checking to see if an
3422 address could be batched with a previous (undelivered) one; under
3423 certain circumstances this could lead to multiple deliveries to the same
3426 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3427 in its include files, and this causes problems building Exim.
3429 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3430 header_syntax) but Exim was just ignoring anything given after a slash.
3431 In particular, this caused confusion with an attempt to use "verify =
3432 reverse_host_lookup/defer_ok". An error is now given when options are
3433 supplied for verify items that do not have them. (Maybe reverse_host_
3434 lookup should have a defer_ok option, but that's a different point.)
3436 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3437 defined by RFC 821) to 2048, because there were problems with some AUTH
3438 commands, and RFC 1869 says the size should be increased for extended
3439 SMTP commands that take arguments.
3441 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3444 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3445 "unknown" error; now it says that the functionality isn't in the binary.
3447 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3448 an address' error message when a string expansion fails (syntax or
3449 whatever). Otherwise the password may appear in the log. Following change
3450 PH/42 below, there is no longer a chance of it appearing in a bounce
3453 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3455 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3456 of its addresses was ever tried. (Bugzilla bug #2.)
3458 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3459 the result incorrectly in the debug output. (It correctly added a newline
3460 to what was transported.)
3462 TF/01 Added $received_time.
3464 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3465 commented out examples of how to interface to a virus scanner and to
3466 SpamAssassin. Also added commented examples of av_scanner and
3467 spamd_address settings.
3469 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3470 and controls are allowed in which ACLs. There were a couple of minor
3471 errors. Some of the entries in the conditions table (which is a table of
3472 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3473 negation of where the condition IS allowed.
3475 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3477 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3478 header file does not have a version number, so I've had to invent a new
3479 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3480 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3481 radiusclient), but it was contributed by a Radius user.
3483 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3484 files or directories, for OpenSSL.
3486 PH/19 When an Exim process that is running as root has to create an Exim log
3487 file, it does so in a subprocess that runs as exim:exim so as to get the
3488 ownership right at creation (otherwise, other Exim processes might see
3489 the file with the wrong ownership). There was no test for failure of this
3490 fork() call, which would lead to the process getting stuck as it waited
3491 for a non-existent subprocess. Forks do occasionally fail when resources
3492 run out. I reviewed all the other calls to fork(); they all seem to check
3495 PH/20 When checking for unexpected SMTP input at connect time (before writing
3496 the banner), Exim was not dealing correctly with a non-positive return
3497 from the read() function. If the client had disconnected by this time,
3498 the result was a log entry for a synchronization error with an empty
3499 string after "input=" when read() returned zero. If read() returned -1
3500 (an event I could not check), uninitialized data bytes were printed.
3501 There were reports of junk text (parts of files, etc) appearing after
3504 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3506 PH/22 Added support for macro redefinition, and (re)definition in between
3507 driver and ACL definitions.
3509 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3510 forgetting to use the resulting value; it was using the unexpanded value.
3512 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3513 hadn't been configured. The fix is from Juergen Kreileder, who
3514 understands it better than I do:
3516 "Here's what I see happening with three configured cyrus_sasl
3517 authenticators configured (plain, login, cram-md5):
3519 On startup auth_cyrus_sasl_init() gets called for each of these.
3520 This means three calls to sasl_listmech() without a specified mech_list.
3521 => SASL tests which mechs of all available mechs actually work
3522 => three warnings about OTP not working
3523 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3525 With the patch, sasl_listmech() also gets called three times. But now
3526 SASL's mech_list option is set to the server_mech specified in the the
3527 authenticator. Or in other words, the answer from sasl_listmech()
3528 gets limited to just the mech you're testing for (which is different
3530 => the return list contains just 'plain' or 'login', 'cram-md5' or
3531 nothing depending on the value of ob->server_mech.
3533 I've just tested the patch: Authentication still works fine,
3534 unavailable mechs specified in the exim configuration are still
3535 caught, and the auth.log warnings about OTP are gone."
3537 PH/25 When debugging is enabled, the contents of the command line are added
3538 to the debugging output, even when log_selector=+arguments is not
3541 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3542 answer is "GNU", and only if the return is "GNU/something" is the answer
3545 PH/27 $acl_verify_message is now set immediately after the failure of a
3546 verification in an ACL, and so is available in subsequent modifiers. In
3547 particular, the message can be preserved by coding like this:
3549 warn !verify = sender
3550 set acl_m0 = $acl_verify_message
3552 Previously, $acl_verify_message was set only while expanding "message"
3553 and "log_message" when a very denied access.
3555 PH/28 Modified OS/os.c-Linux with
3557 -#ifndef OS_LOAD_AVERAGE
3558 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3560 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3561 nomenclature these days.)
3563 PH/29 Installed patch from the Sieve maintainer that adds the options
3564 sieve_useraddress and sieve_subaddress to the redirect router.
3566 PH/30 In these circumstances:
3567 . Two addresses routed to the same list of hosts;
3568 . First host does not offer TLS;
3569 . First host accepts first address;
3570 . First host gives temporary error to second address;
3571 . Second host offers TLS and a TLS session is established;
3572 . Second host accepts second address.
3573 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3574 and peerdn, if requested) that were in fact used only for the second
3577 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3578 attention to any local part prefix or suffix that was matched by the
3579 router that accepted the address. It now behaves in the same way as it
3580 does for delivery: the affixes are removed from the local part unless
3581 rcpt_include_affixes is set on the transport.
3583 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3584 timeout during the DATA phase of an incoming message.
3586 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3587 applied a patch sanctioned by the Sieve maintainer.
3589 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3590 the uid or gid is negative. A case of a negative gid caused this to be
3591 noticed. The fix allows for either to be negative.
3593 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3594 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3597 PH/36 The change PH/12 above was broken. Fixed it.
3599 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3600 the grounds that routing the same address twice would always produce the
3601 same answer. This might have been true once, but it is certainly no
3602 longer true now. Routing a child address may depend on the previous
3603 routing that produced that child. Some complicated redirection strategies
3604 went wrong when messages had multiple recipients, and made Exim's
3605 behaviour dependent on the order in which the addresses were given.
3607 I have moved the duplicate checking until after the routing is complete.
3608 Exim scans the addresses that are assigned to local and remote
3609 transports, and removes any duplicates. This means that more work will be
3610 done, as duplicates will always all be routed, but duplicates are
3611 presumably rare, so I don't expect this is of any significance.
3613 For deliveries to pipes, files, and autoreplies, the duplicate checking
3614 still happens during the routing process, since they are not going to be
3617 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3618 It corrects a timeout issue with spamd. This is Ian's comment: "The
3619 background is that sometimes spamd either never reads data from a
3620 connection it has accepted, or it never writes response data. The exiscan
3621 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3622 blindly assumes that writes won't block so it may never time out."
3624 PH/39 Allow G after quota size as well as K and M.
3626 PH/40 The value set for $authenticated_id in an authenticator may not contain
3627 binary zeroes or newlines because the value is written to log lines and
3628 to spool files. There was no check on this. Now the value is run through
3629 the string_printing() function so that such characters are converted to
3630 printable escape sequences.
3632 PH/41 $message_linecount is a new variable that contains the total number of
3633 lines in the message. Compare $body_linecount, which is the count for the
3636 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3637 bounce and delay warning messages, except in certain special cases, which
3640 (a) An SMTP error message from a remote host;
3641 (b) A message specified in a :fail: redirection;
3642 (c) A message specified in a "fail" command in a system filter;
3643 (d) A message specified in a FAIL return from the queryprogram router;
3644 (e) A message specified by the cannot_route_message router option.
3646 In these cases only, Exim does include the error details in bounce and
3647 warning messages. There are also a few cases where bland messages such
3648 as "unrouteable address" or "local delivery error" are given.
3650 PH/43 $value is now also set for the "else" part of a ${run expansion.
3652 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3653 being worked on, but at least Exim now implements the latest version to
3656 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3657 process to complete was treated as a delivery failure, a timeout while
3658 writing the message to the pipe was logged, but erroneously treated as a
3659 successful delivery. Such timeouts include transport filter timeouts. For
3660 consistency with the overall process timeout, these timeouts are now
3661 treated as errors, giving rise to delivery failures by default. However,
3662 there is now a new Boolean option for the pipe transport called
3663 timeout_defer, which, if set TRUE, converts the failures into defers for
3664 both kinds of timeout. A transport filter timeout is now identified in
3667 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3668 systems where "make" and "gmake" are different, calling "gmake" at top
3669 level broke things. I've arranged for the value of $(MAKE) to be passed
3670 from the Makefile to this script so that it can call the same version of
3674 A note about Exim versions 4.44 and 4.50
3675 ----------------------------------------
3677 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3678 changes of various kinds. As a consequence, a big documentation update was
3679 needed. This delayed the release for rather longer than seemed good, especially
3680 in the light of a couple of (minor) security issues. Therefore, the changes
3681 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3682 release. So 4.44 and 4.50 are in effect two different branches that both start
3685 I have left the 4.50 change log unchanged; it contains all the changes since
3686 4.43. The change log for 4.44 is below; many of its items are identical to
3687 those for 4.50. This seems to be the most sensible way to preserve the
3688 historical information.
3694 1. Minor wording change to the doc/README.SIEVE file.
3696 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3697 computation of the current number of files was incorrect.
3699 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3700 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3703 4. Give more explanation in the error message when the command for a transport
3704 filter fails to execute.
3706 5. There are several places where Exim runs a non-Exim command in a
3707 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3708 was being done only for the command run by the queryprogram router. It is
3709 now done for all such subprocesses. The other cases are: ${run, transport
3710 filters, and the commands run by the lmtp and pipe transports.
3712 6. Added CONFIGURE_GROUP build-time option.
3714 7. Some older OS have a limit of 256 on the maximum number of file
3715 descriptors. Exim was using setrlimit() to set 1000 as a large value
3716 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3717 systems. I've change it so that if it can't get 1000, it tries for 256.
3719 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3720 was an oversight, and furthermore, ever since the addition of extra
3721 controls (e.g. 4.43/32), the checks on when to allow different forms of
3722 "control" were broken. There should now be diagnostics for all cases when a
3723 control that does not make sense is encountered.
3725 9. Added the /retain_sender option to "control=submission".
3727 10. $recipients is now available in the predata ACL (oversight).
3729 11. Tidy the search cache before the fork to do a delivery from a message
3730 received from the command line. Otherwise the child will trigger a lookup
3731 failure and thereby defer the delivery if it tries to use (for example) a
3732 cached ldap connection that the parent has called unbind on.
3734 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3735 of $address_data from the recipient verification was clobbered by the
3736 sender verification.
3738 13. The value of address_data from a sender verification is now available in
3739 $sender_address_data in subsequent conditions in the ACL statement.
3741 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3743 15. Added a new option "connect=<time>" to callout options, to set a different
3746 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3747 was its contents. (It was OK if the option was not defined at all.)
3749 17. A "Completed" log line is now written for messages that are removed from
3750 the spool by the -Mrm option.
3752 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3753 information about exactly what failed.
3755 19. Added -dd to debug only the daemon process.
3757 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3758 handles timeouts, both on the server side and network timeouts. Renamed the
3759 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3761 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3762 It is now set to "smtps".
3764 22. $host_address is now set to the target address during the checking of
3765 ignore_target_hosts.
3767 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3768 being passed; this would have caused $sender_host_name to have been used if
3769 matching the list had actually called for a host name (not very likely,
3770 since this list is usually IP addresses). A host name is now passed as
3773 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3774 code that specifies a non-restarting handler (typically sigaction() in
3775 modern systems) in an attempt to fix a rare and obscure crash bug.
3777 25. Narrowed the window for a race in the daemon that could cause it to ignore
3778 SIGCHLD signals. This is not a major problem, because they are used only to
3779 wake it up if nothing else does.
3781 26. A malformed maildirsize file could cause Exim to calculate negative values
3782 for the mailbox size or file count. Odd effects could occur as a result.
3783 The maildirsize information is now recalculated if the size or filecount
3786 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3787 support for a long time. Removed HAVE_SYS_VFS_H.
3789 28. Installed the latest version of exipick from John Jetmore.
3791 29. In an address list, if the pattern was not a regular expression, an empty
3792 subject address (from a bounce message) matched only if the pattern was an
3793 empty string. Non-empty patterns were not even tested. This was the wrong
3794 because it is perfectly reasonable to use an empty address as part of a
3795 database query. An empty address is now tested by patterns that are
3796 lookups. However, all the other forms of pattern expect the subject to
3797 contain a local part and a domain, and therefore, for them, an empty
3798 address still always fails if the pattern is not itself empty.
3800 30. Exim went into a mad DNS loop when attempting to do a callout where the
3801 host was specified on an smtp transport, and looking it up yielded more
3802 than one IP address.
3804 31. Re-factored the code for checking spool and log partition space into a
3805 function that finds that data and another that does the check. The former
3806 is then used to implement four new variables: $spool_space, $log_space,
3807 $spool_inodes, and $log_inodes.
3809 32. The RFC2047 encoding function was originally intended for short strings
3810 such as real names; it was not keeping to the 75-character limit for
3811 encoded words that the RFC imposes. It now respects the limit, and
3812 generates multiple encoded words if necessary. To be on the safe side, I
3813 have increased the buffer size for the ${rfc2047: expansion operator from
3816 33. It is now permitted to omit both strings after an "if" condition; if the
3817 condition is true, the result is "true". As before, when the second string
3818 is omitted, a false condition yields an empty string. This makes it less
3819 cumbersome to write custom ACL and router conditions.
3821 34. Failure to deliver a bounce message always caused it to be frozen, even if
3822 there was an errors_to setting on the router. The errors_to setting is now
3825 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3826 canonical form (fully expanded) before being placed in
3827 $sender_host_address.
3829 36. The table in the code that translates DNS record types into text (T_A to
3830 "A" for instance) was missing entries for NS and CNAME. It is just possible
3831 that this could have caused confusion if both these types were looked up
3832 for the same domain, because the text type is used as part of Exim's
3833 per-process caching. But the chance of anyone hitting this buglet seems
3836 37. The dnsdb lookup has been extended in a number of ways.
3838 (1) There is a new type, "zns", which walks up the domain tree until it
3839 finds some nameserver records. It should be used with care.
3841 (2) There is a new type, "mxh", which is like "mx" except that it returns
3842 just the host names, not the priorities.
3844 (3) It is now possible to give a list of domains (or IP addresses) to be
3845 looked up. The behaviour when one of the lookups defers can be
3846 controlled by a keyword.
3848 (4) It is now possible to specify the separator character for use when
3849 multiple records are returned.
3851 38. The dnslists ACL condition has been extended: it is now possible to supply
3852 a list of IP addresses and/or domains to be looked up in a particular DNS
3855 39. Added log_selector=+queue_time_overall.
3857 40. When running the queue in the test harness, wait just a tad after forking a
3858 delivery process, to get repeatability of debugging output.
3860 41. Include certificate and key file names in error message when GnuTLS fails
3861 to set them up, because the GnuTLS error message doesn't include the name
3862 of the failing file when there is a problem reading it.
3864 42. Allow both -bf and -bF in the same test run.
3866 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3868 44. The "Exiscan patch" is now merged into the mainline Exim source.
3870 45. Sometimes the final signoff response after QUIT could fail to get
3871 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3872 < 0 before doing a fflush(). This bug looks as though it goes back to the
3873 introduction of TLS in release 3.20, but "sometimes" must have been rare
3874 because the tests only now provoked it.
3876 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3877 (this can affect the format of dates).
3879 47. exim_tidydb, when checking for the continued existence of a message for
3880 which it has found a message-specific retry record, was not finding
3881 messages that were in split spool directories. Consequently, it was
3882 deleting retry records that should have stayed in existence.
3884 48. Steve fixed some bugs in eximstats.
3886 49. The SPA authentication driver was not abandoning authentication and moving
3887 on to the next authenticator when an expansion was forced to fail,
3888 contradicting the general specification for all authenticators. Instead it
3889 was generating a temporary error. It now behaves as specified.
3891 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3892 (the order specifies the preference for clients). The order is now AES256,
3893 AES128, 3DES, ARCFOUR128.
3895 51. Small patch to Sieve code - explicitly set From: when generating an
3898 52. Exim crashed if a remote delivery caused a very long error message to be
3899 recorded - for instance if somebody sent an entire SpamAssassin report back
3900 as a large number of 550 error lines. This bug was coincidentally fixed by
3901 increasing the size of one of Exim's internal buffers (big_buffer) that
3902 happened as part of the Exiscan merge. However, to be on the safe side, I
3903 have made the code more robust (and fixed the comments that describe what
3906 53. Now that there can be additional text after "Completed" in log lines (if
3907 the queue_time_overall log selector is set), a one-byte patch to exigrep
3908 was needed to allow it to recognize "Completed" as not the last thing in
3911 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3912 patch that reportedly fixes this has been added. I am not expert enough to
3913 create a test for it. This is what the patch creator wrote:
3915 "I found a little strange behaviour of ldap code when working with
3916 Windows 2003 AD Domain, where users was placed in more than one
3917 Organization Units. When I tried to give exim partial DN, the exit code
3918 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3919 But simultaneously result of request was absolutely normal ldap result,
3920 so I produce this patch..."
3922 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3923 REFERENCE, so I have modified the code to exclude the patch when that macro
3926 55. Some experimental protocols are using DNS PTR records for new purposes. The
3927 keys for these records are domain names, not reversed IP addresses. The
3928 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3929 leaves it alone. Component reversal etc. now happens only for IP addresses.
3932 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3934 57. Double the size of the debug message buffer (to 2048) so that more of very
3935 long debug lines gets shown.
3937 58. The exicyclog utility now does better if the number of log files to keep
3938 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3940 59. Two changes related to the smtp_active_hostname option:
3942 (1) $smtp_active_hostname is now available as a variable.
3943 (2) The default for smtp_banner uses $smtp_active_hostname instead
3944 of $primary_hostname.
3946 60. The host_aton() function is supposed to be passed a string that is known
3947 to be a valid IP address. However, in the case of IPv6 addresses, it was
3948 not checking this. This is a hostage to fortune. Exim now panics and dies
3949 if the condition is not met. A case was found where this could be provoked
3950 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3951 components; fortuitously, this particular loophole had already been fixed
3952 by change 4.50/55 above.
3954 If there are any other similar loopholes, the new check in host_aton()
3955 itself should stop them being exploited. The report I received stated that
3956 data on the command line could provoke the exploit when Exim was running as
3957 exim, but did not say which command line option was involved. All I could
3958 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3959 running as the user.
3962 61. There was a buffer overflow vulnerability in the SPA authentication code
3963 (which came originally from the Samba project). I have added a test to the
3964 spa_base64_to_bits() function which I hope fixes it.
3967 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3968 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3970 63. The daemon start-up calls getloadavg() while still root for those OS that
3971 need the first call to be done as root, but it missed one case: when
3972 deliver_queue_load_max is set with deliver_drop_privilege. This is
3973 necessary for the benefit of the queue runner, because there is no re-exec
3974 when deliver_drop_privilege is set.
3976 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3977 This has been fixed.
3979 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3980 was in use, was not putting the data itself into the right store pool;
3981 consequently, it could be overwritten for a subsequent message in the same
3982 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3985 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3987 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3988 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3989 the function were treating the return as a boolean value, which happened to
3990 work because 0=false and not-0=true, but is not correct code.
3992 68. The host_aton() function was not handling scoped IPv6 addresses (those
3993 with, for example, "%eth0" on the end) correctly.
3995 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3996 negated items (that is, ~something) in unsigned ints. Some compilers
3997 apparently mutter when there is no cast.
3999 70. If an address verification called from an ACL failed, and did not produce a
4000 user-specific message (i.e. there was only a "system" message), nothing was
4001 put in $acl_verify_message. In this situation, it now puts the system
4004 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4005 session; change 4.31/43 added the unwanted input to the log line - except
4006 that it did not do this in the start of session case. It now does.
4008 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4009 This is wrong and can cause the other end to generate a synchronization
4010 error if it is another Exim or anything else that does the synchronization
4011 check. A QUIT command is no longer sent after a timeout.
4013 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4014 during host lookups.
4016 74. The defer_ok option of callout verification was not working if it was used
4017 when verifying addresses in header lines, that is, for this case:
4019 verify = header_sender/callout=defer_ok
4021 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4022 those file descriptors could be used for SMTP connections. If anything
4023 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4024 could be sent to the SMTP client, causing chaos. The daemon now opens
4025 stdin, stdout, and stderr to /dev/null when it puts itself into the
4028 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4029 log by default. The user can override this with suitable Perl magic.
4031 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4032 the log message when discard triggers, was not working for the DATA ACL or
4033 for the non-SMTP ACL.
4035 78. Error message wording change in sieve.c.
4037 79. If smtp_accept_max_per_host was set, the number of connections could be
4038 restricted to fewer than expected, because the daemon was trying to set up
4039 a new connection before checking whether the processes handling previous
4040 connections had finished. The check for completed processes is now done
4041 earlier. On busy systems, this bug wouldn't be noticed because something
4042 else would have woken the daemon, and it would have reaped the completed
4045 80. If a message was submitted locally by a user whose login name contained one
4046 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4047 It caused a spool format error. I have fixed the spool reading code. A
4048 related problem was that the "from" clause in the Received: line became
4049 illegal because of the space(s). It is now covered by ${quote_local_part.
4051 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4054 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4056 83. Patch from Sieve maintainer to fix unterminated string problem in
4057 "vacation" handling.
4059 84. Some minor changes to the Linux configuration files to help with other
4060 OS variants using glibc.
4062 85. One more patch for Sieve to update vacation handling to latest spec.
4065 ----------------------------------------------------
4066 See the note above about the 4.44 and 4.50 releases.
4067 ----------------------------------------------------
4073 1. Change 4.43/35 introduced a bug that caused file counts to be
4074 incorrectly computed when quota_filecount was set in an appendfile
4077 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4078 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4081 3. Give more explanation in the error message when the command for a transport
4082 filter fails to execute.
4084 4. There are several places where Exim runs a non-Exim command in a
4085 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4086 was being done only for the command run by the queryprogram router. It is
4087 now done for all such subprocesses. The other cases are: ${run, transport
4088 filters, and the commands run by the lmtp and pipe transports.
4090 5. Some older OS have a limit of 256 on the maximum number of file
4091 descriptors. Exim was using setrlimit() to set 1000 as a large value
4092 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4093 systems. I've change it so that if it can't get 1000, it tries for 256.
4095 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4096 was an oversight, and furthermore, ever since the addition of extra
4097 controls (e.g. 4.43/32), the checks on when to allow different forms of
4098 "control" were broken. There should now be diagnostics for all cases when a
4099 control that does not make sense is encountered.
4101 7. $recipients is now available in the predata ACL (oversight).
4103 8. Tidy the search cache before the fork to do a delivery from a message
4104 received from the command line. Otherwise the child will trigger a lookup
4105 failure and thereby defer the delivery if it tries to use (for example) a
4106 cached ldap connection that the parent has called unbind on.
4108 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4109 of $address_data from the recipient verification was clobbered by the
4110 sender verification.
4112 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4113 was its contents. (It was OK if the option was not defined at all.)
4115 11. A "Completed" log line is now written for messages that are removed from
4116 the spool by the -Mrm option.
4118 12. $host_address is now set to the target address during the checking of
4119 ignore_target_hosts.
4121 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4122 being passed; this would have caused $sender_host_name to have been used if
4123 matching the list had actually called for a host name (not very likely,
4124 since this list is usually IP addresses). A host name is now passed as
4127 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4128 code that specifies a non-restarting handler (typically sigaction() in
4129 modern systems) in an attempt to fix a rare and obscure crash bug.
4131 15. Narrowed the window for a race in the daemon that could cause it to ignore
4132 SIGCHLD signals. This is not a major problem, because they are used only to
4133 wake it up if nothing else does.
4135 16. A malformed maildirsize file could cause Exim to calculate negative values
4136 for the mailbox size or file count. Odd effects could occur as a result.
4137 The maildirsize information is now recalculated if the size or filecount
4140 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4141 support for a long time. Removed HAVE_SYS_VFS_H.
4143 18. Updated exipick to current release from John Jetmore.
4145 19. Allow an empty sender to be matched against a lookup in an address list.
4146 Previously the only cases considered were a regular expression, or an
4149 20. Exim went into a mad DNS lookup loop when doing a callout where the
4150 host was specified on the transport, if the DNS lookup yielded more than
4153 21. The RFC2047 encoding function was originally intended for short strings
4154 such as real names; it was not keeping to the 75-character limit for
4155 encoded words that the RFC imposes. It now respects the limit, and
4156 generates multiple encoded words if necessary. To be on the safe side, I
4157 have increased the buffer size for the ${rfc2047: expansion operator from
4160 22. Failure to deliver a bounce message always caused it to be frozen, even if
4161 there was an errors_to setting on the router. The errors_to setting is now
4164 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4165 canonical form (fully expanded) before being placed in
4166 $sender_host_address.
4168 24. Updated eximstats to version 1.33
4170 25. Include certificate and key file names in error message when GnuTLS fails
4171 to set them up, because the GnuTLS error message doesn't include the name
4172 of the failing file when there is a problem reading it.
4174 26. Expand error message when OpenSSL has problems setting up cert/key files.
4177 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4178 (this can affect the format of dates).
4180 28. exim_tidydb, when checking for the continued existence of a message for
4181 which it has found a message-specific retry record, was not finding
4182 messages that were in split spool directories. Consequently, it was
4183 deleting retry records that should have stayed in existence.
4185 29. eximstats updated to version 1.35
4186 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4187 1.35 - bugfix such that pie charts by volume are generated correctly
4189 30. The SPA authentication driver was not abandoning authentication and moving
4190 on to the next authenticator when an expansion was forced to fail,
4191 contradicting the general specification for all authenticators. Instead it
4192 was generating a temporary error. It now behaves as specified.
4194 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4195 (the order specifies the preference for clients). The order is now AES256,
4196 AES128, 3DES, ARCFOUR128.
4198 31. Small patch to Sieve code - explicitly set From: when generating an
4201 32. Exim crashed if a remote delivery caused a very long error message to be
4202 recorded - for instance if somebody sent an entire SpamAssassin report back
4203 as a large number of 550 error lines. This bug was coincidentally fixed by
4204 increasing the size of one of Exim's internal buffers (big_buffer) that
4205 happened as part of the Exiscan merge. However, to be on the safe side, I
4206 have made the code more robust (and fixed the comments that describe what
4209 33. Some experimental protocols are using DNS PTR records for new purposes. The
4210 keys for these records are domain names, not reversed IP addresses. The
4211 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4212 leaves it alone. Component reversal etc. now happens only for IP addresses.
4215 34. The host_aton() function is supposed to be passed a string that is known
4216 to be a valid IP address. However, in the case of IPv6 addresses, it was
4217 not checking this. This is a hostage to fortune. Exim now panics and dies
4218 if the condition is not met. A case was found where this could be provoked
4219 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4220 components; fortuitously, this particular loophole had already been fixed
4221 by change 4.50/55 or 4.44/33 above.
4223 If there are any other similar loopholes, the new check in host_aton()
4224 itself should stop them being exploited. The report I received stated that
4225 data on the command line could provoke the exploit when Exim was running as
4226 exim, but did not say which command line option was involved. All I could
4227 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4228 running as the user.
4231 35. There was a buffer overflow vulnerability in the SPA authentication code
4232 (which came originally from the Samba project). I have added a test to the
4233 spa_base64_to_bits() function which I hope fixes it.
4236 36. The daemon start-up calls getloadavg() while still root for those OS that
4237 need the first call to be done as root, but it missed one case: when
4238 deliver_queue_load_max is set with deliver_drop_privilege. This is
4239 necessary for the benefit of the queue runner, because there is no re-exec
4240 when deliver_drop_privilege is set.
4242 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4243 was in use, was not putting the data itself into the right store pool;
4244 consequently, it could be overwritten for a subsequent message in the same
4245 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4248 38. Sometimes the final signoff response after QUIT could fail to get
4249 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4250 < 0 before doing a fflush(). This bug looks as though it goes back to the
4251 introduction of TLS in release 3.20, but "sometimes" must have been rare
4252 because the tests only now provoked it.
4258 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4259 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4260 Now it returns an integer. A number of calls were still expecting a T/F
4261 return. Fortuitously, in all cases, the tests worked in OK situations,
4262 which is the norm. However, things would have gone wrong on any write
4263 failures on the smtp file descriptor. This function is used when sending
4264 messages over SMTP and also when doing verify callouts.
4266 2. When Exim is called to do synchronous delivery of a locally submitted
4267 message (the -odf or -odi options), it no longer closes stderr before doing
4270 3. Implemented the mua_wrapper option.
4272 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4274 5. Implemented the functions header_remove(), header_testname(),
4275 header_add_at_position(), and receive_remove_recipient(), and exported them
4278 6. If an ACL "warn" statement specified the addition of headers, Exim already
4279 inserted X-ACL-Warn: at the start if there was no header name. However, it
4280 was not making this test for the second and subsequent header lines if
4281 there were newlines in the string. This meant that an invalid header could
4282 be inserted if Exim was badly configured.
4284 7. Allow an ACL "warn" statement to add header lines at the start or after all
4285 the Received: headers, as well as at the end.
4287 8. Added the rcpt_4xx retry error code.
4289 9. Added postmaster_mailfrom=xxx to callout verification option.
4291 10. Added mailfrom=xxxx to the callout verification option, for verify=
4294 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4295 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4297 12. Inserted some casts to stop certain compilers warning when using pointer
4298 differences as field lengths or precisions in printf-type calls (mostly
4299 affecting debugging statements).
4301 13. Added optional readline() support for -be (dynamically loaded).
4303 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4304 same clock tick as a message's arrival, so that its received time was the
4305 same as the "first fail" time on the retry record, and that message
4306 remained on the queue past the ultimate address timeout, every queue runner
4307 would try a delivery (because it was past the ultimate address timeout) but
4308 after another failure, the ultimate address timeout, which should have then
4309 bounced the address, did not kick in. This was a "< instead of <=" error;
4310 in most cases the first failure would have been in the next clock tick
4311 after the received time, and all would be well.
4313 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4314 being recognized when the domain list was tested by the match_domain
4315 condition in an expansion string.
4317 16. Added the ${str2b64: operator.
4319 17. Exim was always calling setrlimit() to set a large limit for the number of
4320 processes, without checking whether the existing limit was already
4321 adequate. (It did check for the limit on file descriptors.) Furthermore,
4322 errors from getrlimit() and setrlimit() were being ignored. Now they are
4323 logged to the main and panic logs, but Exim does carry on, to try to do its
4324 job under whatever limits there are.
4326 18. Imported PCRE 5.0.
4328 19. Trivial typo in log message " temporarily refused connection" (the leading
4331 20. If the log selector return_path_on_delivery was set and an address was
4332 redirected to /dev/null, the delivery process crashed because it assumed
4333 that a return path would always be set for a "successful" delivery. In this
4334 case, the whole delivery is bypassed as an optimization, and therefore no
4337 21. Internal re-arrangement: the function for sending a challenge and reading
4338 a response while authentication was assuming a zero-terminated challenge
4339 string. It's now changed to take a pointer and a length, to allow for
4340 binary data in such strings.
4342 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4344 23. Exim was not respecting finduser_retries when seeking the login of the
4345 uid under which it was called; it was always trying 10 times. (The default
4346 setting of finduser_retries is zero.) Also, it was sleeping after the final
4347 failure, which is pointless.
4349 24. Implemented tls_on_connect_ports.
4351 25. Implemented acl_smtp_predata.
4353 26. If the domain in control=submission is set empty, Exim assumes that the
4354 authenticated id is a complete email address when it generates From: or
4355 Sender: header lines.
4357 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4358 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4359 chown and chgrp in /bin and hostname in /usr/bin.
4361 28. Exim was keeping the "process log" file open after each use, just as it
4362 does for the main log. This opens the possibility of it remaining open for
4363 long periods when the USR1 signal hits a daemon. Occasional processlog
4364 errors were reported, that could have been caused by this. Anyway, it seems
4365 much more sensible not to leave this file open at all, so that is what now
4368 29. The long-running daemon process does not normally write to the log once it
4369 has entered its main loop, and it closes the log before doing so. This is
4370 so that log files can straightforwardly be renamed and moved. However,
4371 there are a couple of unusual error situations where the daemon does write
4372 log entries, and I had neglected to close the log afterwards.
4374 30. The text of an SMTP error response that was received during a remote
4375 delivery was being truncated at 512 bytes. This is too short for some of
4376 the long messages that one sometimes sees. I've increased the limit to
4379 31. It is now possible to make retry rules that apply only when a message has a
4380 specific sender, in particular, an empty sender.
4382 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4383 it possible to be selective about when SMTP synchronization is enforced.
4385 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4387 32. Implemented hosts_connection_nolog.
4389 33. Added an ACL for QUIT.
4391 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4394 35. Added mailbox_size and mailbox_filecount to appendfile.
4396 36. Added control = no_multiline_responses to ACLs.
4398 37. There was a bug in the logic of the code that waits for the clock to tick
4399 in the case where the clock went backwards by a substantial amount such
4400 that the microsecond fraction of "now" was more than the microsecond
4401 fraction of "then" (but the whole seconds number was less).
4403 38. Added support for the libradius Radius client library this is found on
4404 FreeBSD (previously only the radiusclient library was supported).
4410 1. When certain lookups returned multiple values in the form name=value, the
4411 quoting of the values was not always being done properly. Specifically:
4412 (a) If the value started with a double quote, but contained no whitespace,
4414 (b) If the value contained whitespace other than a space character (i.e.
4415 tabs or newlines or carriage returns) it was not quoted.
4416 This fix has been applied to the mysql and pgsql lookups by writing a
4417 separate quoting function and calling it from the lookup code. The fix
4418 should probably also be applied to nisplus, ibase and oracle lookups, but
4419 since I cannot test any of those, I have not disturbed their existing code.
4421 2. A hit in the callout cache for a specific address caused a log line with no
4422 reason for rejecting RCPT. Now it says "Previous (cached) callout
4423 verification failure".
4425 3. There was an off-by-one bug in the queryprogram router. An over-long
4426 return line was truncated at 256 instead of 255 characters, thereby
4427 overflowing its buffer with the terminating zero. As well as fixing this, I
4428 have increased the buffer size to 1024 (and made a note to document this).
4430 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4431 when Exim is waiting for an SMTP response from a remote server, Exim
4432 restarts its select() call on the socket, thereby resetting its timeout.
4433 This is not a problem when such interrupts are rare. Somebody set up a cron
4434 job to run exiwhat every 2 minutes, which is less than the normal select()
4435 timeout (5 or 10 minutes). This meant that the select() timeout never
4436 kicked in because it was always reset. I have fixed this by comparing the
4437 time when an interrupt arrives with the time at the start of the first call
4438 to select(). If more time than the timeout has elapsed, the interrupt is
4439 treated as a timeout.
4441 5. Some internal re-factoring in preparation for the addition of Sieve
4442 extensions (by MH). In particular, the "personal" test is moved to a
4443 separate function, and given an option for scanning Cc: and Bcc: (which is
4444 not set for Exim filters).
4446 6. When Exim created an email address using the login of the caller as the
4447 local part (e.g. when creating a From: or Sender: header line), it was not
4448 quoting the local part when it contained special characters such as @.
4450 7. Installed new OpenBSD configuration files.
4452 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4453 try to make them clearer.
4455 9. Callout options, other than the timeout value, were being ignored when
4456 verifying sender addresses in header lines. For example, when using
4458 verify = header_sender/callout=no_cache
4460 the cache was (incorrectly) being used.
4462 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4463 only the Cygwin environment.
4465 11. When return_path_on_delivery was set as a log selector, if different remote
4466 addresses in the same message used different return paths and parallel
4467 remote delivery occurred, the wrong values would sometimes be logged.
4468 (Whenever a remote delivery process finished, the return path value from
4469 the most recently started remote delivery process was logged.)
4471 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4472 lines when AUTH and/or TLS are in use. This is the "received protocol"
4473 field. Exim used to use "asmtp" for authenticated SMTP, without any
4474 indication (in the protocol name) for TLS use. Now it follows the RFC and
4475 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4476 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4477 names appear in log lines as well as in Received: header lines.
4479 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4480 extensions, and comparison tests, and to fix some bugs.
4482 14. Changes to the "personal" filter test:
4484 (1) The test was buggy in that it was just doing the equivalent of
4485 "contains" tests on header lines. For example, if a user's address was
4486 anne@some.where, the "personal" test would incorrectly be true for
4488 To: susanne@some.where
4490 This test is now done by extracting each address from the header in turn,
4491 and checking the entire address. Other tests that are part of "personal"
4492 are now done using regular expressions (for example, to check local parts
4493 of addresses in From: header lines).
4495 (2) The list of non-personal local parts in From: addresses has been
4496 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4497 taken from the Sieve specification recommendations.
4499 (3) If the message contains any header line starting with "List-" it is
4500 treated as non-personal.
4502 (4) The test for "circular" in the Subject: header line has been removed
4503 because it now seems ill-conceived.
4505 15. Minor typos in src/EDITME comments corrected.
4507 16. Installed latest exipick from John Jetmore.
4509 17. If headers_add on a router specified a text string that was too long for
4510 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4511 of string_sprintf() is now avoided.
4513 18. $message_body_size was not set (it was always zero) when running the DATA
4514 ACL and the local_scan() function.
4516 19. For the "mail" command in an Exim filter, no default was being set for
4517 the once_repeat time, causing a random time value to be used if "once" was
4518 specified. (If the value happened to be <= 0, no repeat happened.) The
4519 default is now 0s, meaning "never repeat". The "vacation" command was OK
4520 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4521 (I found it when inspecting the code).
4523 20. There is now an overall timeout for performing a callout verification. It
4524 defaults to 4 times the callout timeout, which applies to individual SMTP
4525 commands during the callout. The overall timeout applies when there is more
4526 than one host that can be tried. The timeout is checked before trying the
4527 next host. This prevents very long delays if there are a large number of
4528 hosts and all are timing out (e.g. when the network connections are timing
4529 out). The value of the overall timeout can be changed by specifying an
4530 additional sub-option for "callout", called "maxwait". For example:
4532 verify = sender/callout=5s,maxwait=20s
4534 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4535 to the end before writing, but this should make it even safer).
4537 22. Exim was forgetting that it had advertised PIPELINING for the second and
4538 subsequent messages on an SMTP connection. It was also not resetting its
4539 memory on STARTTLS and an internal HELO.
4541 23. When Exim logs an SMTP synchronization error within a session, it now
4542 records whether PIPELINING has been advertised or not.
4544 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4545 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4546 rather than long int.
4548 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4550 26. Added the never_mail option to autoreply.
4556 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4557 crash if the getsockname() call failed; this can happen if a connection is
4558 closed very soon after it is established. The problem was simply in the
4559 order in which certain operations were done, causing Exim to try to write
4560 to the SMTP stream before it had set up the file descriptor. The bug has
4561 been fixed by making things happen in the correct order.
4567 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4568 before the connection was closed, thus losing the rejection response.
4570 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4571 some early Solaris releases, but causes trouble in current releases where
4572 socklen_t is defined.
4574 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4577 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4580 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4581 that releasing the top of it at the end releases what was used for sub-
4582 expansions (unless the block got too big). However, discard this block if
4583 the first thing is a variable or header, so that we can use its block when
4584 it is dynamic (useful for very large $message_headers, for example).
4586 6. Lookups now cache *every* query, not just the most recent. A new, separate
4587 store pool is used for this. It can be recovered when all lookup caches are
4588 flushed. Lookups now release memory at the end of their result strings.
4589 This has involved some general refactoring of the lookup sources.
4591 7. Some code has been added to the store_xxx() functions to reduce the amount
4592 of flapping under certain conditions.
4594 8. log_incoming_interface used to affect only the <= reception log lines. Now
4595 it causes the local interface and port to be added to several more SMTP log
4596 lines, for example "SMTP connection from", and rejection lines.
4598 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4600 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4602 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4603 could be overwritten at the end of the current message (or the start of a
4604 new message if it was set in a HELO ACL). The value is now preserved for
4605 the duration of the SMTP connection.
4607 12. If a transport had a headers_rewrite setting, and a matching header line
4608 contained an unqualified address, that address was qualified, even if it
4609 did not match any rewriting rules. The underlying bug was that the values
4610 of the flags that permit the existence of unqualified sender and recipient
4611 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4612 non-local messages, and by -bnq for local messages) were not being
4613 preserved with the message after it was received.
4615 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4616 "next input=" as part of the text comprising the host identity instead of
4617 the correct text. The code was using the same buffer for two different
4618 strings. However, depending on which order the printing function evaluated
4619 its arguments, the bug did not always show up. Under Linux, for example, my
4620 test suite worked just fine.
4622 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4623 doesn't work with some older versions of Perl. It has been changed to "my",
4624 which in any case is probably the better facility to use.
4626 15. A really picky compiler found some instances of statements for creating
4627 error messages that either had too many or two few arguments for the format
4630 16. The size of the buffer for calls to the DNS resolver has been increased
4631 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4632 for addresses that have a lot of PTR records. This alleviates a problem; it
4633 does not fully solve it.
4635 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4636 buffer now truncates the list and logs the incident, which is the same
4637 action as happens when Exim is looking up a host name and its aliases.
4638 Previously in this situation something unpredictable would happen;
4639 sometimes it was "internal error: store_reset failed".
4641 18. If a server dropped the connection unexpectedly when an Exim client was
4642 using GnuTLS and trying to read a response, the client delivery process
4643 crashed while trying to generate an error log message.
4645 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4646 string, for example:
4648 warn message = H1: something\nH2: something
4650 the text was added as a single header line from Exim's point of view
4651 though it ended up OK in the delivered message. However, searching for the
4652 second and subsequent header lines using $h_h2: did not work. This has been
4653 fixed. Similarly, if a system filter added multiple headers in this way,
4654 the routers could not see them.
4656 20. Expanded the error message when iplsearch is called with an invalid key to
4657 suggest using net-iplsearch in a host list.
4659 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4660 ACLs are no longer actually imposed (and a message to that effect is
4663 22. If a "gecos" field in a passwd entry contained escaped characters, in
4664 particular, if it contained a \" sequence, Exim got it wrong when building
4665 a From: or a Sender: header from that name. A second bug also caused
4666 incorrect handling when an unquoted " was present following a character
4667 that needed quoting.
4669 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4670 was not being matched caselessly.
4672 24. Arranged for all hyphens in the exim.8 source to be escaped with
4675 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4676 properly. Recipient callout cache records were still being keyed to include
4677 the sender, even when use_sender was set false. This led to far more
4678 callouts that were necessary. The sender is no longer included in the key
4679 when use_sender is false.
4681 26. Added "control = submission" modifier to ACLs.
4683 27. Added the ${base62d: operator to decode base 62 numbers.
4685 28. dnsdb lookups can now access SRV records.
4687 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4688 the configuration file.
4690 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4691 (-v) mode. This makes the output for a verbose queue run more intelligible.
4693 31. Added a use_postmaster feature to recipient callouts.
4695 32. Added the $body_zerocount variable, containing the number of binary zero
4696 bytes in the message body.
4698 33. The time of last modification of the "new" subdirectory is now used as the
4699 "mailbox time last read" when there is a quota error for a maildir
4702 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4704 35. Added +ignore_unknown as a special item in host lists.
4706 36. Code for decoding IPv6 addresses in host lists is now included, even if
4707 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4708 address was recognized as an IP address, but was then not correctly decoded
4709 into binary, causing unexpected and incorrect effects when compared with
4716 1. Very minor rewording of debugging text in manualroute to say "list of
4717 hosts" instead of "hostlist".
4719 2. If verify=header_syntax was set, and a header line with an unqualified
4720 address (no domain) and a large number of spaces between the end of the
4721 name and the colon was received, the reception process suffered a buffer
4722 overflow, and (when I tested it) crashed. This was caused by some obsolete
4723 code that should have been removed. The fix is to remove it!
4725 3. When running in the test harness, delay a bit after writing a bounce
4726 message to get a bit more predictability in the log output.
4728 4. Added a call to search_tidyup() just before forking a reception process. In
4729 theory, someone could use a lookup in the expansion of smtp_accept_max_
4730 per_host which, without the tidyup, could leave open a database connection.
4732 5. Added the variables $recipient_data and $sender_data which get set from a
4733 lookup success in an ACL "recipients" or "senders" condition, or a router
4734 "senders" option, similar to $domain_data and $local_part_data.
4736 6. Moved the writing of debug_print from before to after the "senders" test
4739 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4740 problems for message scanning, either using a data ACL, or using
4741 local_scan() because the Received: header was not generated till after they
4742 were called (in order to set the time as the time of reception completion).
4743 I have revised the way this works. The header is now generated after the
4744 body is received, but before the ACL or local_scan() are called. After they
4745 are run, the timestamp in the header is updated.
4751 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4752 before starting a queue runner without re-exec. This happened only when
4753 deliver_drop_privilege was set or when the Exim user was set to root. The
4754 effect of the bug was that timeouts during subsequent deliveries caused
4755 crashes instead of being properly handled. The handler is now left at its
4756 default (and expected) setting.
4758 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4759 message, again when deliver_drop_privilege is set or Exim is run as root.
4760 The bug described in (1) was not present in this case, but the tidying up
4761 of the other signals was missing. I have made the two cases consistent.
4763 3. The ignore_target_hosts setting on a manualroute router was being ignored
4764 for hosts that were looked up using the /MX notation.
4766 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4769 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4770 operated on the sender address. After changing the $sender_address to <>
4771 for the sender address verify, Exim was re-instated it as the original
4772 (before rewriting) address, but remembering that it had rewritten it, so it
4773 wasn't rewriting it again. This bug also had the effect of breaking the
4774 sender address verification caching when the sender address was rewritten.
4776 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4777 This has been changed so that if the ip literal address matches
4778 ignore_target_hosts, the router declines.
4780 7. Added expansion conditions match_domain, match_address, and match_local_
4781 part (NOT match_host).
4783 8. The placeholder for the Received: header didn't have a length field set.
4785 9. Added code to Exim itself and to exim_lock to test for a specific race
4786 condition that could lead to file corruption when using MBX delivery. The
4787 issue is with the lockfile that is created in /tmp. If this file is removed
4788 after a process has opened it but before that process has acquired a lock,
4789 there is the potential for a second process to recreate the file and also
4790 acquire a lock. This could lead to two Exim processes writing to the file
4791 at the same time. The added code performs the same test as UW imapd; it
4792 checks after acquiring the lock that its file descriptor still refers to
4793 the same named file.
4795 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4796 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4799 11. Added the smtp_active_hostname option. If used, this will typically be made
4800 to depend on the incoming interface address. Because $interface_address is
4801 not set up until the daemon has forked a reception process, error responses
4802 that can happen earlier (such as "too many connections") no longer contain
4805 12. If an expansion in a condition on a "warn" statement fails because a lookup
4806 defers, the "warn" statement is abandoned, and the next ACL statement is
4807 processed. Previously this caused the whole ACL to be aborted.
4809 13. Added the iplsearch lookup type.
4811 14. Added ident_timeout as a log selector.
4813 15. Added tls_certificate_verified as a log selector.
4815 16. Added a global option tls_require_ciphers (compare the smtp transport
4816 option of the same name). This controls incoming TLS connections.
4818 17. I finally figured out how to make tls_require_ciphers do a similar thing
4819 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4820 before starting the TLS session.
4822 18. Tabs are now shown as \t in -bP output.
4824 19. If the log selector return_path_on_delivery was set, Exim crashed when
4825 bouncing a message because it had too many Received: header lines.
4827 20. If two routers both had headers_remove settings, and the first one included
4828 a superfluous trailing colon, the final name in the first list and the
4829 first name in the second list were incorrectly joined into one item (with a
4830 colon in the middle).
4836 1. Added -C and -D options to the exinext utility, mainly to make it easier
4837 to include in the automated testing, but these could be helpful when
4838 multiple configurations are in use.
4840 2. The exinext utility was not formatting the output nicely when there was
4841 an alternate port involved in the retry record key, nor when there was a
4842 message id as well (for retries that were specific to a specific message
4843 and a specific host). It was also confused by IPv6 addresses, because of
4844 the additional colons they contain. I have fixed the IPv4 problem, and
4845 patched it up to do a reasonable job for IPv6.
4847 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4848 delivery, the log line now contains "pipelined" if PIPELINING was used.
4850 4. An SMTP transport process used to panic and die if the bind() call to set
4851 an explicit outgoing interface failed. This has been changed; it is now
4852 treated in the same way as a connect() failure.
4854 5. A reference to $sender_host_name in the part of a conditional expansion
4855 that was being skipped was still causing a DNS lookup. This no longer
4858 6. The def: expansion condition was not recognizing references to header lines
4859 that used bh_ and bheader_.
4861 7. Added the _cache feature to named lists.
4863 8. The code for checking quota_filecount in the appendfile transport was
4864 allowing one more file than it should have been.
4866 9. For compatibility with Sendmail, the command line option
4874 and sets the incoming protocol and host name (for trusted callers). The
4875 host name and its colon can be omitted when only the protocol is to be set.
4876 Note the Exim already has two private options, -pd and -ps, that refer to
4877 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4878 "s", but I don't think that's a major issue.
4880 10. A number of refactoring changes to the code, none of which should affect
4883 (a) The number of logging options was getting close to filling up the
4884 32-bit word that was used as a bit map. I have split them into two classes:
4885 those that are passed in the argument to log_write(), and those that are
4886 only ever tested independently outside of that function. These are now in
4887 separate 32-bit words, so there is plenty of room for expansion again.
4888 There is no change in the user interface or the logging behaviour.
4890 (b) When building, for example, log lines, the code previously used a
4891 macro that called string_cat() twice, in order to add two strings. This is
4892 not really sufficiently general. Furthermore, there was one instance where
4893 it was actually wrong because one of the argument was used twice, and in
4894 one call a function was used. (As it happened, calling the function twice
4895 did not affect the overall behaviour.) The macro has been replaced by a
4896 function that can join an arbitrary number of extra strings onto a growing
4899 (c) The code for expansion conditions now uses a table and a binary chop
4900 instead of a serial search (which was left over from when there were very
4901 few conditions). Also, it now recognizes conditions like "pam" even when
4902 the relevant support is not compiled in: a suitably worded error message is
4903 given if an attempt is made to use such a condition.
4905 11. Added ${time_interval:xxxxx}.
4907 12. A bug was causing one of the ddress fields not to be passed back correctly
4908 from remote delivery subprocesses. The field in question was not being
4909 subsequently used, so this caused to problems in practice.
4911 13. Added new log selectors queue_time and deliver_time.
4913 14. Might have fixed a bug in maildirsizefile handling that threw up
4914 "unexpected character" debug warnings, and recalculated the data
4915 unnecessarily. In any case, I expanded the warning message to give more
4918 15. Added the message "Restricted characters in address" to the statements in
4919 the default ACL that block characters like @ and % in local parts.
4921 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4922 Three changes have been made:
4924 (a) There was a serious bug; a negative response to MAIL caused the whole
4925 recipient domain to be cached as invalid, thereby blocking all messages
4926 to all local parts at the same domain, from all senders. This bug has
4927 been fixed. The domain is no longer cached after a negative response to
4928 MAIL if the sender used is not empty.
4930 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4933 (c) A new callout option, "use_sender" has been added for people who want
4934 the modified behaviour.
4940 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4943 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4944 indeed breaks things for older releases.
4946 3. Added additional logging to the case where there is a problem reading data
4947 from a filter that is running in a subprocess using a pipe, in order to
4948 try to track down a specific problem.
4950 4. Testing facility fudge: when running in the test harness and attempting
4951 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4952 getting "No route to host". Convert this to a timeout.
4954 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4957 6. Some OS don't have socklen_t but use size_t instead. This affects the
4958 fifth argument of getsockopt() amongst other things. This is now
4959 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4960 can be set for individual OS. I have set it for SunOS5, OSF1, and
4961 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4962 some earlier ones do not.
4964 7. Change 4.30/15 was not doing the test caselessly.
4966 8. The standard form for an IPv6 address literal was being rejected by address
4967 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4968 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4969 this, as well as the form without the "IPv6" on the front (but only when
4970 address literals are enabled, of course).
4972 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4974 10. Exim crashed if a message with an empty sender address specified by -f
4975 encountered a router with an errors_to setting. This could be provoked only
4976 by a command such as
4980 where an empty string was supplied; "<>" did not hit this bug.
4982 11. Installed PCRE release 4.5.
4984 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4985 remained set. It is now erased.
4987 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4988 times from message ids (which are base 36 rather than the normal 62).
4990 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4991 were being counted as actual protocol errors, and logged if the log
4992 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4993 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4994 and DATA following a set of rejected RCPTs do not count as protocol errors.
4995 In other words, Exim assumes they were pipelined, though this may not
4996 actually be the case. Of course, in all cases the client gets an
4997 appropriate error code.
4999 15. If a lookup fails in an ACL condition, a message about the failure may
5000 be available; it is used if testing the ACL cannot continue, because most
5001 such messages specify what the cause of the deferral is. However, some
5002 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5003 that caused an old message to be retained and used if a later statement
5004 caused a defer, replacing the real cause of the deferral.
5006 16. If an IP address had so many PTR records that the DNS lookup buffer
5007 was not large enough to hold them, Exim could crash while trying to process
5008 the truncated data. It now detects and logs this case.
5010 17. Further to 4.21/58, another change has been made: if (and only if) the
5011 first line of a message (the first header line) ends with CRLF, a bare LF
5012 in a subsequent header line has a space inserted after it, so as not to
5013 terminate the header.
5015 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5016 unnecessarily, used atoi() instead of strtol(), and didn't check the
5017 termination when getting file sizes from file names by regex.
5019 19. Completely re-implemented the support for maildirsize files, in the light
5020 of a number of problems with the previous contributed implementation
5021 (4.30/29). In particular:
5023 . If the quota is zero, the maildirsize file is maintained, but no quota is
5026 . If the maildir directory does not exist, it is created before any attempt
5027 to write a maildirsize file.
5029 . The quota value in the file is just a cache; if the quota is changed in
5030 the transport, the new value overrides.
5032 . A regular expression is available for excluding directories from the
5035 20. The autoreply transport checks the characters in options that define the
5036 message's headers; it allows continued headers, but it was checking with
5037 isspace() after an embedded newline instead of explicitly looking for a
5040 21. If all the "regular" hosts to which an address was routed had passed their
5041 expiry times, and had not reached their retry times, the address was
5042 bounced, even if fallback hosts were defined. Now Exim should go on to try
5045 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5046 equivalent code in the SMTP transport. Some hosts send humungous responses
5047 to HELO/EHLO, more than 1024 it seems.
5049 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5050 gives compiler warnings in some environments. I've now done it "properly",
5053 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5054 (because of problems with the built-in one) was declared to return uschar *
5055 instead of char *, causing compiler failure.
5057 25. Fixed a file descriptor leak when processing alias/forward files.
5059 26. Fixed a minor format string issue in dbfn.c.
5061 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5063 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5064 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5067 29. When a sender address is verified, it is cached, to save repeating the test
5068 when there is more than one recipient in a message. However, when the
5069 verification involves a callout, it is possible for different callout
5070 options to be set for different recipients. It is too complicated to keep
5071 track of this in the cache, so now Exim always runs a verification when a
5072 callout is required, relying on the callout cache for the optimization.
5073 The overhead is duplication of the address routing, but this should not be
5076 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5077 to be verified with callout=postmaster, and the main callout worked but the
5078 postmaster check failed, the verification correctly failed. However, if a
5079 subsequent RCPT command asked for sender verification *without* the
5080 postmaster check, incorrect caching caused this verification also to fail,
5083 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5084 it was not caching the DNS options (qualify_single, search_parents) that
5085 were used when the lookup failed. A subsequent lookup with different
5086 options therefore always gave the same answer, though there were cases
5087 where it should not have. (Example: a "domains = !$mx_any" option on a
5088 dnslookup router: the "domains" option is always processed without any
5089 widening, but the router might have qualify_single set.) Now Exim uses the
5090 cached value only when the same options are set.
5092 32. Added John Jetmore's "exipick" utility to the distribution.
5094 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5095 than a timeout (e.g. a certificate is required, and is not provided), an
5096 Exim server now closes the connection immediately. Previously it waited for
5097 the client to close - but if the client is SSL, it seems that they each
5098 wait for each other, leading to a delay before one of them times out.
5100 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5101 maintained 0.8.x compatibility because I don't think many are using it, and
5102 it is clearly obsolete.
5104 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5107 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5108 client certificate was expired. A simple patch fixes this, though I don't
5109 understand the full logic of why the verify callback is called multiple
5112 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5113 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5114 which causes problems with some clients (such as the Certicom SSL Plus
5115 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5116 disables the coutermeasure allowing Eudora to connect."
5118 38. Exim was not checking that a write() to a log file succeeded. This could
5119 lead to Bad Things if a log got too big, in particular if it hit a file
5120 size limit. Exim now panics and dies if it cannot write to a log file, just
5121 as it does if it cannot open a log file.
5123 39. Modified OS/Makefile-Linux so that it now contains
5125 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5127 The two -D definitions ensure that Exim is compiled with large file
5128 support, which makes it possible to handle log files that are bigger than
5131 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5132 instance) a domain was checked against a named list that involved a lookup,
5133 causing $domain_data to be set, then another domain was checked against the
5134 same list, then the first domain was re-checked, the value of $domain_data
5135 after the final check could be wrong. In particular, if the second check
5136 failed, it could be set empty. This bug probably also applied to
5139 41. The strip_trailing_dot option was not being applied to the address given
5140 with the -f command-line option.
5142 42. The code for reading a message's header from the spool was incrementing
5143 $received_count, but never initializing it. This meant that the value was
5144 incorrect (doubled) while delivering a message in the same process in which
5145 it was received. In the most common configuration of Exim, this never
5146 happens - a fresh exec is done - but it can happen when
5147 deliver_drop_privilege is set.
5149 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5150 it now includes up to 150 characters of the unexpected data in the log
5153 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5154 and building data strings. The size of both of these buffers was 10 000
5155 bytes - far larger than anybody would *ever* want, thought I. Needless to
5156 say, somebody hit the limit. I have increased the maximum line length to
5157 20 000 and the maximum data length of concatenated lines to 100 000. I have
5158 also fixed two bugs, because there was no checking on these buffers. Tsk,
5159 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5160 buffer is too small.
5162 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5163 lsearch lookups. Now it does.
5165 46. When parsing a route_list item in a manualroute router, a fixed-length
5166 buffer was used for the list of hosts. I made this 1024 bytes long,
5167 thinking that nobody would ever have a list of hosts that long. Wrong.
5168 Somebody had a whole pile of complicated expansion conditions, and the
5169 string was silently truncated, leading to an expansion error. It turns out
5170 that it is easier to change to an unlimited length (owing to other changes
5171 that have happened since this code was originally written) than to build
5172 structure for giving a limitation error. The length of the item that
5173 expands into the list of hosts is now unlimited.
5175 47. The lsearch lookup could not handle data where the length of text line was
5176 more than 4095 characters. Such lines were truncated, leading to shortened
5177 data being returned. It should now handle lines of any length.
5179 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5180 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5183 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5184 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5185 no longer happens in comment lines. A list of replacements is now placed
5186 at the head of all of the source files, except those whose only change is
5187 to replace PERL_COMMAND in the very first #! line.
5189 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5190 messages on the queue, with a bottom-up merge sort, using code contributed
5191 by Michael Haardt. This should make operations like -bp somewhat faster on
5192 large queues. It won't affect queue runners, except when queue_run_in_order
5195 51. Installed eximstats 1.31 in the distribution.
5197 52. Added support for SRV lookups to the dnslookup router.
5199 53. If an ACL referred to $message_body or $message_body_end, the value was not
5200 reset for any messages that followed in the same SMTP session.
5202 54. The store-handling optimization for building very long strings was not
5203 differentiating between the different store pools. I don't think this
5204 actually made any difference in practice, but I've tidied it.
5206 55. While running the routers to verify a sender address, $sender_address
5207 was still set to the sender address. This is wrong, because when routing to
5208 send a bounce to the sender, it would be empty. Therefore, I have changed
5209 it so that, while verifying a sender address, $sender_address is set to <>.
5210 (There is no change to what happens when verifying a recipient address.)
5212 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5213 target A or AAAA records (if not already returned) without resetting the
5214 qualify_single or search_parents options of the DNS resolver. These are
5215 inappropriate in this case because the targets of MX and SRV records must
5216 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5217 target that, when qualified, matched something in the local domain. These
5218 two options are now turned off when doing these lookups.
5220 57. It seems that at least some releases of Reiserfs (which does not have the
5221 concept of a fixed number of inodes) returns zero and not -1 for the
5222 number of available inodes. This interacted badly with check_spool_inodes,
5223 which assumed that -1 was the "no such thing" setting. What I have done is
5224 to check that the total number of inodes is greater than zero before doing
5225 the test of how many are available.
5227 58. When a "warn" ACL statement has a log_message modifier, the message is
5228 remembered, and not repeated. This is to avoid a lot of repetition when a
5229 message has many recipients that cause the same warning to be written.
5230 Howewer, Exim was preserving the list of already written lines for an
5231 entire SMTP session, which doesn't seem right. The memory is now reset if a
5232 new message is started.
5234 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5235 debugging output unless log_rewrite was also set.
5237 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5238 of (int)(handle) when we know that handle contains (void *)(-1).
5240 61. The Exim daemon panic-logs an error return when it closes the incoming
5241 connection. However "connection reset by peer" seems to be common, and
5242 isn't really an error worthy of noting specially, so that particular error
5245 62. When Exim is trying to find all the local interfaces, it used to panic and
5246 die if the ioctl to get the interface flags failed. However, it seems that
5247 on at least one OS (Solaris 9) it is possible to have an interface that is
5248 included in the list of interfaces, but for which you get a failure error
5249 for this call. This happens when the interface is not "plumbed" into a
5250 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5251 failure of the "get flags" call assumes that the interface is down.
5253 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5254 makes life easier for people who are doing arithmetic on fields extracted
5255 from dates, where you often get leading zeros that should not be
5256 interpreted as octal.
5258 64. Added qualify_domain to the redirect router, to override the global
5261 65. If a pathologically long header line contained very many addresses (the
5262 report of this problem mentioned 10 000) and each of them was rewritten,
5263 Exim could use up a very large amount of memory. (It kept on making new
5264 copies of the header line as it rewrote, and never released the old ones.)
5265 At the expense of a bit more processing, the header rewriting function has
5266 been changed so that it no longer eats memory in this way.
5268 66. The generation of the Received: header has been moved from the time that a
5269 message starts to be received, to the time that it finishes. The timestamp
5270 in the Received: header should now be very close to that of the <= log
5271 line. There are two side-effects of this change:
5273 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5274 logged header lines no longer include the local Received: line, because
5275 it has not yet been created. The same applies to a copy of the message
5276 that is returned to a non-SMTP sender when a message is rejected.
5278 (b) When a filter file is tested using -bf, no additional Received: header
5279 is added to the test message. After some thought, I decided that this
5282 This change does not affect the value of $received_for. It is still set
5283 after address rewriting, but before local_scan() is called.
5285 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5287 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5288 gave an unhelpful panic error message, and a defer error. I have managed to
5289 change this behaviour so that it now rejects any supplied certificate,
5290 which seems right, as the list of acceptable certificates is empty.
5292 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5293 gave an unhelpful defer error. I have not managed to make this reject any
5294 supplied certificates, but the error message it gives is "no certificate
5295 supplied", which is not helpful.
5297 70. exigrep's output now also includes lines that are not associated with any
5298 message, but which match the given pattern. Implemented by a patch from
5299 Martin Sluka, which also tidied up the Perl a bit.
5301 71. Recipient callout verification, like sender verification, was using <> in
5302 the MAIL FROM command. This isn't really the right thing, since the actual
5303 sender may affect whether the remote host accepts the recipient or not. I
5304 have changed it to use the actual sender in the callout; this means that
5305 the cache record is now keyed on a recipient/sender pair, not just the
5306 recipient address. There doesn't seem to be a real danger of callout loops,
5307 since a callout by the remote host to check the sender would use <>.
5308 [SEE ABOVE: changed after hitting problems.]
5310 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5311 temporary errors. However, in the case of such a code being given after
5312 the end of a data transmission (i.e. after ".") Exim was failing to write
5313 a retry record for the message. (Yes, there was some broken host that was
5314 actually sending 8xx at this point.)
5316 73. An unknown lookup type in a host list could cause Exim to panic-die when
5317 the list was checked. (An example that provoked this was putting <; in the
5318 middle of a list instead of at the start.) If this happened during a DATA
5319 ACL check, a -D file could be left lying around. This kind of configuration
5320 error no longer causes Exim to die; instead it causes a defer errror. The
5321 incident is still logged to the main and panic logs.
5323 74. Buglet left over from Exim 3 conversion. The message "too many messages
5324 in one connection" was written to the rejectlog but not the mainlog, except
5325 when address rewriting (yes!) was being logged.
5327 75. Added write_rejectlog option.
5329 76. When a system filter was run not as root (that is, when system_filter_user
5330 was set), the values of the $n variables were not being returned to the
5331 main process; thus, they were not subsequently available in the $sn
5334 77. Added +return_path_on_delivery log selector.
5336 78. A connection timeout was being treated differently from recipients deferred
5337 when testing hosts_max_try with a message that was older than the host's
5338 retry timeout. (The host should not be counted, thus allowing all hosts to
5339 be tried at least once before bouncing.) This may have been the cause of an
5340 occasionally reported bug whereby a message would remain on the queue
5341 longer than the retry timeout, but would be bounced if a delivery was
5342 forced. I say "may" because I never totally pinned down the problem;
5343 setting up timeout/retry tests is difficult. See also the next item.
5345 79. The ultimate address timeout was not being applied to errors that involved
5346 a combination of host plus message (for example, a timeout on a MAIL
5347 command). When an address resolved to a number of possible hosts, and they
5348 were not all tried for each delivery (e.g. because of hosts_max_try), a
5349 message could remain on the queue longer than the retry timeout.
5351 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5354 81. Fixed an obscure SMTP outgoing bug which required at least the following
5355 conditions: (a) there was another message waiting for the same server;
5356 (b) the server returned 5xx to all RCPT commands in the first message so
5357 that the message was not completed; (c) the server dropped the connection
5358 or gave a negative response to the RSET that Exim sends to abort the
5359 transaction. The observed case was a dropped connection after DATA that had
5360 been sent in pipelining mode. That is, the server had advertised PIPELINING
5361 but was not implementing it correctly. The effect of the bug was incorrect
5362 behaviour, such as trying another host, and this could lead to a crash.
5368 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5369 and daemon.c were passed as pointers to ints; they should have been
5370 pointers to socklen_t variables (which are typically unsigned ints).
5372 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5375 3. Fixed a really odd bug that affected only the testing scheme; patching a
5376 certain fixed string in the binary changed the value of another string that
5377 happened to be identical to the end of the original first string.
5379 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5380 name", it returns that address as the IP address. On some operating
5381 systems (e.g. Solaris), it also passes back the IP address string as the
5382 "host name". However, on others (e.g. Linux), it passes back an empty
5383 string. Exim wasn't checking for this, and was changing the host name to an
5384 empty string, assuming it had been canonicized.
5386 5. Although rare, it is permitted to have more than one PTR record for a given
5387 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5388 all the names associated with an address, because they do in Solaris.
5389 However, it seems that they do not in Linux for data that comes from the
5390 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5391 I found this out when I moved to a new Linux workstation and tried to run
5392 the Exim test suite.
5394 To get round this problem I have changed the code so that it now does its
5395 own call to the DNS to look up PTR records when searching for a host name.
5396 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5397 addresses that are only in /etc/hosts are still found.
5399 This behaviour is, however, controlled by an option called host_lookup_
5400 order, which defaults to "bydns:byaddr". If people want to use the other
5401 order, or indeed, just use one or the other means of lookup, they can
5402 specify it in this variable.
5404 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5405 some operating systems, this comes back from gethostbyaddr() as an empty
5406 string, and this is what Exim used to test for. However, it seems that in
5407 other systems, "." is yielded. Exim now tests for this case too.
5409 7. The values of check_spool_space and check_log_space are now held internally
5410 as a number of kilobytes instead of an absolute number of bytes. If a
5411 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5412 kilobyte. This means that much larger values can be stored.
5414 8. Exim monitor: an attempt to get the action menu when not actually pointing
5415 at a message produces an empty menu entitled "No message selected". This
5416 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5417 no entries in it ("Shell widget menu has zero width and/or height"). So I
5418 have added a single, blank menu entry in this case.
5420 9. Added ${quote_local_part.
5422 10. MIME decoding is now applied to the contents of Subject: header lines when
5425 11. Now that a reference to $sender_host_address automatically causes a reverse
5426 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5427 host lookup before query-style lookups in lists that might use this
5428 variable. This has therefore been abolished, and the "net-" prefix is no
5429 longer necessary for query-style lookups.
5431 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5432 have been a typo for LFLAGS, so it has been changed.
5434 13. The install script calls Exim with "-C /dev/null" in order to find the
5435 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5436 to be output. Howeve, since Exim outputs its version number before the
5437 error, it didn't break the script. It just looked ugly. I fixed this by
5438 always allowing "-C /dev/null" if the caller is root.
5440 14. Ignore overlarge ACL variable number when reading spool file - insurance
5441 against a later release with more variables having written the file.
5443 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5444 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5445 this, as well as the form without the "IPv6" on the front.
5447 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5448 OS/Makefile-Darwin file.
5450 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5451 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5453 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5454 variable. However, some OS use ints for inodes. I've added cast to long int
5455 to get rid of the compiler warning.
5457 19. I had forgotten to lock out "/../" in configuration file names when
5458 ALT_CONFIG_PREFIX was set.
5460 20. Routers used for verification do not need to specify transports. However,
5461 if such a router generated a host list, and callout was configured, Exim
5462 crashed, because it could not find a port number from the (non-existent)
5463 transport. It now assumes port 25 in this circumstance.
5465 21. Added the -t option to exigrep.
5467 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5468 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5469 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5470 them is set, LOOKUP_LSEARCH is forced.
5472 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5474 24. Added sender and host information to the "rejected by local_scan()" log
5475 line; previously there was no indication of these.
5477 25. Added .include_if_exists.
5479 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5480 Linux. It turns out that not all file systems support this. Apparently some
5481 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5482 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5483 not supported on the file descriptor, is now ignored when Exim is trying to
5484 sync a directory. This applies only to Linux.
5486 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5488 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5489 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5492 29. Added support for maildirsize files from supplied patch (modified a bit).
5494 30. The use of :fail: followed by an empty string could lead Exim to respond to
5495 sender verification failures with (e.g.):
5497 550 Verification failed for <xxx>
5498 550 Sender verify failed
5500 where the first response line was missing the '-' that indicates it is not
5501 the final line of the response.
5503 31. The loop for finding the name of the user that called Exim had a hardwired
5504 limit of 10; it now uses the value of finduser_retries, which is used for
5505 all other user lookups.
5507 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5510 33. Exim was neglecting to zero errno before one call of strtol() when
5511 expanding a string and expecting an integer value. On some systems this
5512 resulted in spurious "integer overflow" errors. Also, it was casting the
5513 result into an int without checking.
5515 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5516 did not work. The code looks as if it has *never* worked, though it appears
5517 to have been documented since at least releast 1.62. I have made it work.
5519 35. The "timeout_DNS" error in retry rules, also documented since at least
5520 1.62, also never worked. As it isn't clear exactly what this means, and
5521 clearly it isn't a major issue, I have abolished the feature by treating it
5522 as "timeout", and writing a warning to the main and panic logs.
5524 36. The display of retry rules for -brt wasn't always showing the error code
5527 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5528 timeout_connect_A, timeout_connect_MX.
5530 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5531 to the empty sender.
5533 39. The daemon was not analysing the content of -oX till after it had closed
5534 stderr and disconnected from the controlling terminal. This meant that any
5535 syntax errors were only noted on the panic log, and the return code from
5536 the command was 0. By re-arranging the code a little, I've made the
5537 decoding happen first, so such errors now appear on stderr, and the return
5538 code is 1. However, the actual setting up of the sockets still happens in
5539 the disconnected process, so errors there are still only recorded on the
5542 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5543 connections (as happens on some IP stacks) was logged at start up time as
5544 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5545 it from "IPv6 and IPv4", which means that two separate sockets are being
5548 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5549 says whether AF_INET or AF_INET6 was passed as an argument.
5551 42. Exiwhat output was messed up when time zones were included in log
5554 43. Exiwhat now gives more information about the daemon's listening ports,
5555 and whether -tls-on-connect was used.
5557 44. The "port" option of the smtp transport is now expanded.
5559 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5560 silently ignored. Now an error message is written to the main and panic
5563 46. There's a new ACL modifier called "logwrite" which writes to a log file
5564 as soon as it is encountered.
5566 47. Added $local_user_uid and $local_user_gid at routing time.
5568 48. Exim crashed when trying to verify a sender address that was being
5571 49. Exim was recognizing only a space character after ".include". It now also
5572 recognizes a tab character.
5574 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5575 extracting the relevant information from the specification. The man page no
5576 longer contains scrambled data for the -d option, and I've added a section
5577 at the front about calling Exim under different names.
5579 51. Added "extra_headers" argument to the "mail" command in filter files.
5581 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5584 53. Installed eximstats 1.29.
5586 54. Added transport_filter_timeout as a generic transport option.
5588 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5589 Cc: header lines. This was required by RFC 822, but it not required by RFC
5592 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5593 incoming messages that did not have them. Now it does so only if the
5594 message originates locally, that is, if there is no associated remote host
5595 address. When Resent- header lines are present, this applies to the Resent-
5596 lines rather than the non-Resent- lines.
5598 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5599 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5601 58. Messages for configuration errors now include the name of the main
5602 configuration file - useful now that there may be more than one file in a
5603 list (.included file names were always shown).
5605 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5606 for those rare installations that do not start the daemon as root or run it
5607 setuid root. I've cut out the call to initgroups() if the daemon is not
5610 60. The Exim user and group can now be bound into the binary as text strings
5611 that are looked up at the start of Exim's processing.
5613 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5615 62. Added $mailstore_basename variable.
5617 63. Installed patch to sieve.c from Michael Haardt.
5619 64. When Exim failed to open the panic log after failing to open the main log,
5620 the original message it was trying to log was written to stderr and debug
5621 output, but if they were not available (the usual case in production), it
5622 was lost. Now it is written to syslog before the two lines that record the
5623 failures to open the logs.
5625 65. Users' Exim filters run in subprocesses under the user's uid. It is
5626 possible for a "deliver" command or an alias in a "personal" command to
5627 provoke an address rewrite. If logging of address rewriting is configured,
5628 this fails because the process is not running as root or exim. There may be
5629 a better way of dealing with this, but for the moment (because 4.30 needs
5630 to be released), I have disabled address rewrite logging when running a
5631 filter in a non-root, non-exim process.
5637 1. The buildconfig auxiliary program wasn't quoting the value set for
5638 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5639 not defined. This bug was masked in 4.22 by the effect that was fixed in
5642 2. Some messages that were rejected after a message id was allocated were
5643 shown as "incomplete" by exigrep. It no longer does this for messages that
5644 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5646 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5647 have allow_domain_literals set, the ID did not get logged in the <= line.
5648 Domain literals are now always recognized in Message-ID: header lines.
5650 4. The first argument for a ${extract expansion item is the key name or field
5651 number. Leading and trailing spaces in this item were not being ignored,
5652 causing some misleading effects.
5654 5. When deliver_drop_privilege was set, single queue runner processes started
5655 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5656 same command in the process it spins off) were not dropping privilege.
5658 6. When the daemon running as "exim" started a queue runner, it always
5659 re-executed Exim in the spun-off process. This is a waste of effort when
5660 deliver_drop_privilege is set. The new process now just calls the
5661 queue-runner function directly.
5667 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5670 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5671 ignored. Though the use of -oP was forcing the writing of a pid file, it
5672 was always written to the default place.
5674 3. If the message "no IP address found for host xxxx" is generated during
5675 incoming verification, it is now followed by identification of the incoming
5676 connection (so you can more easily find what provoked it).
5678 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5680 5. Added some features to "harden" Exim a bit more against certain attacks:
5682 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5683 be put in Local/Makefile. This is like the never_users runtime option,
5684 but it cannot be overridden. The default setting is "root".
5686 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5687 prefix string with which any file named in a -C command line option
5690 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5691 is retained for -C and -D only if the caller of Exim is root. Without
5692 it, the exim user may also use -C and -D and retain privilege.
5694 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5695 command line option is disabled.
5697 6. Macro names set by the -D option must start with an upper case letter, just
5698 like macro names defined in the configuration file.
5700 7. Added "dereference=" facility to LDAP.
5702 8. Two instances of the typo "uknown" in the source files are fixed.
5704 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5705 the Configure-Makefile script screwed up while processing it.
5707 10. Incorporated PCRE 4.4.
5709 11. The SMTP synchronization check was not operating right at the start of an
5710 SMTP session. For example, it could not catch a HELO sent before the client
5711 waited for the greeting. There is now a check for outstanding input at the
5712 point when the greeting is written. Because of the duplex, asynchronous
5713 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5714 way, but not yet received, when the check is performed.
5716 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5717 on TCP/IP sockets, because this apparently causes some broken clients to
5720 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5721 unchanged) from the Cygwin maintainer.
5723 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5724 was supported instead of testing for mbx. Effectively a typo.
5726 15. The spa authenticator server code was not checking that the input it
5727 received was valid base64.
5729 16. The debug output line for the "set" modifier in ACLs was not showing the
5730 name of the variable that was being set.
5732 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5734 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5735 lookup on its own. Something else was needed to trigger the lookup. For
5736 example, a match in host_lookup or the need for a host name in a host list.
5737 Now, if $sender_host_name is referenced and the host name has not yet been
5738 looked up, a lookup is performed. If the lookup fails, the variable remains
5739 empty, and $host_lookup_failed is set to "1".
5741 19. Added "eqi" as a case-independent comparison operator.
5743 20. The saslauthd authentication condition could segfault if neither service
5744 nor realm was specified.
5746 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5747 error message that was logged was misleading, and incoming SMTP
5748 connections were dropped. The message is now more accurate, and temporary
5749 errors are given to SMTP connections.
5751 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5752 (except RSET) with a 421 error, until QUIT is received. However, it was
5753 failing to send a response to QUIT.
5755 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5756 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5757 ACL is now run after the helo_try_verify_hosts code.
5759 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5760 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5761 case-independent, so other case variants are also recognized). Apparently
5762 some systems use these upper case variants.
5764 25. If more than two messages were waiting for the same host, and a transport
5765 filter was specified for the transport, Exim sent two messages over the
5766 same TCP/IP connection, and then failed with "socket operation on non-
5767 socket" when it tried to send the third.
5769 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5771 27. The extern definition of crypt16() in expand.c was not being excluded when
5772 the OS had its own crypt16() function.
5774 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5775 as a preferred synonym for return_size_limit, both as an option and as an
5778 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5780 30. Changed the default configuration ACL to relax the local part checking rule
5781 for addresses that are not in any local domains. For these addresses,
5782 slashes and pipe symbols are allowed within local parts, but the sequence
5783 /../ is explicitly forbidden.
5785 31. SPA server authentication was not clearing the challenge buffer before
5788 32. log_message in a "warn" ACL statement was writing to the reject log as
5789 well as to the main log, which contradicts the documentation and doesn't
5790 seem right (because no rejection is happening). So I have stopped it.
5792 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5793 However, I am unable to do any testing of this.
5795 34. Fixed an infelicity in the appendfile transport. When checking directories
5796 for a mailbox, to see if any needed to be created, it was accidentally
5797 using path names with one or more superfluous leading slashes; tracing
5798 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5800 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5801 contents are added to the log line that is written for every discarded
5802 recipient. (Previously a log_message setting was ignored.)
5804 36. The ${quote: operator now quotes the string if it is empty.
5806 37. The install script runs exim in order to find its version number. If for
5807 some reason other than non-existence or emptiness, which it checks, it
5808 could not run './exim', it was installing it with an empty version number,
5809 i.e. as "exim-". This error state is now caught, and the installation is
5812 38. An argument was missing from the function that creates an error message
5813 when Exim fails to connect to the socket for saslauthd authentication.
5814 This could cause Exim to crash, or give a corrupted message.
5816 39. Added isip, isip4, and isip6 to ${if conditions.
5818 40. The ACL variables $acl_xx are now saved with the message, and can be
5819 accessed later in routers, transports, and filters.
5821 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5822 strings in the file are not string-expanded.
5824 42. If a MAIL command specified a SIZE value that was too large to fit into an
5825 int variable, the check against message_size_limit failed. Such values are
5826 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5827 day this will have to be increased, but I don't think I want to be around
5828 when emails are that large.
5835 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5836 iconv() is not standard in FreeBSD.
5838 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5839 IPv6 enabled. The observed symptom was a segmentation fault on return from
5840 the function os_common_find_running_interfaces() in src/os.c.
5842 3. In the check_special_case() function in daemon.c I had used "errno" as an
5843 argument name, which causes warnings on some systems. This was basically a
5844 typo, since it was named "eno" in the comments!
5846 4. The code that waits for the clock to tick (at a resolution of some fraction
5847 of a second) so as to ensure message-id uniqueness was always waiting for
5848 at least one whole tick, when it could have waited for less. [This is
5849 almost certainly not relevant at current processor speeds, where it is
5850 unlikely to ever wait at all. But we try to future-proof.]
5852 5. The function that sleeps for a time interval that includes fractions of a
5853 second contained a race. It did not block SIGALRM between setting the
5854 timer, and suspending (a couple of lines later). If the interval was short
5855 and the sigsuspend() was delayed until after it had expired, the suspension
5856 never ended. On busy systems this could lead to processes getting stuck for
5859 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5860 process, before it forks any delivery processes. The open lookup caching
5861 mechanism meant that the open file or database connection was passed into
5862 the delivery process. The problem was that delivery processes always tidy
5863 up cached lookup data. This could cause a problem for the next delivery
5864 process started by the queue runner, because the external queue runner
5865 process does not know about the closure. So the next delivery process
5866 still has data in the lookup cache. In the case of a file lookup, there was
5867 no problem because closing a file descriptor in a subprocess doesn't affect
5868 the parent. However, if the lookup was caching a connection to a database,
5869 the connection was closed, and the second delivery process was likely to
5870 see errors such as "PGSQL: query failed: server closed the connection
5871 unexpectedly". The problem has been fixed by closing all cached lookups
5872 in a queue runner before running a delivery process.
5874 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5875 seem to have the "const" qualifier which it has on other OS. I've
5878 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5879 *of the same type* (client or server) with the same public name that an
5880 error should be diagnosed.
5882 9. When Exim looked up a host name for an IP address, but failed to find the
5883 original IP address when looking up the host name (a safety check), it
5884 output the message "<ip address> does not match any IP for NULL", which was
5885 confusing, to say the least. The bug was that the host name should have
5886 appeared instead of "NULL".
5888 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5889 user that is built into the binary, and the -C or -D options is used, root
5890 privilege is dropped before the configuration file is read. In addition,
5891 logging is switched to stderr instead of the normal log files. If the
5892 configuration then re-defines the Exim user, the unprivileged environment
5893 is probably not what is expected, so Exim logs a panic warning message (but
5896 However, if deliver_drop_privilege is set, the unprivileged state may well
5897 be exactly what is intended, so the warning has been cut out in that case,
5898 and Exim is allowed to try to write to its normal log files.
5904 1. smtp_return_error_details was not giving details for temporary sender
5905 or receiver verification errors.
5907 2. Diagnose a configuration error if two authenticators have the same public
5910 3. Exim used not to create the message log file for a message until the first
5911 delivery attempt. This could be confusing when incoming messages were held
5912 for policy or load reasons. The message log file is now created at the time
5913 the message is received, and an initial "Received" line is written to it.
5915 4. The automatically generated man page for command line options had a minor
5916 bug that caused no ill effects; however, a more serious problem was that
5917 the procedure for building the man page automatically didn't always
5918 operate. Consequently, release 4.20 contains an out-of-date version. This
5919 shouldn't happen again.
5921 5. When building Exim with embedded Perl support, the script that builds the
5922 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5923 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5925 6. The freeze_tell option was not being used for messages that were frozen on
5926 arrival, either by an ACL or by local_scan().
5928 7. Added the smtp_incomplete_transaction log selector.
5930 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5931 was accepting AUTH without a new EHLO.
5933 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5934 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5937 10. Logging of TCP/IP connections (when configured) now happens in the main
5938 daemon process instead of the child process, so that the TCP/IP connection
5939 count is more accurate (but it can never be perfect).
5941 11. The use of "drop" in a nested ACL was not being handled correctly in the
5942 outer ACL. Now, if condition failure induced by the nested "drop" causes
5943 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5944 or "require"), the connection is dropped.
5946 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5947 that yield "discard" can now be used with an "accept" or a "discard" verb,
5948 but an error is generated for any others (because I can't see a useful way
5949 to define what should happen).
5951 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5952 are now processed in the same way as lines in the Exim configuration file.
5953 In particular, continuation lines are supported.
5955 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5957 15. Added -ti meaning -t -i.
5959 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5960 domains, and warn by logging if others are found.
5962 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5963 returned from GIFCONF when Exim is trying to find the list of interfaces on
5964 a host. The code in os.c has been modified to copy each ifreq to an aligned
5965 structure in all cases.
5967 Also, in some cases, the returned ifreq's were being copied to a 'struct
5968 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5969 means the last couple of bytes of an IPv6 address could be chopped if the
5970 ifreq contained only a normal sockaddr (14 bytes storage).
5972 18. Named domain lists were not supported in the hosts_treat_as_local option.
5973 An entry such as +xxxx was not recognized, and was treated as a literal
5976 19. Ensure that header lines added by a DATA ACL are included in the reject log
5977 if the ACL subsequently rejects the message.
5979 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5980 MD5 (which is deprecated).
5982 21. When testing a filter file using -bf, Exim was writing a message when it
5983 took the sender from a "From " line in the message, but it was not doing so
5984 when it took $return_path from a Return-Path: header line. It now does.
5986 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5987 with a valid header line field name (a series of printing characters
5988 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5990 23. Changed "disc" in the source to "disk" to conform to the documentation and
5991 the book and for uniformity.
5993 24. Ignore Sendmail's -Ooption=value command line item.
5995 25. When execve() failed while trying to run a command in a pipe transport,
5996 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5997 could be confused with a return value of 69 from the command itself. This
5998 has been changed to 127, the value the shell returns if it is asked to run
5999 a non-existent command. The wording for the related log line suggests a
6000 non-existent command as the problem.
6002 26. If received_header_text expands to an empty string, do not add a Received:
6003 header line to the message. (Well, it adds a token one on the spool, but
6004 marks it "old" so that it doesn't get used or transmitted.)
6006 27. Installed eximstats 1.28 (addition of -nt option).
6008 28. There was no check for failure on the call to getsockname() in the daemon
6009 code. This can fail if there is a shortage of resources on the system, with
6010 ENOMEM, for example. A temporary error is now given on failure.
6012 29. Contrary to the C standard, it seems that in some environments, the
6013 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6014 program. Exim now does this explicitly; it affects the formatting of
6015 timestamps using strftime().
6017 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6018 complaints. I've now initialized all the variables, to avoid this.
6020 32. Header lines added by a system filter were not being "seen" during
6021 transport-time rewrites.
6023 33. The info_callback() function passed to OpenSSL is set up with type void
6024 (*)(SSL *, int, int), as described somewhere. However, when calling the
6025 function (actually a macro) that sets it up, the type void(*)() is
6026 expected. I've put in a cast to prevent warnings from picky compilers.
6028 34. If a DNS black list lookup found a CNAME record, but there were no A
6029 records associated with the domain it pointed at, Exim crashed.
6031 35. If a DNS black list lookup returned more than one A record, Exim ignored
6032 all but the first. It now scans all returned addresses if a particular IP
6033 value is being sought. In this situation, the contents of the
6034 $dnslist_value variable are a list of all the addresses, separated by a
6037 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6038 to accept a host name and all its aliases if the forward lookup for any of
6039 them yielded the IP address of the incoming connection. Now it accepts only
6040 those names whose forward lookup yields the correct IP address. Any other
6041 names are discarded. This closes a loophole whereby a rogue DNS
6042 administrator could create reverse DNS records to break through a
6043 wildcarded host restriction in an ACL.
6045 37. If a user filter or a system filter that ran in a subprocess used any of
6046 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6047 the wrong values were passed to the pipe command ($thisaddress had the
6048 value of $0, $0 had the value of $1, etc). This bug was introduced by
6049 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6051 38. Improved the line breaking for long SMTP error messages from ACLs.
6052 Previously, if there was no break point between 40 and 75 characters, Exim
6053 left the rest of the message alone. Two changes have been made: (a) I've
6054 reduced the minimum length to 35 characters; (b) if it can't find a break
6055 point between 35 and 75 characters, it looks ahead and uses the first one
6056 that it finds. This may give the occasional overlong line, but at least the
6057 remaining text gets split now.
6059 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6060 file descriptors might be low, and that setting 1000 would always raise it.
6061 It turns out that in some environments, the limit is already over 1000 and
6062 that lowering it causes trouble. So now Exim takes care not to decrease it.
6064 40. When delivering a message, the value of $return_path is set to $sender_
6065 address at the start of routing (routers may change the value). By an
6066 oversight, this default was not being set up when an address was tested by
6067 -bt or -bv, which affected the outcome if any router or filter referred to
6070 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6071 log only when "message" or "log_message" are set. However, if one of the
6072 conditions was an address verification, or a call to a nested ACL, the
6073 messages generated by the underlying test were being passed through. This
6074 no longer happens. The underlying message is available in $acl_verify_
6075 message for both "message" and "log_message" expansions, so it can be
6076 passed through if needed.
6078 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6079 new expansion $bh_ to give the encoded byte string without charset
6080 translation. Translation happens only if iconv() is available; HAVE_ICONV
6081 indicates this at build time. HEADERS_CHARSET gives the charset to
6082 translate to; headers_charset can change it in the configuration, and
6083 "headers charset" can change it in an individual filter file.
6085 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6086 that creates RFC 2047 encoded "words" labels them as that charset instead
6087 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6088 expansion operator; (ii) when Exim creates a From: line for a local
6089 message; (iii) when a header line is rewritten to include a "phrase" part.
6091 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6092 buggy, causing it to skip the first lines of messages whose message ID
6093 ended in 'D'. This would not have bitten before Exim release 4.14, because
6094 message IDs were unlikely to end in 'D' before then. The effect was to have
6095 incorrect size information for certain domains.
6097 45. #include "config.h" was missing at the start of the crypt16.c module. This
6098 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6101 46. If there was a timeout during a "random" callout check, Exim treated it as
6102 a failure of the random address, and carried on sending RSET and the real
6103 address. If the delay was just some slowness somewhere, the response to the
6104 original RCPT would be taken as a response to RSET and so on, causing
6105 mayhem of various kinds.
6107 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6108 when I implemented it. It didn't allow for the fact that some option values
6109 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6110 the right test for positive values.
6112 48. Domain names in DNS records are case-independent. Exim always looks them up
6113 in lower case. Some resolvers return domain names in exactly the case they
6114 appear in the zone file, that is, they may contain uppercase letters. Not
6115 all resolvers do this - some return always lower case. Exim was treating a
6116 change of case by a resolver as a change of domain, similar to a widening
6117 of a domain abbreviation. This triggered its re-routing code and so it was
6118 trying to route what was effectively the same domain again. This normally
6119 caused routing to fail (because the router wouldn't handle the domain
6120 twice). Now Exim checks for this case specially, and just changes the
6121 casing of the domain that it ultimately uses when it transmits the message
6124 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6127 50. If a filter generated a file delivery with a non-absolute name (possible if
6128 no home directory exists for the router), the forbid_file option was not
6131 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6132 the existing equality matching.
6134 52. Exim was using ints instead of ino_t variables in some places where it was
6135 dealing with inode numbers.
6137 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6138 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6139 called TMPDIR, and if it finds it is different, it changes its value.
6141 54. The smtp_printf() function is now made available to local_scan() so
6142 additional output lines can be written before returning. There is also an
6143 smtp_fflush() function to enable the detection of a dropped connection.
6144 The variables smtp_input and smtp_batched_input are exported to
6147 55. Changed the default runtime configuration: the message "Unknown user"
6148 has been removed from the ACL, and instead placed on the localuser router,
6149 using the cannot_route_message feature. This means that any verification
6150 failures that generate their own messages won't get overridden. Similarly,
6151 the "Unrouteable address" message that was in the ACL for unverifiable
6152 relay addresses has also been removed.
6154 56. Added hosts_avoid_esmtp to the smtp transport.
6156 57. The exicyclog script was not checking for the esoteric option
6157 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6158 will work only if exicyclog is run under the appropriate euid.
6160 58. Following a discussion on the list, the rules by which Exim recognises line
6161 endings on incoming messages have been changed. The -dropcr and drop_cr
6162 options are now no-ops, retained only for backwards compatibility. The
6163 following line terminators are recognized: LF CRLF CR. However, special
6164 processing applies to CR:
6166 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6167 nor a local message in the state where . is a terminator.
6169 (ii) If a bare CR is encountered in a header line, an extra space is added
6170 after the line terminator so as not to end the header. The reasoning
6171 behind this is that bare CRs in header lines are most likely either
6172 to be mistakes, or people trying to play silly games.
6174 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6175 was being incorrectly given as 18 bytes larger than it should have been.
6176 This is a VOB (very old bug).
6178 60. This may never have affected anything current, but just in case it has:
6179 When the local host is found other than at the start of a list of hosts,
6180 the local host, those with the same MX, and any that follow, are discarded.
6181 When the list in question was part of a longer list of hosts, the following
6182 hosts (not currently being processed) were also being discarded. This no
6183 longer happens. I'm not sure if this situation could ever has previously
6186 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6189 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6190 header. This is something that is recommended in a new Internet Draft, and
6191 is something that is documented as being done by Sendmail. There are two
6192 possible values. For messages generated by the autoreply transport, Exim
6195 Auto-Submitted: auto-replied
6197 whereas for all other generated messages (e.g. bounces) it adds
6199 Auto-Submitted: auto-generated
6201 63. The "personal" condition in filters now includes a test for the
6202 Auto-Submitted: header. If it contains the string "auto-" the message it
6203 not considered personal.
6205 64. Added rcpt_include_affixes as a generic transport option.
6207 65. Added queue_only_override (default true).
6209 66. Added the syslog_duplication option.
6211 67. If what should have been the first header line of a message consisted of
6212 a space followed by a colon, Exim was mis-interpreting it as a header line.
6213 It isn't of course - it is syntactically invalid and should therefore be
6214 treated as the start of the message body. The misbehaviour could have
6215 caused a number of strange effects, including loss of data in subsequent
6216 header lines, and spool format errors.
6218 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6219 client host had authenticated. This control can now be exercised by an ACL
6220 for more flexibility.
6222 69. By default, callouts do not happen when testing with -bh. There is now a
6223 variant, -bhc, which does actually run the callout code, including
6224 consulting and updating the callout cache.
6226 70. Added support for saslauthd authentication, courtesy of Alexander
6229 71. If statvfs() failed on the spool or log directories while checking their
6230 size for availability, Exim confusingly gave the error "space shortage".
6231 Furthermore, in debugging mode it crashed with a floating point exception.
6232 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6233 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6234 serious problem, Exim now writes to the main and panic logs when this
6235 happens, with details of the failure. It then refuses to accept the
6236 incoming message, giving the message "spool directory problem" or "log
6237 directory problem" with a 421 code for SMTP messages.
6239 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6240 0, 1, and 2 exist, because some OS complain for execs without them (see
6241 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6242 descriptors. However, the code omitted to check that the open succeeded,
6243 causing mysterious errors if for some reason the permissions on /dev/null
6244 got screwed. Now Exim writes a message to the main and panic logs, and
6245 bombs out if it can't open /dev/null.
6247 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6248 interact so that it is all more flexible. It is supposed to remain
6249 backwards compatible. Also added extra_local_interfaces.
6251 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6252 to bomb out with an assertion failure - to the client this appears as a
6253 connection drop. This problem occurs in the part of the code that was taken
6254 from the Samba project. Fortunately, the assertion is in a very simple
6255 function, so I have fixed this by reproducing the function inline in the
6256 one place where it is called, and arranging for authentication to fail
6257 instead of killing the process with assert().
6259 75. The SPA client code was not working when the server requested OEM rather
6260 than Unicode encoding.
6262 76. Added code to make require_files with a specific uid setting more usable in
6263 the case where statting the file as root fails - usually a non-root-mounted
6264 NFS file system. When this happens and the failure is EACCES, Exim now
6265 forks a subprocess and does the per-uid checking as the relevant uid.
6267 77. Added process_log_path.
6269 78. If log_file_path was not explicitly set, a setting of check_log_space or
6270 check_log_inodes was ignored.
6272 79. If a space check for the spool or log partitions fails, the incident is now
6273 logged. Of course, in the latter case the data may get lost...
6275 80. Added the %p formatting code to string_format() so that it can be used to
6276 print addresses in debug_print(). Adjusted all the address printing in the
6277 debugging in store.c to use %p rather than %d.
6279 81. There was a concern that a line of code in smtp_in.c could overflow a
6280 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6281 initially expressed, the concern was not well-founded, because trailing
6282 spaces are removed early. However, if the trailing spaces were followed by
6283 a NULL, they did not get removed, so the overflow was possible. Two fixes
6286 (a) I re-wrote the offending code in a cleaner fashion.
6287 (b) If an incoming SMTP command contains a NULL character, it is rejected
6290 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6291 runs initgroups(), so that if the Exim user is in any additional groups,
6292 they will be used during message reception.
6298 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob