1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
8 and 3.4.4 - once the server is enabled to respond to an OCSP request
9 it does even when not requested, resulting in a stapling non-aware
10 client dropping the TLS connection.
12 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
13 support variable-length bit vectors. No functional change.
15 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
16 The I= interface field on outgoing lines is now after the H= remote
17 host field, same as incoming lines. There is a separate
18 outgoing_interface log selector which allows you to disable the
24 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
27 JH/02 The smtp transport option "multi_domain" is now expanded.
29 JH/03 The smtp transport now requests PRDR by default, if the server offers
32 JH/04 Certificate name checking on server certificates, when exim is a client,
33 is now done by default. The transport option tls_verify_cert_hostnames
34 can be used to disable this per-host. The build option
35 EXPERIMENTAL_CERTNAMES is withdrawn.
37 JH/05 The value of the tls_verify_certificates smtp transport and main options
38 default to the word "system" to access the system default CA bundle.
39 For GnuTLS, only version 3.0.20 or later.
41 JH/06 Verification of the server certificate for a TLS connection is now tried
42 (but not required) by default. The verification status is now logged by
43 default, for both outbound TLS and client-certificate supplying inbound
46 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
49 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
50 Status Notification (bounce) messages are now MIME format per RFC 3464.
51 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
52 under the control of the dsn_advertise_hosts option, and routers may
53 have a dsn_lasthop option.
55 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
56 default, modifiable by a malware= option. The list separator for
57 the options can now be changed in the usual way. Bug 68.
59 JH/10 The smtp_receive_timeout main option is now expanded before use.
61 JH/11 The incoming_interface log option now also enables logging of the
62 local interface on delivery outgoing connections.
64 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
65 if the interface and destination host and port all match.
67 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
70 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
71 Patch from Andrew Lewis.
73 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
74 now supports optional time-restrictions, weighting, and priority
75 modifiers per server. Patch originally by <rommer@active.by>.
77 JH/16 The spamd_address main option now supports a mixed list of local
78 and remote servers. Remote servers can be IPv6 addresses, and
81 JH/17 Bug 68: The spamd_address main option now supports an optional
82 timeout value per server.
84 JH/18 Bug 1581: Router and transport options headers_add/remove can
85 now have the list separator specified.
87 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
90 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
93 JH/21 Support for the A6 type of dns record is withdrawn.
95 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
96 rather than the verbs used.
98 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
99 from 255 to 1024 chars.
101 JH/24 Verification callouts now attempt to use TLS by default.
103 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
104 are generic router options now. The defaults didn't change.
106 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
107 Original patch from Alexander Shikoff, worked over by JH.
109 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
110 files if ZCAT_COMMAND is not executable.
112 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
114 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
116 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
117 Normally benign, it bites when the pair was led to by a CNAME;
118 modern usage is to not canoicalize the domain to a CNAME target
119 (and we were inconsistent anyway for A-only vs AAAA+A).
121 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
123 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
124 when evaluating $sender_host_dnssec.
126 JH/31 Check the HELO verification lookup for DNSSEC, adding new
127 $sender_helo_dnssec variable.
129 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
131 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
133 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
135 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
136 documented as working, but never had. Support all but $spam_report.
138 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
139 added for tls authenticator.
144 TL/01 When running the test suite, the README says that variables such as
145 no_msglog_check are global and can be placed anywhere in a specific
146 test's script, however it was observed that placement needed to be near
147 the beginning for it to behave that way. Changed the runtest perl
148 script to read through the entire script once to detect and set these
149 variables, reset to the beginning of the script, and then run through
150 the script parsing/test process like normal.
152 TL/02 The BSD's have an arc4random API. One of the functions to induce
153 adding randomness was arc4random_stir(), but it has been removed in
154 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
155 function when detected.
157 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
158 cause callback expansion.
160 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
161 syntax errors in an expansion can be treated as a string instead of
162 logging or causing an error, due to the internal use of bool_lax
163 instead of bool when processing it.
165 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
166 server certificates when making smtp deliveries.
168 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
170 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
172 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
174 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
175 Merged patch from Sebastian Wiedenroth.
177 JH/05 Fix results-pipe from transport process. Several recipients, combined
178 with certificate use, exposed issues where response data items split
179 over buffer boundaries were not parsed properly. This eventually
180 resulted in duplicates being sent. This issue only became common enough
181 to notice due to the introduction of conection certificate information,
182 the item size being so much larger. Found and fixed by Wolfgang Breyha.
184 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
185 size buffer was used, resulting in syntax errors when an expansion
188 JH/07 Add support for directories of certificates when compiled with a GnuTLS
189 version 3.3.6 or later.
191 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
192 is EXPERIMENTAL_EVENT, the main-configuration and transport options
193 both become "event_action", the variables become $event_name, $event_data
194 and $event_defer_errno. There is a new variable $verify_mode, usable in
195 routers, transports and related events. The tls:cert event is now also
196 raised for inbound connections, if the main configuration event_action
199 TL/06 In test suite, disable OCSP for old versions of openssl which contained
200 early OCSP support, but no stapling (appears to be less than 1.0.0).
202 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
203 server certificate names available under the smtp transport option
204 "tls_verify_cert_hostname" now do not permit multi-component wildcard
207 JH/10 Time-related extraction expansions from certificates now use the main
208 option "timezone" setting for output formatting, and are consistent
209 between OpenSSL and GnuTLS compilations. Bug 1541.
211 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
212 encoded parameter in the incoming message. Bug 1558.
214 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
215 include certificate info, eximon was claiming there were spoolfile
218 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
220 JH/14 Log delivery-related information more consistently, using the sequence
221 "H=<name> [<ip>]" wherever possible.
223 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
224 are problematic for Debian distribution, omit them from the release
227 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
229 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
232 JH/17 Fix a null-indirection in certextract expansions when a nondefault
233 output list separator was used.
238 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
239 checkers that were complaining about end of non-void function with no
242 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
243 This was a regression intruduced in 4.83 by another bugfix.
245 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
247 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
248 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
254 TF/01 Correctly close the server side of TLS when forking for delivery.
256 When a message was received over SMTP with TLS, Exim failed to clear up
257 the incoming connection properly after forking off the child process to
258 deliver the message. In some situations the subsequent outgoing
259 delivery connection happened to have the same fd number as the incoming
260 connection previously had. Exim would try to use TLS and fail, logging
261 a "Bad file descriptor" error.
263 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
264 utilities have not been installed.
266 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
267 temporary space as the ACL may create new global variables.
269 TL/01 LDAP support uses per connection or global context settings, depending
270 upon the detected version of the libraries at build time.
272 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
273 to extract and use the src ip:port in logging and expansions as if it
274 were a direct connection from the outside internet. PPv2 support was
275 updated based on HAProxy spec change in May 2014.
277 JH/02 Add ${listextract {number}{list}{success}{fail}}.
279 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
280 Properly escape header and check for NULL return.
282 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
285 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
287 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
288 characters in header names, implemented as a verify condition.
289 Contributed by Michael Fischer v. Mollard.
291 TL/05 Rename SPF condition results err_perm and err_temp to standardized
292 results permerror and temperror. Previous values are deprecated but
293 still accepted. In a future release, err_perm and err_temp will be
294 completely removed, which will be a backward incompatibility if the
295 ACL tests for either of these two old results. Patch contributed by
296 user bes-internal on the mailing list.
298 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
300 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
301 selectors, in both main and reject logs.
303 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
306 JH/07 Add malware type "sock" for talking to simple daemon.
308 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
310 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
311 routers/transports under cutthrough routing.
313 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
314 numbers. Touch up "bool" conditional to keep the same definition.
316 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
318 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
320 JH/12 Expand items in router/transport headers_add or headers_remove lists
321 individually rather than the list as a whole. Bug 1452.
323 Required for reasonable handling of multiple headers_ options when
324 they may be empty; requires that headers_remove items with embedded
325 colons must have them doubled (or the list-separator changed).
327 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
328 view the policy declared in the DMARC record. Currently, $dmarc_status
329 is a combined value of both the record presence and the result of the
332 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
334 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
335 dnslookup router and the smtp transport (applying to the forward
338 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
339 of ldap servers used for a specific lookup. Patch provided by Heiko
342 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
343 New variable $lookup_dnssec_authenticated for observability.
345 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
346 Patch submitted by Lars Timman.
348 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
350 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
351 Requires trusted mode and valid format message id, aborts otherwise.
352 Patch contributed by Heiko Schlichting.
354 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
355 certextract with support for various fields. Bug 1358.
357 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
358 is requested by default, modifiable by smtp transport option
361 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
362 operate on certificate variables to give certificate fingerprints
363 Also new ${sha256:cert_variable}.
365 JH/23 The PRDR feature is moved from being Experimental into the mainline.
367 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
368 Christian Aistleitner.
370 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
372 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
373 file. Patch from Wolfgang Breyha.
375 JH/25 Expand the coverage of the delivery $host and $host_address to
376 client authenticators run in verify callout. Bug 1476.
378 JH/26 Port service names are now accepted for tls_on_connect_ports, to
379 align with daemon_smtp_ports. Bug 72.
381 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
382 support and error reporting did not work properly.
384 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
385 and is readable. Patch from Andrew Colin Kissa.
387 TL/14 Enhance documentation of ${run expansion and how it parses the
388 commandline after expansion, particularly in the case when an
389 unquoted variable expansion results in an empty value.
391 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
393 PP/02 Fix internal collision of T_APL on systems which support RFC3123
394 by renaming away from it. Addresses GH issue 15, reported by
397 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
399 TL/15 SECURITY: prevent double expansion in math comparison functions
400 (can expand unsanitized data). Not remotely exploitable.
407 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
409 PP/02 Make -n do something, by making it not do something.
410 When combined with -bP, the name of an option is not output.
412 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
415 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
416 $sender_host_name and config options to manage this, and basic check
419 PP/05 DSCP support for outbound connections and control modifier for inbound.
421 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
422 (Only plugin which currently uses this is kerberos4, which nobody should
423 be using, but we should make it available and other future plugins might
424 conceivably use it, even though it would break NAT; stuff *should* be
425 using channel bindings instead).
427 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
428 name; added for Sendmail compatibility; requires admin caller.
429 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
430 just ignore it); requires trusted caller.
431 Also parse but ignore: -Ac -Am -X<logfile>
434 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
436 TL/02 Add +smtp_confirmation as a default logging option.
438 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
439 Patch by Magnus Holmgren from 2007-02-20.
441 TL/04 Bugzilla 1281 - Spec typo.
442 Bugzilla 1283 - Spec typo.
443 Bugzilla 1290 - Spec grammar fixes.
445 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
447 TL/06 Add Experimental DMARC support using libopendmarc libraries.
449 TL/07 Fix an out of order global option causing a segfault. Reported to dev
450 mailing list by by Dmitry Isaikin.
452 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
454 JH/02 Support "G" suffix to numbers in ${if comparisons.
456 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
458 NM/01 Bugzilla 1197 - Spec typo
459 Bugzilla 1196 - Spec examples corrections
461 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
463 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
464 gnutls_enable_pkcs11, but renamed to more accurately indicate its
467 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
468 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
470 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
471 "acl {{name}{arg}...}", and optional args on acl condition
474 JH/05 Permit multiple router/transport headers_add/remove lines.
476 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
478 JH/07 Avoid using a waiting database for a single-message-only transport.
479 Performance patch from Paul Fisher. Bugzilla 1262.
481 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
484 JH/09 Add $headers_added variable, with content from use of ACL modifier
485 add_header (but not yet added to the message). Bugzilla 199.
487 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
488 Pulled from Bugzilla 817 by Wolfgang Breyha.
490 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
492 (nb: this is the same fix as in Exim 4.80.1)
494 JH/11 Add A= logging on delivery lines, and a client_set_id option on
497 JH/12 Add optional authenticated_sender logging to A= and a log_selector
500 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
502 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
503 advertise SMTP AUTH mechanism to us, instead of a generic
504 protocol violation error. Also, make Exim more robust to bad
505 data from the Dovecot auth socket.
507 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
509 When a queue runner is handling a message, Exim first routes the
510 recipient addresses, during which it prunes them based on the retry
511 hints database. After that it attempts to deliver the message to
512 any remaining recipients. It then updates the hints database using
515 So if a recipient address works intermittently, it can get repeatedly
516 deferred at routing time. The retry hints record remains fresh so the
517 address never reaches the final cutoff time.
519 This is a fairly common occurrence when a user is bumping up against
520 their storage quota. Exim had some logic in its local delivery code
521 to deal with this. However it did not apply to per-recipient defers
522 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
524 This change adds a proper retry rule check during routing so that the
525 final cutoff time is checked against the message's age. We only do
526 this check if there is an address retry record and there is not a
527 domain retry record; this implies that previous attempts to handle
528 the address had the retry_use_local_parts option turned on. We use
529 this as an approximation for the destination being like a local
530 delivery, as in LMTP.
532 I suspect this new check makes the old local delivery cutoff check
533 redundant, but I have not verified this so I left the code in place.
535 TF/02 Correct gecos expansion when From: is a prefix of the username.
537 Test 0254 submits a message to Exim with the header
541 When I ran the test suite under the user fanf2, Exim expanded
542 the header to contain my full name, whereas it should have added
543 a Resent-Sender: header. It erroneously treats any prefix of the
544 username as equal to the username.
546 This change corrects that bug.
548 GF/01 DCC debug and logging tidyup
549 Error conditions log to paniclog rather than rejectlog.
550 Debug lines prefixed by "DCC: " to remove any ambiguity.
552 TF/03 Avoid unnecessary rebuilds of lookup-related code.
554 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
555 Bug spotted by Jeremy Harris; was flawed since initial commit.
556 Would have resulted in OCSP responses post-SNI triggering an Exim
557 NULL dereference and crash.
559 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
561 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
562 Bug detection, analysis and fix by Samuel Thibault.
563 Bugzilla 1331, Debian bug #698092.
565 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
567 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
568 Server implementation by Todd Lyons, client by JH.
569 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
570 config variable "prdr_enable" controls whether the server
571 advertises the facility. If the client requests PRDR a new
572 acl_data_smtp_prdr ACL is called once for each recipient, after
573 the body content is received and before the acl_smtp_data ACL.
574 The client is controlled by bolth of: a hosts_try_prdr option
575 on the smtp transport, and the server advertisement.
576 Default client logging of deliveries and rejections involving
577 PRDR are flagged with the string "PRDR".
579 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
580 fclose(). Diagnosis by Todd Lyons.
582 PP/17 Update configure.default to handle IPv6 localhost better.
583 Patch by Alain Williams (plus minor tweaks).
586 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
587 This is now consistent with GnuTLS, and is now documented: the
588 previous undocumented portable approach to treating the option as
589 unset was to force an expansion failure. That still works, and
590 an empty string is now equivalent.
592 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
593 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
594 not performing validation itself.
596 PP/20 Added force_command boolean option to pipe transport.
597 Patch from Nick Koston, of cPanel Inc.
599 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
602 TF/04 Added udpsend ACL modifer and hexquote expansion operator
604 PP/21 Fix eximon continuous updating with timestamped log-files.
605 Broken in a format-string cleanup in 4.80, missed when I repaired the
606 other false fix of the same issue.
607 Report and fix from Heiko Schlichting.
610 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
611 Report from Prashanth Katuri.
613 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
614 It's SecureTransport, so affects any MacOS clients which use the
615 system-integrated TLS libraries, including email clients.
617 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
618 using a MIME ACL for non-SMTP local injection.
619 Report and assistance in diagnosis by Warren Baker.
621 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
623 JH/16 Fix comparisons for 64b. Bugzilla 1385.
625 TL/09 Add expansion variable $authenticated_fail_id to keep track of
626 last id that failed so it may be referenced in subsequent ACL's.
628 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
631 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
632 ldap library initialization, allowing self-signed CA's to be
633 used. Also properly sets require_cert option later in code by
634 using NULL (global ldap config) instead of ldap handle (per
635 session). Bug diagnosis and testing by alxgomz.
637 TL/12 Enhanced documentation in the ratelimit.pl script provided in
638 the src/util/ subdirectory.
640 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
641 renamed to Transport Post Delivery Action by Jeremy Harris, as
644 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
645 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
646 redis_servers = needs to be configured which will be used by the redis
647 lookup. Patch from Warren Baker, of The Packet Hub.
649 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
651 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
652 hostname or reverse DNS when processing a host list. Used suggestions
653 from multiple comments on this bug.
655 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
657 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
658 Missed a few lines, added it to make the runtest require no keyboard
661 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
662 contains upper case chars. Make router use caseful_local_part.
664 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
665 support when GnuTLS has been built with p11-kit.
671 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
673 This, or similar/improved, will also be change PP/11 of 4.82.
679 PP/01 Handle short writes when writing local log-files.
680 In practice, only affects FreeBSD (8 onwards).
681 Bugzilla 1053, with thanks to Dmitry Isaikin.
683 NM/01 Bugzilla 949 - Documentation tweak
685 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
688 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
690 PP/02 Implemented gsasl authenticator.
692 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
694 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
695 `pkg-config foo` for cflags/libs.
697 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
698 with rest of GSASL and with heimdal_gssapi.
700 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
701 `pkg-config foo` for cflags/libs for the TLS implementation.
703 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
704 properties get this fed in as external SSF. A number of robustness
705 and debugging improvements to the cyrus_sasl authenticator.
707 PP/08 cyrus_sasl server now expands the server_realm option.
709 PP/09 Bugzilla 1214 - Log authentication information in reject log.
710 Patch by Jeremy Harris.
712 PP/10 Added dbmjz lookup type.
714 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
716 PP/12 MAIL args handles TAB as well as SP, for better interop with
717 non-compliant senders.
718 Analysis and variant patch by Todd Lyons.
720 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
721 Bug report from Lars Müller <lars@samba.org> (via SUSE),
722 Patch from Dirk Mueller <dmueller@suse.com>
724 PP/13 tls_peerdn now print-escaped for spool files.
725 Observed some $tls_peerdn in wild which contained \n, which resulted
726 in spool file corruption.
728 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
729 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
730 or write after TLS renegotiation, which otherwise led to messages
733 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
734 as a tracking header (ie: a signed header comes before the signature).
735 Patch from Wolfgang Breyha.
737 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
738 comma-sep list; embedded commas doubled.
740 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
742 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
744 Report and patch from Dmitry Banschikov.
746 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
747 Removed SSL_clear() after SSL_new() which led to protocol negotiation
748 failures. We appear to now support TLS1.1+ with Exim.
750 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
751 lets Exim select keys and certificates based upon TLS SNI from client.
752 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
753 before an outbound SMTP session. New log_selector, +tls_sni.
755 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
756 NULL dereference. Report and patch from Alun Jones.
758 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
759 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
760 Not seeing resolver debug output on NetBSD, but suspect this is a
761 resolver implementation change.
763 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
764 Left warnings. Added "eximon gdb" invocation mode.
766 PP/21 Defaulting "accept_8bitmime" to true, not false.
768 PP/22 Added -bw for inetd wait mode support.
770 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
771 locate the relevant includes and libraries. Made this the default.
773 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
774 Bugzilla 1246, report and most of solution from Tomasz Kusy.
776 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
777 This may cause build issues on older platforms.
779 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
780 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
781 gnutls_require_mac & gnutls_require_protocols (no longer supported).
782 Added SNI support via GnuTLS too.
783 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
785 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
787 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
788 Applied second patch from Janne, implementing suggestion to default
789 multiple-strings-in-record handling to match SPF spec.
791 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
793 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
794 read-only, out of scope).
795 Patch from Wolfgang Breyha, report from Stuart Northfield.
797 PP/29 Fix three issues highlighted by clang analyser static analysis.
798 Only crash-plausible issue would require the Cambridge-specific
799 iplookup router and a misconfiguration.
800 Report from Marcin Mirosław.
802 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
804 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
805 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
806 As part of this, removing so much warning spew let me fix some minor
807 real issues in debug logging.
809 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
810 assignment on my part. Fixed.
812 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
813 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
814 Janne Snabb (who went above and beyond: thank you).
816 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
817 string otherwise requires a connection and a bunch more work and it's
818 relatively easy to get wrong. Should also expose TLS library linkage
821 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
822 64-bit ${eval} (JH/03).
824 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
825 GNU libc to support some of the 64-bit stuff, should not lead to
826 conflicts. Defined before os.h is pulled in, so if a given platform
827 needs to override this, it can.
829 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
830 protection layer was required, which is not implemented.
831 Bugzilla 1254, patch from Wolfgang Breyha.
833 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
834 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
835 tls_dhparam take prime identifiers. Also unbreak combination of
836 OpenSSL+DH_params+TLSSNI.
838 PP/39 Disable SSLv2 by default in OpenSSL support.
844 PP/01 Solaris build fix for Oracle's LDAP libraries.
845 Bugzilla 1109, patch from Stephen Usher.
847 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
849 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
852 TF/02 Fix a couple more cases where we did not log the error message
853 when unlink() failed. See also change 4.74-TF/03.
855 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
856 lock up or crash if it happened to be inside a call to libc when it
857 got a SIGUSR1 from exiwhat.
859 The SIGUSR1 handler appends the current process status to the process
860 log which is later printed by exiwhat. It used to use the general
861 purpose logging code to do this, but several functions it calls are
862 not safe for signals.
864 The new output code in the SIGUSR1 handler is specific to the process
865 log, and simple enough that it's easy to inspect for signal safety.
866 Removing some special cases also simplifies the general logging code.
867 Removing the spurious timestamps from the process log simplifies
870 TF/04 Improved ratelimit ACL condition.
872 The /noupdate option has been deprecated in favour of /readonly which
873 has clearer semantics. The /leaky, /strict, and /readonly update modes
874 are mutually exclusive. The update mode is no longer included in the
875 database key; it just determines when the database is updated. (This
876 means that when you upgrde Exim will forget old rate measurements.)
878 Exim now checks that the per_* options are used with an update mode that
879 makes sense for the current ACL. For example, when Exim is processing a
880 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
881 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
882 must specify per_mail/readonly. If you omit the update mode it defaults to
883 /leaky where that makes sense (as before) or /readonly where required.
885 The /noupdate option is now undocumented but still supported for
886 backwards compatibility. It is equivalent to /readonly except that in
887 ACLs where /readonly is required you may specify /leaky/noupdate or
888 /strict/noupdate which are treated the same as /readonly.
890 A useful new feature is the /count= option. This is a generalization
891 of the per_byte option, so that you can measure the throughput of other
892 aggregate values. For example, the per_byte option is now equivalent
893 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
895 The per_rcpt option has been generalized using the /count= mechanism
896 (though it's more complicated than the per_byte equivalence). When it is
897 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
898 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
899 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
900 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
901 Note that using per_rcpt with a non-readonly update mode in more than
902 one ACL will cause the recipients to be double-counted. (The per_mail
903 and per_byte options don't have this problem.)
905 The handling of very low rates has changed slightly. If the computed rate
906 is less than the event's count (usually one) then this event is the first
907 after a long gap. In this case the rate is set to the same as this event's
908 count, so that the first message of a spam run is counted properly.
910 The major new feature is a mechanism for counting the rate of unique
911 events. The new per_addr option counts the number of different
912 recipients that someone has sent messages to in the last time period. It
913 behaves like per_rcpt if all the recipient addresses are different, but
914 duplicate recipient addresses do not increase the measured rate. Like
915 the /count= option this is a general mechanism, so the per_addr option
916 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
917 example, measure the rate that a client uses different sender addresses
918 with the options per_mail/unique=$sender_address. There are further
919 details in the main documentation.
921 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
923 TF/06 Removed a few PCRE remnants.
925 TF/07 Automatically extract Exim's version number from tags in the git
926 repository when doing development or release builds.
928 PP/02 Raise smtp_cmd_buffer_size to 16kB.
929 Bugzilla 879. Patch from Paul Fisher.
931 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
932 Heavily based on revision 40f9a89a from Simon Arlott's tree.
935 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
937 PP/05 Variable $av_failed, true if the AV scanner deferred.
938 Bugzilla 1078. Patch from John Horne.
940 PP/06 Stop make process more reliably on build failure.
941 Bugzilla 1087. Patch from Heiko Schlittermann.
943 PP/07 Make maildir_use_size_file an _expandable_ boolean.
944 Bugzilla 1089. Patch from Heiko Schlittermann.
946 PP/08 Handle ${run} returning more data than OS pipe buffer size.
947 Bugzilla 1131. Patch from Holger Weiß.
949 PP/09 Handle IPv6 addresses with SPF.
950 Bugzilla 860. Patch from Wolfgang Breyha.
952 PP/10 GnuTLS: support TLS 1.2 & 1.1.
954 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
957 PP/11 match_* no longer expand right-hand-side by default.
958 New compile-time build option, EXPAND_LISTMATCH_RHS.
959 New expansion conditions, "inlist", "inlisti".
961 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
963 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
965 PP/14 fix log_write() format string regression from TF/03.
966 Bugzilla 1152. Patch from Dmitry Isaikin.
972 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
974 PP/02 Harmonised TLS library version reporting; only show if debugging.
975 Layout now matches that introduced for other libraries in 4.74 PP/03.
977 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
979 PP/04 New "dns_use_edns0" global option.
981 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
984 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
985 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
987 TK/01 Updated PolarSSL code to 0.14.2.
988 Bugzilla 1097. Patch from Andreas Metzler.
990 PP/07 Catch divide-by-zero in ${eval:...}.
993 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
996 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
997 format-string attack -- SECURITY: remote arbitrary code execution.
999 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1000 time unintentionally subject to list matching rules, letting the header
1001 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1002 arbitrary string expansion). This allowed for information disclosure.
1004 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1005 INT_MIN/-1 -- value coerced to INT_MAX.
1011 NM/01 Workround for PCRE version dependency in version reporting
1014 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1015 This fixes portability to compilers other than gcc, notably
1016 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1018 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1019 makefiles for portability to HP-UX and POSIX correctness.
1021 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1022 Also via indented variable definition in the Makefile.
1023 (Debugging by Oliver Heesakkers).
1025 PP/02 Restore caching of spamd results with expanded spamd_address.
1026 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1028 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1029 Improves build reliability. Fix from: Frank Elsner
1031 NM/02 Fix wide character breakage in the rfc2047 coding
1032 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1034 NM/03 Allow underscore in dnslist lookups
1035 Fixes bug 1026. Patch from Graeme Fowler
1037 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1038 Code patches from Adam Ciarcinski of NetBSD.
1040 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1043 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1044 is logged, to avoid truncation. Patch from John Horne.
1046 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1047 Patch from Jakob Hirsch.
1049 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1050 SQL string expansion failure details.
1051 Patch from Andrey Oktyabrski.
1053 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1054 Patch from Simon Arlott.
1056 PP/09 New lookups functionality failed to compile on old gcc which rejects
1057 extern declarations in function scope.
1058 Patch from Oliver Fleischmann
1060 PP/10 Use sig_atomic_t for flags set from signal handlers.
1061 Check getgroups() return and improve debugging.
1062 Fixed developed for diagnosis in bug 927 (which turned out to be
1065 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1066 Patch from Mark Zealey.
1068 PP/12 Bugzilla 1056: Improved spamd server selection.
1069 Patch from Mark Zealey.
1071 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1072 Based on patch from Heiko Schlittermann.
1074 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1075 Patch from Uwe Doering, sign-off by Michael Haardt.
1077 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1078 variable declaration deep within a block. Bug and patch from
1081 PP/15 lookups-Makefile IRIX compatibilty coercion.
1083 PP/16 Make DISABLE_DKIM build knob functional.
1085 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1086 Patch by Simon Arlott
1088 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1089 variable argument macros. Our copy now differs from upstream.
1095 TF/01 Failure to get a lock on a hints database can have serious
1096 consequences so log it to the panic log.
1098 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1099 controlled using the smtp_confirmation log selector.
1101 TF/03 Include the error message when we fail to unlink a spool file.
1103 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1104 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1105 for maintaining out-of-tree patches for some time.
1107 PP/01 Bugzilla 139: Documentation and portability issues.
1108 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1109 Handle per-OS dynamic-module compilation flags.
1111 PP/02 Let /dev/null have normal permissions.
1112 The 4.73 fixes were a little too stringent and complained about the
1113 permissions on /dev/null. Exempt it from some checks.
1114 Reported by Andreas M. Kirchwitz.
1116 PP/03 Report version information for many libraries, including
1117 Exim version information for dynamically loaded libraries. Created
1118 version.h, now support a version extension string for distributors
1119 who patch heavily. Dynamic module ABI change.
1121 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1122 privilege escalation vulnerability whereby the Exim run-time user
1123 can cause root to append content of the attacker's choosing to
1126 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1129 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1130 If dropping privileges for untrusted macros, we disabled normal logging
1131 on the basis that it would fail; for the Exim run-time user, this is not
1132 the case, and it resulted in successful deliveries going unlogged.
1133 Fixed. Reported by Andreas Metzler.
1139 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1140 only prepend for the Resent-* case. Fixes regression introduced in
1141 Exim 4.70 by NM/22 for Bugzilla 607.
1143 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1144 increasing numbers of administrators be bitten by this.
1146 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1148 PP/03 Bugzilla 994: added openssl_options main configuration option.
1150 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1152 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1154 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1156 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1157 without a peer certificate, leading to a segfault because of an
1158 assumption that peers always have certificates. Be a little more
1159 paranoid. Problem reported by Martin Tscholak.
1161 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1162 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1163 NB: ClamAV planning to remove STREAM in "middle of 2010".
1164 CL also introduces -bmalware, various -d+acl logging additions and
1165 more caution in buffer sizes.
1167 PP/09 Implemented reverse_ip expansion operator.
1169 PP/10 Bugzilla 937: provide a "debug" ACL control.
1171 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1173 PP/12 Bugzilla 973: Implement --version.
1175 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1177 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1179 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1181 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1182 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1183 ignore trailing whitespace.
1185 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1187 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1190 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1191 Notification from Dr Andrew Aitchison.
1193 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1194 ExtendedDetectionInfo response format.
1195 Notification from John Horne.
1197 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1200 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1201 XSL and documented dependency on system catalogs, with examples of how
1204 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1207 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1208 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1209 configuration file which is writeable by the Exim user or group.
1211 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1212 of configuration files to cover files specified with the -C option if
1213 they are going to be used with root privileges, not just the default
1216 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1217 option (effectively making it always true).
1219 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1220 files to be used while preserving root privileges.
1222 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1223 that rogue child processes cannot use them.
1225 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1226 run-time user, instead of root.
1228 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1229 Exim run-time user without dropping privileges.
1231 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1232 result string, instead of calling string_vformat() twice with the same
1235 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1236 for other users. Others should always drop root privileges if they use
1237 -C on the command line, even for a whitelisted configure file.
1239 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1241 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1247 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1248 $data_path, and $header_path variables; fixed documentation bugs and
1251 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1252 exipick to access non-standard spools, including the "frozen" queue
1255 NM/01 Bugzilla 965: Support mysql stored procedures.
1256 Patch from Alain Williams
1258 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1260 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1261 Patch from Andreas Metzler
1263 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1264 Patch from Kirill Miazine
1266 NM/05 Bugzilla 671: Added umask to procmail example.
1268 JJ/03 installed exipick 20100323.0, fixing doc bug
1270 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1271 directory. Notification and patch from Dan Rosenberg.
1273 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1275 TK/02 Improve log output when DKIM signing operation fails.
1277 MH/01 Treat the transport option dkim_domain as a colon separated
1278 list, not as a single string, and sign the message with each element,
1279 omitting multiple occurences of the same signer.
1281 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1282 Bugzilla 985, 986. Patch by Simon Arlott
1284 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1285 Patch by Simon Arlott
1287 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1288 MBX locking. Notification from Dan Rosenberg.
1294 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1296 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1298 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1300 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1302 NM/04 Bugzilla 915: Flag broken perl installation during build.
1308 TK/01 Added patch by Johannes Berg that expands the main option
1309 "spamd_address" if it starts with a dollar sign.
1311 TK/02 Write list of recipients to X-Envelope-Sender header when building
1312 the mbox-format spool file for content scanning (suggested by Jakob
1315 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1316 (http://www.dcc-servers.net/) support via dccifd. Activated by
1317 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1319 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1320 by Mark Daniel Reidel <mr@df.eu>.
1322 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1323 When building exim an external PCRE library is now needed -
1324 PCRE is a system library on the majority of modern systems.
1325 See entry on PCRE_LIBS in EDITME file.
1327 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1328 conversation. Added nologin parameter to request.
1329 Patch contributed by Kirill Miazine.
1331 TF/01 Do not log submission mode rewrites if they do not change the address.
1333 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1335 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1336 log files in place. Contributed by Roberto Lima.
1338 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1340 TF/03 Bugzilla 615: When checking the local_parts router precondition
1341 after a local_part_suffix or local_part_prefix option, Exim now
1342 does not use the address's named list lookup cache, since this
1343 contains cached lookups for the whole local part.
1345 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1346 Robert Millan. Documentation is in experimental-spec.txt.
1348 TF/04 Bugzilla 668: Fix parallel build (make -j).
1350 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1352 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1353 Patch provided by Jan Srzednicki.
1355 TF/05 Leading white space used to be stripped from $spam_report which
1356 wrecked the formatting. Now it is preserved.
1358 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1359 that they are available at delivery time.
1361 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1363 TF/08 TLS error reporting now respects the incoming_interface and
1364 incoming_port log selectors.
1366 TF/09 Produce a more useful error message if an SMTP transport's hosts
1367 setting expands to an empty string.
1369 NM/06 Bugzilla 744: EXPN did not work under TLS.
1370 Patch provided by Phil Pennock.
1372 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1373 Patch provided by Richard Godbee.
1375 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1376 acl_smtp_notquit, added index entry.
1378 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1379 Patch provided by Eugene Bujak.
1381 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1382 accept(). Patch provided by Maxim Dounin.
1384 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1385 Patch provided by Phil Pennock.
1387 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1389 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1390 Patch provided by Brad "anomie" Jorsch.
1392 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1393 Patch provided by Dean Brooks.
1395 TK/05 Add native DKIM support (does not depend on external libraries).
1397 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1398 Patch provided by Graeme Fowler.
1400 NM/16 Bugzilla 851: Documentation example syntax fix.
1402 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1404 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1407 NM/19 Bugzilla 745: TLS version reporting.
1408 Patch provided by Phil Pennock.
1410 NM/20 Bugzilla 167: bool: condition support.
1411 Patch provided by Phil Pennock.
1413 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1414 clients. Patch provided by Phil Pennock.
1416 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1417 Patch provided by Brad "anomie" Jorsch.
1419 NM/23 Bugzilla 687: Fix misparses in eximstats.
1420 Patch provided by Heiko Schlittermann.
1422 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1423 Patch provided by Heiko Schlittermann.
1425 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1426 plus update to original patch.
1428 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1430 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1431 Patch provided by David Brownlee.
1433 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1435 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1437 NM/30 Bugzilla 888: TLS documentation bugfixes.
1439 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1441 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1442 Unlike the original bugzilla I have changed all shell scripts in src tree.
1444 NM/33 Bugzilla 898: Transport filter timeout fix.
1445 Patch by Todd Rinaldo.
1447 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1448 Patch by Serge Demonchaux.
1450 NM/35 Bugzilla 39: Base64 decode bug fixes.
1451 Patch by Jakob Hirsch.
1453 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1455 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1457 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1459 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1465 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1466 ALT-N's libdkim that I have put here:
1467 http://duncanthrax.net/exim-experimental/
1469 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1470 were called 'true' and it seems that C99 defines that as a reserved
1471 keyword to be used with 'bool' variable types. That means you could
1472 not include C99-style headers which use bools without triggering
1473 build errors in sieve.c.
1475 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1476 as mailq or other aliases. Changed the --help handling significantly
1477 to do whats expected. exim_usage() emits usage/help information.
1479 SC/01 Added the -bylocaldomain option to eximstats.
1481 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1483 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1485 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1491 PH/01 Another patch from the Sieve maintainer.
1493 PH/02 When an IPv6 address is converted to a string for single-key lookup
1494 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1495 dots are used instead of colons so that keys in lsearch files need not
1496 contain colons. This was done some time before quoting was made available
1497 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1498 (notated using the quote facility) so as to distinguish them from IPv4
1499 keys. This meant that lookups for IP addresses in host lists did not work
1500 for iplsearch lookups.
1502 This has been fixed by arranging for IPv6 addresses to be expressed with
1503 colons if the lookup type is iplsearch. This is not incompatible, because
1504 previously such lookups could never work.
1506 The situation is now rather anomolous, since one *can* have colons in
1507 ordinary lsearch keys. However, making the change in all cases is
1508 incompatible and would probably break a number of configurations.
1510 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1513 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1514 conversion specification without a maximum field width, thereby enabling
1515 a rogue spamd server to cause a buffer overflow. While nobody in their
1516 right mind would setup Exim to query an untrusted spamd server, an
1517 attacker that gains access to a server running spamd could potentially
1518 exploit this vulnerability to run arbitrary code as the Exim user.
1520 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1521 $primary_hostname instead of what libspf2 thinks the hosts name is.
1523 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1524 a directory entry by the name of the lookup key. Previously, if a
1525 symlink pointed to a non-existing file or a file in a directory that
1526 Exim lacked permissions to read, a lookup for a key matching that
1527 symlink would fail. Now it is enough that a matching directory entry
1528 exists, symlink or not. (Bugzilla 503.)
1530 PH/03 The body_linecount and body_zerocount variables are now exported in the
1533 PH/04 Added the $dnslist_matched variable.
1535 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1536 This means they are set thereafter only if the connection becomes
1539 PH/06 Added the client_condition to authenticators so that some can be skipped
1540 by clients under certain conditions.
1542 PH/07 The error message for a badly-placed control=no_multiline_responses left
1543 "_responses" off the end of the name.
1545 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1547 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1548 (without spaces) instead of just copying the configuration text.
1550 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1552 PH/11 Added $max_received_linelength.
1554 PH/12 Added +ignore_defer and +include_defer to host lists.
1556 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1557 way in which PCRE > 7.0 is built.
1559 PH/14 Implemented queue_only_load_latch.
1561 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1562 MAIL command. The effect was to mangle the value on 64-bit systems.
1564 PH/16 Another patch from the Sieve maintainer.
1566 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1568 PH/18 If a system quota error occurred while trying to create the file for
1569 a maildir delivery, the message "Mailbox is full" was not appended to the
1570 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1571 only to a quota excession during the actual writing of the file.
1573 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1574 characters?) which causes problems in log lines. The DN values are now
1575 passed through string_printing() before being added to log lines.
1577 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1578 and InterBase are left for another time.)
1580 PH/21 Added message_body_newlines option.
1582 PH/22 Guard against possible overflow in moan_check_errorcopy().
1584 PH/23 POSIX allows open() to be a macro; guard against that.
1586 PH/24 If the recipient of an error message contained an @ in the local part
1587 (suitably quoted, of course), incorrect values were put in $domain and
1588 $local_part during the evaluation of errors_copy.
1594 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1595 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1598 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1599 issue a MAIL command.
1601 PH/02 In an ACL statement such as
1603 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1605 if a client was not listed at all, or was listed with a value other than
1606 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1607 the condition was not true (as it should be), so access was not denied.
1608 The bug was that the ! inversion was incorrectly passed on to the second
1609 item. This has been fixed.
1611 PH/03 Added additional dnslists conditions == and =& which are different from
1612 = and & when the dns lookup returns more than one IP address.
1614 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1615 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1617 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1618 FSYNC, which compiles an option called disable_fsync that allows for
1619 bypassing fsync(). The documentation is heavily laced with warnings.
1621 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1623 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1624 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1625 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1626 including adding "make clean"; (3) Added -fPIC when compiling the test
1627 dynamically loaded module, to get rid of a warning.
1629 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1630 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1631 = 0s. The bug is otherwise harmless.
1633 PH/07 There was a bug in the dovecot authenticator such that the value of
1634 $auth1 could be overwritten, and so not correctly preserved, after a
1635 successful authentication. This usually meant that the value preserved by
1636 the server_setid option was incorrect.
1638 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1640 PH/09 Installed PCRE release 7.0.
1642 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1643 run for batched SMTP input. It is now run at the start of every message
1644 in the batch. While fixing this I discovered that the process information
1645 (output by running exiwhat) was not always getting set for -bs and -bS
1646 input. This is fixed, and it now also says "batched" for BSMTP.
1648 PH/11 Added control=no_pipelining.
1650 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1651 patch, slightly modified), and move the expansion of helo_data till after
1652 the connection is made in the smtp transport (so it can use these
1655 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1657 PH/14 Added log_selector = +pid.
1659 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1661 PH/16 Add ${if forany and ${if forall.
1663 PH/17 Added dsn_from option to vary the From: line in DSNs.
1665 PH/18 Flush SMTP output before performing a callout, unless control =
1666 no_callout_flush is set.
1668 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1669 was true (the default) a successful delivery failed to delete the retry
1670 item, thus causing premature timeout of the address. The bug is now
1673 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1675 PH/21 Long custom messages for fakedefer and fakereject are now split up
1676 into multiline reponses in the same way that messages for "deny" and
1677 other ACL rejections are.
1679 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1680 with slight modification.
1682 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1683 draft, changing the syntax and factoring some duplicate code".
1685 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1686 for deliveries of the second and subsequent messages over the same SMTP
1689 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1690 ${reduce, with only minor "tidies".
1692 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1694 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1695 expansion side effects.
1697 PH/27 When a message times out after an over-quota error from an Exim-imposed
1698 quota, the bounce message says "mailbox is full". This message was not
1699 being given when it was a system quota that was exceeded. It now should
1702 MH/03 Made $recipients available in local_scan(). local_scan() already has
1703 better access to the recipient list through recipients_list[], but
1704 $recipients can be useful in postmaster-provided expansion strings.
1706 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1707 in the case of a MAIL command with additional options following the
1708 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1709 were accidentally chopped off.
1711 PH/29 SMTP synchronization checks are implemented when a command is read -
1712 there is a check that no more input is waiting when there shouldn't be
1713 any. However, for some commands, a delay in an ACL can mean that it is
1714 some time before the response is written. In this time, more input might
1715 arrive, invalidly. So now there are extra checks after an ACL has run for
1716 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1717 pipelining has not been advertised.
1719 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1721 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1722 returned in the SMTP response when smtp_return_error_details was false.
1723 This has been fixed.
1725 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1726 instead of the C I/O that was originally supplied, because problems were
1727 reported on Solaris.
1729 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1730 Exim which did not show up earlier: it was assuming that a call to
1731 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1732 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1733 was a macro that became an assignment, so it seemed to work. This has
1734 changed to a proper function call with a void return, hence the compile
1735 error. Exim's code has been fixed.
1737 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1740 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1742 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1744 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1745 because it turns out to be too big or there is a timeout) there is no
1746 "Completed" line in the log. When some messages of this type were
1747 selected by exigrep, they were listed as "not completed". Others were
1748 picked up by some special patterns. I have improved the selection
1749 criteria to be more general.
1751 PH/38 The host_find_failed option in the manualroute router can now be set
1752 to "ignore", to completely ignore a host whose IP address cannot be
1753 found. If all hosts are ignored, the behaviour is controlled by the new
1754 host_all_ignored option.
1756 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1757 homing or because of multiple MX records with /mx) generated more than
1758 one IP address, and the following item turned out to be the local host,
1759 all the secondary addresses of the first item were incorrectly removed
1760 from the list, along with the local host and any following hosts (which
1761 is what is supposed to happen).
1763 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1764 whoever called Exim into the -H file. In the case of the daemon it was
1765 behaving confusingly. When first started, it used values for whoever
1766 started the daemon, but after a SIGHUP it used the Exim user (because it
1767 calls itself on a restart). I have changed the code so that it now always
1770 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1771 message are rejected with the same error (e.g. no authentication or bad
1772 sender address), and a DATA command is nevertheless sent (as can happen
1773 with PIPELINING or a stupid MUA), the error message that was given to the
1774 RCPT commands is included in the rejection of the DATA command. This is
1775 intended to be helpful for MUAs that show only the final error to their
1778 PH/42 Another patch from the Sieve maintainer.
1780 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1781 Eximstats - Fixed some broken HTML links and added missing column headers
1783 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1784 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1786 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1788 PH/43 Yet another patch from the Sieve maintainer.
1790 PH/44 I found a way to check for a TCP/IP connection going away before sending
1791 the response to the final '.' that terminates a message, but only in the
1792 case where the client has not sent further data following the '.'
1793 (unfortunately, this is allowed). However, in many cases there won't be
1794 any further data because there won't be any more messages to send. A call
1795 to select() can be used: if it shows that the input is "ready", there is
1796 either input waiting, or the socket has been closed. An attempt to read
1797 the next input character can distinguish the two cases. Previously, Exim
1798 would have sent an OK response which the client would never have see.
1799 This could lead to message repetition. This fix should cure that, at
1800 least in a lot of common cases.
1802 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1803 advertised in response to EHLO.
1809 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1810 fixed by 4.65/MH/01 (is this a record?) are fixed:
1812 (i) An empty string was always treated as zero by the numeric comparison
1813 operators. This behaviour has been restored.
1815 (ii) It is documented that the numeric comparison operators always treat
1816 their arguments as decimal numbers. This was broken in that numbers
1817 starting with 0 were being interpreted as octal.
1819 While fixing these problems I realized that there was another issue that
1820 hadn't been noticed. Values of message_size_limit (both the global option
1821 and the transport option) were treated as octal if they started with 0.
1822 The documentation was vague. These values are now always treated as
1823 decimal, and I will make that clear in the documentation.
1829 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1830 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1833 MH/01 Don't check that the operands of numeric comparison operators are
1834 integers when their expansion is in "skipping" mode (fixes bug
1835 introduced by 4.64-PH/07).
1837 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1838 child addresses, Exim now panics and dies. Previously, because the count
1839 is held in a short int, deliveries were likely to be lost. As such a
1840 large number of recipients for a single message is ridiculous
1841 (performance will be very, very poor), I have chosen to impose a limit
1842 rather than extend the field.
1848 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1849 leftover -K file (the existence of which was triggered by #402).
1850 While we were at it, introduced process PID as part of the -K
1851 filename. This should rule out race conditions when creating
1854 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1855 processing considerably. Previous code took too long for large mails,
1856 triggering a timeout which in turn triggers #401.
1858 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1859 in the DK code in transports.c. sendfile() is not really portable,
1860 hence the _LINUX specificness.
1862 TF/01 In the add_headers option to the mail command in an Exim filter,
1863 there was a bug that Exim would claim a syntax error in any
1864 header after the first one which had an odd number of characters
1867 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1868 callout verification, Exim cached a "reject" for the entire domain. This
1869 is correct for most verifications, but it is not correct for a recipient
1870 verification with use_sender or use_postmaster set, because in that case
1871 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1872 case of MAIL FROM:<> rejection from other early rejections (e.g.
1873 rejection of HELO). When verifying a recipient using a non-null MAIL
1874 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1875 Whatever the result of the callout, the value of the domain cache is
1876 left unchanged (for any other kind of callout, getting as far as trying
1877 RCPT means that the domain itself is ok).
1879 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1882 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1883 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1884 session, the daemon ignores ECONNECT errors and logs others; it now
1885 ignores EPIPE as well.
1887 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1888 (quoted-printable decoding).
1890 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1891 later the small subsequent patch to fix an introduced bug.
1893 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1895 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1897 PH/08 An error is now given if message_size_limit is specified negative.
1899 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1900 to be given (somewhat) arbitrary names.
1902 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1905 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1906 miscellaneous code fixes
1908 PH/10 Added the log_reject_target ACL modifier to specify where to log
1911 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1912 hostname. This is wrong, because it relates to the incoming message (and
1913 probably the interface on which it is arriving) and not to the outgoing
1914 callout (which could be using a different interface). This has been
1915 changed to use the value of the helo_data option from the smtp transport
1916 instead - this is what is used when a message is actually being sent. If
1917 there is no remote transport (possible with a router that sets up host
1918 addresses), $smtp_active_hostname is used.
1920 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1921 tweaks were necessary in order to get it to work (see also 21 below):
1922 (a) The code assumed that strncpy() returns a negative number on buffer
1923 overflow, which isn't the case. Replaced with Exim's string_format()
1925 (b) There were several signed/unsigned issues. I just did the minimum
1926 hacking in of casts. There is scope for a larger refactoring.
1927 (c) The code used strcasecmp() which is not a standard C function.
1928 Replaced with Exim's strcmpic() function.
1929 (d) The code set only $1; it now sets $auth1 as well.
1930 (e) A simple test gave the error "authentication client didn't specify
1931 service in request". It would seem that Dovecot has changed its
1932 interface. Fortunately there's a specification; I followed it and
1933 changed what the client sends and it appears to be working now.
1935 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1938 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1939 address A is aliased to B and C, where B exists and C does not. Without
1940 -v the output is "A verified" because verification stops after a
1941 successful redirection if more than one address is generated. However,
1942 with -v the child addresses are also verified. Exim was outputting "A
1943 failed to verify" and then showing the successful verification for C,
1944 with its parentage. It now outputs "B failed to verify", showing B's
1945 parentage before showing the successful verification of C.
1947 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1948 look up a TXT record in a specific list after matching in a combined
1951 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1952 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1953 they consult the DNS. I had assumed they would set it the way they
1954 wanted; and indeed my experiments on Linux seem to show that in some
1955 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1956 To be on the safe side, however, I have now made the interface to
1957 host_find_byname() similar to host_find_bydns(), with an argument
1958 containing the DNS resolver options. The host_find_byname() function now
1959 sets these options at its start, just as host_find_bydns() does. The smtp
1960 transport options dns_qualify_single and dns_search_parents are passed to
1961 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1962 of host_find_byname() use the default settings of RES_DEFNAMES
1963 (qualify_single) but not RES_DNSRCH (search_parents).
1965 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1966 spool_read_header() do less string testing, by means of a preliminary
1967 switch on the second character of optional "-foo" lines. (This is
1968 overdue, caused by the large number of possibilities that now exist.
1969 Originally there were few.) While I was there, I also converted the
1970 str(n)cmp tests so they don't re-test the leading "-" and the first
1971 character, in the hope this might squeeze out yet more improvement.
1973 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1974 flag allowing group syntax was set by the header_syntax check but not
1975 turned off, possible causing trouble later; (2) The flag was not being
1976 set at all for the header_verify test, causing "group"-style headers to
1977 be rejected. I have now set it in this case, and also caused header_
1978 verify to ignore an empty address taken from a group. While doing this, I
1979 came across some other cases where the code for allowing group syntax
1980 while scanning a header line wasn't quite right (mostly, not resetting
1981 the flag correctly in the right place). These bugs could have caused
1982 trouble for malformed header lines. I hope it is now all correct.
1984 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1985 with the "reply" argument non-NULL. The code, however (which originally
1986 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1987 but it didn't always do it. This confused somebody who was copying the
1988 code for some other use. I have removed all the tests.
1990 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1991 feature that was used to support insecure browsers during the U.S. crypto
1992 embargo. It requires special client support, and Exim is probably the
1993 only MTA that supported it -- and would never use it because real RSA is
1994 always available. This code has been removed, because it had the bad
1995 effect of slowing Exim down by computing (never used) parameters for the
1996 RSA_EXPORT functionality.
1998 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1999 authenticator to fail if there's a tab character in the incoming data
2000 (there should never be unless someone is messing about, as it's supposed
2001 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2002 if the connection is using TLS or if the remote IP is the same as the
2003 local IP, and the "valid-client-cert option" if a client certificate has
2006 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2007 authenticators. This can be used for authorization after authentication
2008 succeeds. (In the case of plaintext, it servers for both authentication
2011 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2012 if any retry times were supplied.
2014 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2015 connection, where there is no client IP address to check. In this
2016 situation, the verify now always succeeds.
2018 PH/25 Applied John Jetmore's -Mset patch.
2020 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2022 PH/27 In a string expansion for a processed (not raw) header when multiple
2023 headers of the same name were present, leading whitespace was being
2024 removed from all of them, but trailing whitespace was being removed only
2025 from the last one. Now trailing whitespace is removed from each header
2026 before concatenation. Completely empty headers in a concatenation (as
2027 before) are ignored.
2029 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2030 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2032 PH/29 [Removed. This was a change that I later backed out, and forgot to
2033 correct the ChangeLog entry (that I had efficiently created) before
2034 committing the later change.]
2036 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2037 address errors (4xx response to RCPT) over the same connection as other
2038 messages routed to the same hosts. Such deliveries are always "forced",
2039 so retry times are not inspected. This resulted in far too many retries
2040 for the affected addresses. The effect occurred only when there were more
2041 hosts than the hosts_max_try setting in the smtp transport when it had
2042 the 4xx errors. Those hosts that it had tried were not added to the list
2043 of hosts for which the message was waiting, so if all were tried, there
2044 was no problem. Two fixes have been applied:
2046 (i) If there are any address or message errors in an SMTP delivery, none
2047 of the hosts (tried or untried) are now added to the list of hosts
2048 for which the message is waiting, so the message should not be a
2049 candidate for sending over the same connection that was used for a
2050 successful delivery of some other message. This seems entirely
2051 reasonable: after all the message is NOT "waiting for some host".
2052 This is so "obvious" that I'm not sure why it wasn't done
2053 previously. Hope I haven't missed anything, but it can't do any
2054 harm, as the worst effect is to miss an optimization.
2056 (ii) If, despite (i), such a delivery is accidentally attempted, the
2057 routing retry time is respected, so at least it doesn't keep
2058 hammering the server.
2060 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2061 in ${readsocket because some servers need this prod.
2063 PH/32 Added some extra debug output when updating a wait-xxx database.
2065 PH/33 The hint "could be header name not terminated by colon", which has been
2066 given for certain expansion errors for a long time, was not being given
2067 for the ${if def:h_colon_omitted{... case.
2069 PH/34 The spec says: "With one important exception, whenever a domain list is
2070 being scanned, $domain contains the subject domain." There was at least
2071 one case where this was not true.
2073 PH/35 The error "getsockname() failed: connection reset by peer" was being
2074 written to the panic log as well as the main log, but it isn't really
2075 panic-worthy as it just means the connection died rather early on. I have
2076 removed the panic log writing for the ECONNRESET error when getsockname()
2079 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2080 runs only) independently of the message's sender address. This meant
2081 that, if the 4xx error was in fact related to the sender, a different
2082 message to the same recipient with a different sender could confuse
2083 things. In particualar, this can happen when sending to a greylisting
2084 server, but other circumstances could also provoke similar problems.
2085 I have changed the default so that the retry time for these errors is now
2086 based a combination of the sender and recipient addresses. This change
2087 can be overridden by setting address_retry_include_sender=false in the
2090 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2091 remote server are returned as part of bounce messages. This was not
2092 happening for LMTP over a pipe (the lmtp transport), but now it is the
2093 same for both kinds of LMTP.
2095 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2096 in header lines that were in fact CNAMEs. This is no longer the case.
2098 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2099 and queue runs started by the daemon processed all messages. This has
2100 been fixed so that -R and -S can now usefully be given with -q<time>.
2102 PH/40 Import PCRE release 6.7 (fixes some bugs).
2104 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2106 PH/42 Give an error if -q is specified more than once.
2108 PH/43 Renamed the variables $interface_address and $interface_port as
2109 $received_ip_address and $received_port, to make it clear that these
2110 values apply to message reception, and not to the outgoing interface when
2111 a message is delivered. (The old names remain recognized, of course.)
2113 PH/44 There was no timeout on the connect() call when using a Unix domain
2114 socket in the ${readsocket expansion. There now is.
2116 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2117 be meaningful with "accept".
2119 SC/01 Eximstats V1.43
2120 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2122 SC/02 Eximstats V1.44
2123 Use a glob alias rather than an array ref in the generated
2124 parser. This improves both readability and performance.
2126 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2127 Collect SpamAssassin and rejection statistics.
2128 Don't display local sender or destination tables unless
2129 there is data to show.
2130 Added average volumes into the top table text output.
2132 SC/04 Eximstats V1.46
2133 Collect data on the number of addresses (recipients)
2134 as well as the number of messages.
2136 SC/05 Eximstats V1.47
2137 Added 'Message too big' to the list of mail rejection
2138 reasons (thanks to Marco Gaiarin).
2140 SC/06 Eximstats V1.48
2141 Mainlog lines which have GMT offsets and are too short to
2142 have a flag are now skipped.
2144 SC/07 Eximstats V1.49 (Alain Williams)
2145 Added the -emptyok flag.
2147 SC/08 Eximstats V1.50
2148 Fixes for obtaining the IP address from reject messages.
2150 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2151 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2152 whitesspace changes from 4.64-PH/27
2154 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2157 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2158 are found, allow negative numbers in numeric criteria)
2160 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2162 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2163 to match changes made in 4.64-PH/43
2165 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2167 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2168 "domain" is actually an IP address, and force a failure. This locks out
2169 those revolvers/nameservers that support "A-for-A" lookups, in
2170 contravention of the specifications.
2172 PH/48 When a host name was looked up from an IP address, and the subsequent
2173 forward lookup of the name timed out, the host name was left in
2174 $sender_host_name, contrary to the specification.
2176 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2177 restricted to single-key lookups, Exim was not diagnosing an error if
2178 * or *@ was used with a query-style lookup.
2180 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2182 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2183 long ago, but noone interested enough thought of it. Let's just say that
2184 the "1.1" means that there are some new functions that weren't there at
2185 some point in the past.
2187 PH/51 Error processing for expansion failure of helo_data from an smtp
2188 transport during callout processing was broken.
2190 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2191 tested/used via the -bh/-bhc/-bs options.
2193 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2194 bug, fixed in subsequent PCRE releases).
2196 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2197 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2199 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2205 SC/01 Use a glob alias rather than an array ref in eximstats generated
2206 parser. This improves both readability and performance.
2208 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2209 Don't display local sender or destination tables in eximstats unless
2210 there is data to show.
2211 Added average volumes into the eximstats top table text output.
2213 SC/03 Collect data on the number of addresses (recipients) as well
2214 as the number of messages in eximstats.
2216 TF/01 Correct an error in the documentation for the redirect router. Exim
2217 does (usually) call initgroups() when daemonizing.
2219 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2220 with consistent privilege compared to when running as a daemon.
2222 TF/03 Note in the spec that $authenticated_id is not set for local
2223 submissions from trusted users.
2225 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2226 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2228 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2229 by adding some example configuration directives to the default
2230 configuration file. A little bit of work is required to uncomment the
2231 directives and define how usernames and passwords are checked, but
2232 there is now a framework to start from.
2234 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2235 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2236 without this. I don't know how relevant this is to other LDAP libraries.
2238 PH/02 Add the verb name to the "unknown ACL verb" error.
2240 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2242 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2244 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2245 directory not expanded when it should be if an expanded home directory
2246 was set for the address (which is overridden by the transport).
2248 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2251 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2252 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2253 because it is too late at that time, and has no effect.
2255 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2256 security issue with \' (bugzilla #107). I could not use the
2257 PQescapeStringConn() function, because it needs a PGconn value as one of
2260 PH/08 When testing addresses using -bt, indicate those final addresses that
2261 are duplicates that would not cause an additional delivery. At least one
2262 person was confused, thinking that -bt output corresponded to deliveries.
2263 (Suppressing duplicates isn't a good idea as you lose the information
2264 about possibly different redirections that led to the duplicates.)
2266 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2267 systems where poll() doesn't work, in particular OS X.
2269 PH/10 Added more information to debugging output for retry time not reached.
2271 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2272 operations in malware.c.
2274 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2277 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2278 syslog_duplication set false, log lines that would normally be written
2279 both the the main log and to the reject log were not written to syslog at
2282 PH/14 In the default configuration, change the use of "message" in ACL warn
2283 statements to "add_header".
2285 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2286 not followed by a command (e.g. "seen endif").
2288 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2289 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2292 PH/17 Added extra conditions to the default value of delay_warning_condition
2296 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2297 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2298 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2301 The Auto-Submitted: and various List- headers are standardised, whereas I
2302 don't think Precedence: ever was.
2304 PH/18 Refactored debugging code in route_finduser() to show more information,
2305 in particular, the error code if getpwnam() issues one.
2307 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2308 This is apparently needed in addition to the PH/07 change above to avoid
2309 any possible encoding problems.
2311 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2312 but not after initializing Perl.
2314 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2315 output them only if debugging. By default they are written stderr,
2316 apparently, which is not desirable.
2318 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2321 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2324 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2326 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2327 authenticated or an ident call has been made. Suppress the default
2328 values for $authenticated_id and $authenticated_sender (but permit -oMai
2329 and -oMas) when testing with -bh.
2331 PH/24 Re-jigged the order of the tests in the default configuration so that the
2332 tests for valid domains and recipients precede the DNS black list and CSA
2333 tests, on the grounds that those ones are more expensive.
2335 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2336 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2337 This bug exists in every version of Exim that I still have, right back to
2340 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2341 However, an attempt to turn on case-sensitivity in a regex key by
2342 including (?-i) didn't work because the subject string was already
2343 lowercased, and the effects were non-intuitive. It turns out that a
2344 one-line patch can be used to allow (?-i) to work as expected.
2350 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2351 other effects) broke the use of negated acl sub-conditions.
2353 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2356 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2357 "Deny" causes Exim to reject the incoming connection with a 554 error.
2358 Unfortunately, if there is a major crisis, such as a disk failure,
2359 tcp-wrappers gives "deny", whereas what one would like would be some
2360 kind of temporary error. A kludge has been added to help with this.
2361 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2362 554 error is used if errno is still zero or contains ENOENT (which occurs
2363 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2366 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2368 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2369 errors. Otherwise a message that provokes a temporary error (when other
2370 messages do not) can cause a whole host to time out.
2372 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2373 addresses were routed directly to files or pipes from a redirect router.
2374 File deliveries just didn't batch; pipe deliveries might have suffered
2377 PH/06 A failure to get a lock for a hints database would erroneously always say
2378 "Failed to get write lock", even when it was really a read lock.
2380 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2381 of 0600. This has been changed to use the value of the lockfile_mode
2382 option (which defaults to 0600).
2384 PH/08 Applied small patch from the Sieve maintainer.
2386 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2387 folder from quota calculations, a direct delivery into this folder messed
2388 up the contents of the maildirsize file. This was because the regex was
2389 used only to exclude .Trash (or whatever) when the size of the mailbox
2390 was calculated. There was no check that a delivery was happening into an
2391 excluded directory. This bug has been fixed by ignoring all quota
2392 processing for deliveries into excluded directories.
2394 PH/10 Added the maildirfolder_create_regex option to appendfile.
2400 PH/01 The code for finding all the local interface addresses on a FreeBSD
2401 system running IPv6 was broken. This may well have applied to all BSD
2402 systems, as well as to others that have similar system calls. The broken
2403 code found IPv4 interfaces correctly, but gave incorrect values for the
2404 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2405 that it would not match correctly against @[] and not recognize the IPv6
2408 PH/02 The ipliteral router was not recognizing addresses of the form user@
2409 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2411 PH/03 Added disable_ipv6.
2413 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2414 decoded form, because it is most often used to construct To: headers
2415 lines in autoreplies, and the decoded form may well be syntactically
2416 invalid. However, $reply_address has leading white space removed, and all
2417 newlines turned into spaces so that the autoreply transport does not
2420 PH/05 If group was specified without a user on a router, and no group or user
2421 was specified on a transport, the group from the router was ignored.
2423 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2424 for visible compile-time settings that can be used to change these
2425 numbers, for those that want even more. Backwards compatibility with old
2426 spool files has been maintained. However, going back to a previous Exim
2427 release will lost any variables that are in spool files.
2429 PH/07 Two small changes when running in the test harness: increase delay when
2430 passing a TCP/IP connection to a new process, in case the original
2431 process has to generate a bounce, and remove special handling of
2432 127.0.0.2 (sic), which is no longer necessary.
2434 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2435 be the same on different OS.
2437 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2440 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2441 whether --show-vars was specified or not
2443 JJ/02 exipick: Added support for new ACL variable spool format introduced
2446 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2447 syntactically invalid From: or Reply-to: line, and a filter used this to
2448 generate an autoreply, and therefore failed to obtain an address for the
2449 autoreply, Exim could try to deliver to a non-existent relative file
2450 name, causing unrelated and misleading errors. What now happens is that
2451 it logs this as a hard delivery error, but does not attempt to create a
2454 PH/11 The exinext utility has a -C option for testing purposes, but although
2455 the given file was scanned by exinext itself; it wasn't being passed on
2456 when Exim was called.
2458 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2459 an end-of-file indication when reading a command response.
2461 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2462 compiled. In many other places in Exim, IPv6 addresses are always
2463 recognized, so I have changed this. It also means that IPv4 domain
2464 literals of the form [IPV4:n.n.n.n] are now always recognized.
2466 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2467 used if the router is not running as root, for example, when verifying at
2468 ACL time, or when using -bh. The debugging output from this situation was
2469 non-existent - all you got was a failure to exec. I have made two
2472 (a) Failures to set uid/gid, the current directory, or a process leader
2473 in a subprocess such as that created by queryprogram now generate
2474 suitable debugging ouput when -d is set.
2476 (b) The queryprogram router detects when it is not running as root,
2477 outputs suitable debugging information if -d is set, and then runs
2478 the subprocess without attempting to change uid/gid.
2480 PH/15 Minor change to Makefile for building test_host (undocumented testing
2483 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2484 additional section of a DNS packet that returns MX or SRV records.
2485 Instead, it always explicitly searches for A/AAAA records. This avoids
2486 major problems that occur when a DNS server includes only records of one
2487 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2488 fixed another bug: if SRV records were looked up and the corresponding
2489 address records were *not* found in the additional section, the port
2490 values from the SRV records were lost.
2492 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2493 using the correct key (the original address) when searching the retry
2494 rules in order to find which one to use for generating the retry hint.
2496 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2497 adding the default one. Similarly, if it contains a Reply-To: header, the
2498 errors_reply_to option, if set, is not used.
2500 PH/19 When calculating a retry time, Exim used to measure the "time since
2501 failure" by looking at the "first failed" field in the retry record. Now
2502 it does not use this if it is later than than the arrival time of the
2503 message. Instead it uses the arrival time. This makes for better
2504 behaviour in cases where some deliveries succeed, thus re-setting the
2505 "first failed" field. An example is a quota failure for a huge message
2506 when small messages continue to be delivered. Without this change, the
2507 "time since failure" will always be short, possible causing more frequent
2508 delivery attempts for the huge message than are intended.
2509 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2511 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2512 $1, $2, $3) because the numerical variables can be reset during some
2513 expansion items (e.g. "match"), thereby losing the authentication data.
2515 PH/21 Make -bV show the size of off_t variables so that the test suite can
2516 decide whether to run tests for quotas > 2G.
2518 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2519 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2520 filecount value is greater than 2G or if a quota value is greater than 2G
2521 on a system where the size of off_t is not greater than 4, a panic error
2524 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2525 never match. The debug and -bh output now contains an explicit error
2526 message indicating a malformed IPv4 address or mask.
2528 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2529 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2530 PH/23 above applies.
2532 PH/25 Do not write to syslog when running in the test harness. The only
2533 occasion when this arises is a failure to open the main or panic logs
2534 (for which there is an explicit test).
2536 PH/26 Added the /no_tell option to "control=freeze".
2538 PH/27 If a host name lookup failed very early in a connection, for example, if
2539 the IP address matched host_lookup and the reverse lookup yielded a name
2540 that did not have a forward lookup, an error message of the form "no IP
2541 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2542 could be logged. Now it outputs the IP address instead of "NULL".
2544 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2545 allows the sender and the authenticated sender to be set when
2546 submitting a message from within Exim. Since child_open_exim() is
2547 documented for local_scan(), the new function should be too.
2549 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2550 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2551 results in an empty string is now treated as unset.
2553 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2555 PH/31 Added sender_verify_fail logging option.
2557 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2558 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2559 tidied the source and removed it altogether.
2561 PH/33 When a queue run was abandoned because the load average was too high, a
2562 log line was always written; now it is written only if the queue_run log
2563 selector is set. In addition, the log line for abandonment now contains
2564 information about the queue run such as the pid. This is always present
2565 in "start" and "stop" lines but was omitted from the "abandon" line.
2567 PH/34 Omit spaces between a header name and the colon in the error message that
2568 is given when verify = headers_syntax fails (if there are lots of them,
2569 the message gets confusing).
2571 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2572 names, as there are now some PTR records that contain slashes. This check
2573 is only to protect against broken name servers that fall over on strange
2574 characters, so the fact that it applies to all lookups doesn't matter.
2576 PH/36 Now that the new test suite is complete, we can remove some of the
2577 special code in Exim that was needed for the old test suite. For example,
2578 sorting DNS records because real resolvers return them in an arbitrary
2579 order. The new test suite's fake resolver always returns records in the
2582 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2583 bounces) except when queue_only is set, to avoid logging races between
2584 the different processes.
2586 PH/38 Panic-die if .include specifies a non-absolute path.
2588 PH/39 A tweak to the "H" retry rule from its user.
2590 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2591 a label. They prevented compilation on older perls.
2593 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2594 a warning to be raised on newish perls.
2596 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2597 on queue. Changes to match documented behaviour of showing count of
2598 messages matching specified criteria.
2600 PH/40 Changed the default ident timeout from 30s to 5s.
2602 PH/41 Added support for the use of login_cap features, on those BSD systems
2603 that have them, for controlling the resources used by pipe deliveries.
2605 PH/42 The content-scanning code uses fopen() to create files in which to put
2606 message data. Previously it was not paying any attention to the mode of
2607 the files. Exim runs with umask(0) because the rest of the code creates
2608 files with open(), and sets the required mode explicitly. Thus, these
2609 files were ending up world-writeable. This was not a big issue, because,
2610 being within the spool directory, they were not world-accessible. I have
2611 created a function called modefopen, which takes an additional mode
2612 argument. It sets umask(777), creates the file, chmods it to the required
2613 mode, then resets the umask. All the relevant calls to fopen() in the
2614 content scanning code have been changed to use this function.
2616 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2617 to 24 hours. This avoids potential overflow problems when processing G
2618 and H retry rules. I suspect nobody ever tinkers with this value.
2620 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2622 PH/45 When the plaintext authenticator is running as a client, the server's
2623 challenges are checked to ensure they are valid base64 strings. By
2624 default, the authentication attempt is cancelled if an invalid string is
2625 received. Setting client_ignore_invalid_base64 true ignores these errors.
2626 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2627 they are received. Thus, the responses can be made to depend on the
2628 challenges. If an invalid string is ignored, an empty string is placed in
2631 PH/46 Messages that are created by the autoreply transport now contains a
2632 References: header, in accordance with RFCs 2822 and 3834.
2634 PH/47 Added authenticated_sender_force to the smtp transport.
2636 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2638 PH/49 Installed latest patch from the Sieve maintainer.
2640 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2641 was also set, the appendfile transport was unnecessarily scanning a
2642 directory of message files (e.g. for maildir delivery) to find the count
2643 of files (along with the size), even though it did not need this
2644 information. It now does the scan only if it needs to find either the
2645 size of the count of files.
2647 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2649 PH/52 Two bugs concerned with error handling when the smtp transport is
2652 (i) Exim was not creating retry information for temporary errors given
2653 for individual recipients after the DATA command when the smtp transport
2654 was used in LMTP mode. This meant that they could be retried too
2655 frequently, and not timed out correctly.
2657 (ii) Exim was setting the flag that allows error details to be returned
2658 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2659 recipients that were returned after the DATA command.
2661 PH/53 This is related to PH/52, but is more general: for any failing address,
2662 when detailed error information was permitted to be returned to the
2663 sender, but the error was temporary, then after the final timeout, only
2664 "retry timeout exceeded" was returned. Now it returns the full error as
2665 well as "retry timeout exceeded".
2667 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2668 do this, and (what is worse) MTAs that accept it.
2670 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2671 will now be deprecated.
2673 PH/56 New os.c-cygwin from the Cygwin maintainer.
2675 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2676 formats (previously only available in exim formats via -bpr, -bpru,
2677 and -bpra. Now also available in native and exiqgrep formats)
2679 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2680 with very large, slow to parse queues
2682 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2684 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2686 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2687 responses to authentication challenges, though it was showing the
2688 challenges; (ii) I've removed the CR characters from the debug output for
2691 PH/58 Allow for the insertion of a newline as well as a space when a string
2692 is turned into more than one encoded-word during RFC 2047 encoding. The
2693 Sieve code now uses this.
2695 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2696 data_4xx, lost_connection, tls_required.
2698 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2699 message was being sent as an SMTP response.
2701 PH/61 Add -l and -k options to exicyclog.
2703 PH/62 When verifying, if an address was redirected to one new address, so that
2704 verification continued, and the new address failed or deferred after
2705 having set something in $address_data, the value of $address_data was not
2706 passed back to the ACL. This was different to the case when no
2707 redirection occurred. The value is now passed back in both cases.
2709 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2710 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2711 use login_cap.h, so on its own it isn't the distinguishing feature. The
2712 new name refers directly to the setclassresources() function.
2714 PH/65 Added configuration files for NetBSD3.
2716 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2718 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2719 is preferred over IPv4.
2721 PH/68 The bounce_return_message and bounce_return_body options were not being
2722 honoured for bounces generated during the reception of non-SMTP messages.
2723 In particular, this applied to messages rejected by the ACL. This bug has
2724 been fixed. However, if bounce_return_message is true and bounce_return_
2725 body is false, the headers that are returned for a non-SMTP message
2726 include only those that have been read before the error was detected.
2727 (In the case of an ACL rejection, they have all been read.)
2729 PH/69 The HTML version of the specification is now built in a directory called
2730 spec_html instead of spec.html, because the latter looks like a path with
2731 a MIME-type, and this confuses some software.
2733 PH/70 Catch two compiler warnings in sieve.c.
2735 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2736 function verify_get_ident() calls ip_connect() to connect a socket, but
2737 if the "connect()" function timed out, ip_connect() used to close the
2738 socket. However, verify_get_ident() also closes the socket later, and in
2739 between Exim writes to the log, which may get opened at this point. When
2740 the socket was closed in ip_connect(), the log could get the same file
2741 descriptor number as the socket. This naturally causes chaos. The fix is
2742 not to close the socket in ip_connect(); the socket should be closed by
2743 the function that creates it. There was only one place in the code where
2744 this was missing, in the iplookup router, which I don't think anybody now
2745 uses, but I've fixed it anyway.
2747 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2748 well as to direct DNS lookups. Otherwise the handling of names in host
2749 lists is inconsistent and therefore confusing.
2755 PH/01 Two changes to the default runtime configuration:
2757 (1) Move the checks for relay_from_hosts and authenticated clients from
2758 after to before the (commented out) DNS black list checks.
2760 (2) Add control=submission to the relay_from_hosts and authenticated
2761 clients checks, on the grounds that messages accepted by these
2762 statements are most likely to be submissions.
2764 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2766 (1) Generate an error if the third argument for the ${prvs expansion is
2769 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2772 (3) Reset the variables that are obtained from the first argument of
2773 ${prvscheck and used in the second argument before leaving the code,
2774 because their memory is reclaimed, so using them afterwards may do
2777 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2778 one (it's much easier than Tom thought :-).
2780 (5) Because of (4), we can now allow for the use of $prvscheck_result
2781 inside the third argument.
2783 PH/03 For some reason, the default setting of PATH when running a command from
2784 a pipe transport was just "/usr/bin". I have changed it to
2787 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2788 anything to be listed in the output from -bV.
2790 PH/05 When a filter generated an autoreply, the entire To: header line was
2791 quoted in the delivery log line, like this:
2793 => >A.N.Other <ano@some.domain> <original@ddress> ...
2795 This has been changed so that it extracts the operative address. There
2796 may be more than one such address. If so, they are comma-separated, like
2799 => >ano@some.domain,ona@other.domain <original@ddress> ...
2801 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2802 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2803 not being looked up in the rDNS to get a host name, Exim was showing the
2804 IP address twice in Received: lines, even though the IP addresses were
2805 identical. For example:
2807 Received: from [1.2.3.4] (helo=[1.2.3.4])
2809 However, if the real host name was known, it was omitting the HELO data
2810 if it matched the actual IP address. This has been tidied up so that it
2811 doesn't show the same IP address twice.
2813 PH/07 When both +timestamp and +memory debugging was on, the value given by
2814 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2815 called by the debug printing, thereby overwriting the timestamp buffer.
2816 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2818 PH/08 When the original message was included in an autoreply transport, it
2819 always said "this is a copy of the message, including all the headers",
2820 even if body_only or headers_only was set. It now gives an appropriate
2823 PH/09 Applied a patch from the Sieve maintainer which:
2825 o fixes some comments
2826 o adds the (disabled) notify extension core
2827 o adds some debug output for the result of if/elsif tests
2828 o points to the current vacation draft in the documentation
2829 and documents the missing references header update
2833 o fixes a bug in processing the envelope test (when testing
2834 multiple envelope elements, the last element determinted the
2837 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2838 Electronic Mail") by including:
2840 Auto-submitted: auto-generated
2842 in the messages that it generates (bounce messages and others, such as
2843 warnings). In the case of bounce messages for non-SMTP mesages, there was
2844 also a typo: it was using "Auto_submitted" (underscore instead of
2845 hyphen). Since every message generated by Exim is necessarily in response
2846 to another message, thes have all been changed to:
2848 Auto-Submitted: auto-replied
2850 in accordance with these statements in the RFC:
2852 The auto-replied keyword:
2854 - SHOULD be used on messages sent in direct response to another
2855 message by an automatic process,
2857 - MUST NOT be used on manually-generated messages,
2859 - MAY be used on Delivery Status Notifications (DSNs) and Message
2860 Disposition Notifications (MDNs),
2862 - MUST NOT be used on messages generated by automatic or periodic
2863 processes, except for messages which are automatic responses to
2866 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2867 to the default Received: header definition.
2869 PH/12 Added log selector acl_warn_skipped (default on).
2871 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2872 variables because (a) they are in the wrong storage pool and (b) even if
2873 they were copied, it wouldn't work properly because of the caching.
2875 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2876 checking when decoding. Apparently there are clients that generate
2877 overlong encoded strings. Why am I not surprised?
2879 PH/15 If the first argument of "${if match_address" was not empty, but did not
2880 contain an "@" character, Exim crashed. Now it writes a panic log message
2881 and treats the condition as false.
2883 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2885 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2886 extension "envelope-auth". The code is finished and in agreement with
2887 other implementations, but there is no documentation so far and in fact,
2888 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2889 not changing the active code.
2891 Print executed "if" and "elsif" statements when debugging is used. This
2892 helps a great deal to understand what a filter does.
2894 Document more things not specified clearly in RFC3028. I had all this
2895 sorted out, when out of a sudden new issues came to my mind. Oops."
2897 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2900 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2901 canonical form (as documented). However, after a host name lookup from
2902 the IP address, check_host() was doing a simple string comparison with
2903 addresses acquired from the DNS when checking that the found name did
2904 have the original IP as one of its addresses. Since any found IPv6
2905 addresses are likely to be in abbreviated form, the comparison could
2906 fail. Luckily, there already exists a function for doing the comparison
2907 by converting both addresses to binary, so now that is used instead of
2908 the text comparison.
2910 PH/20 There was another similar case to PH/19, when a complete host name was
2911 given in a host list; looking up its IP address could give an abbreviated
2912 form, whereas the current host's name might or might not be abbreviated.
2913 The same fix has been applied.
2919 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2920 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2923 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2924 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2926 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2928 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2929 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2930 Also, for "Auto-Submitted", treat anything other than "no" as
2931 non-personal, in accordance with RFC 3834. (Previously it treated
2932 anything starting "auto-" as non-personal.)
2934 TF/01 The control=submission/name=... option had a problem with syntax
2935 errors if the name included a slash character. The /name= option
2936 now slurps the rest of the string, so it can include any characters
2937 but it must come last in the list of options (after /sender_retain
2940 PH/05 Some modifications to the interface to the fake nameserver for the new
2948 TK/01 Added the "success_on_redirect" address verification option. See
2949 NewStuff for rationale and an example.
2951 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2953 PH/02 Patch to exigrep to allow it to work on syslog lines.
2955 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2956 fread() to skip over the body file's header line, because in Cygwin the
2957 header line is locked and is inaccessible.
2959 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2960 co-exist for some time) to make it clear that it is the Exim ID that is
2961 referenced, not the Message-ID: header line.
2963 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2964 string_format() function, because snprintf() does not exist on all
2967 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2968 ${stat: expansion item.
2970 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2971 protocol synchronization error", to keep the pedants happy.
2973 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2974 well as for IRIX systems, when gcc is being used. See the host.c source
2977 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2979 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2982 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2983 transport and to the smtp transport in LMTP mode.
2985 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2987 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2988 run a filter in a subprocess. This could lead to confusion in subsequent
2989 lookups in the parent process. There should also be a search_tidyup() at
2990 the end of the subprocess.
2992 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2993 only if the host matched helo_try_verify_hosts, which caused the
2994 verification to occur when the EHLO/HELO command was issued. The ACL just
2995 tested the remembered result. Now, if a previous verification attempt has
2996 not happened, "verify = helo" does it there and then.
2998 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3000 TK/03 Fix log output including CR from clamd.
3002 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3003 exist provoked a memory error which could cause a segfault.
3005 PH/15 Installed PCRE 6.2
3007 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3009 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3010 of the problem. Specifically, suggested +O2 rather than +O1 for the
3013 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3015 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3016 record, the port was not correctly logged when the outgoing_port log
3017 selector was set (it logged the transort's default port).
3019 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3020 fallback_hosts, and "hosts" in the smtp transport.
3022 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3023 host errors such as "Connection refused".
3025 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3026 authentication with radiusclient 0.4.9:
3028 - Error returned from rc_read_config was caught wrongly
3029 - Username/password not passed on to radius server due to wrong length.
3031 The presumption is that some radiusclient API changes for 4.51/PH/17
3032 were not taken care of correctly. The code is still untested by me (my
3033 Linux distribution still has 0.3.2 of radiusclient), but it was
3034 contributed by a Radius user.
3036 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3037 expanding the "port" option of the smtp transport.
3039 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3040 while reading a MIME header. Thanks to Tom Hughes for a patch.
3042 PH/24 Include config.h inside local_scan.h so that configuration settings are
3045 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3046 that in an ACL for RCPT (for example), you can examine exactly what was
3049 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3050 commands, but it was not correctly comparing the address with the actual
3051 client host address. Thus, it would show the EHLO address in Received:
3052 header lines when this was not necessary.
3054 PH/27 Added the % operator to ${eval:}.
3056 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3057 it should be ignoring failures (because with -C, for example, it has lost
3058 privilege). It wasn't ignoring creation failures other than "already
3061 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3064 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3065 because an input line was too long, either on its own, or by virtue of
3066 too many continuations, the temporary file was not being removed, and the
3067 return code was incorrect.
3069 PH/31 Missing "BOOL" in function definition in filtertest.c.
3071 PH/32 Applied Sieve patches from the maintainer.
3073 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3075 PH/33 Added "verify = not_blind".
3077 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3078 Local/Makefile (with some defaults set). These are used in built scripts
3079 such as exicyclog, but they have never been used in the exim_install
3080 script (though there are many overriding facilities there). I have
3081 arranged that the exim_install script now takes note of these two
3084 PH/35 Installed configuration files for Dragonfly.
3086 PH/36 When a locally submitted message by a trusted user did not contain a
3087 From: header, and the sender address was obtained from -f or from an SMTP
3088 MAIL command, and the trusted user did not use -F to supply a sender
3089 name, $originator_name was incorrectly used when constructing a From:
3090 header. Furthermore, $originator_name was used for submission mode
3091 messages from external hosts without From: headers in a similar way,
3092 which is clearly wrong.
3094 PH/37 Added control=suppress_local_fixups.
3096 PH/38 When log_selector = +received_sender was set, and the addition of the
3097 sender made the log line's construction buffer exactly full, or one byte
3098 less than full, an overflow happened when the terminating "\n" was
3101 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3102 when the result of a list match is failure because a DNS lookup failed.
3104 PH/40 RM_COMMAND is now used in the building process.
3106 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3107 the "build-* directories that it finds.
3109 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3110 address in a domain literal was a prefix of an interface address.
3112 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3113 when verifying a sender address, unless rewrite_headers is false.
3115 PH/44 Wrote a long comment about why errors_to addresses are verified as
3116 recipients, not senders.
3118 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3119 the ratelimit ACL was added.
3121 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3123 PH/46 Added extra information about PostgreSQL errors to the error string.
3125 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3126 avoiding the need to install special zones in a real server. This is
3127 backwards compatible; if it can't find the fake resolver, it drops back.
3128 Thus, both old and new test suites can be run.
3130 TF/02 Added util/ratelimit.pl
3132 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3133 clock is set back in time.
3135 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3136 Candler <B.Candler@pobox.com>.
3138 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3139 OK for addresses that are the result of redirections.
3141 PH/48 A number of further additions for the benefit of the new test suite,
3142 including a fake gethostbyname() that interfaces to the fake DNS resolver
3145 TF/06 The fix for widen_domains has also been applied to qualify_single and
3146 search_parents which are the other dnslookup options that can cause
3149 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3152 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3154 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3155 service" if it encountered an MX record with an empty target hostname.
3156 The message is now "an MX or SRV record indicated no SMTP service".
3158 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3159 if the DNS of the sending site is misconfigured. This is quite a
3160 common situation. This change restores the behaviour of treating a
3161 helo verification defer as a failure.
3163 PH/51 If self=fail was set on a router, the bounce message did not include the
3164 actual error message.
3170 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3172 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3173 command itself ended in error, the underlying message about the transport
3174 filter timeout was being overwritten with the pipe command error. Now the
3175 underlying error message should be appended to the second error message.
3177 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3179 PH/02 Reduce the amount of output that "make" produces by default. Full output
3180 can still be requested.
3182 PH/03 The warning log line about a condition test deferring for a "warn" verb
3183 was being output only once per connection, rather than after each
3184 occurrence (because it was using the same function as for successful
3185 "warn" verbs). This seems wrong, so I have changed it.
3187 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3188 it should not have, which might have caused a crash in the right
3189 circumstances, but probably never did.
3191 PH/04 Installed a modified version of Tony Finch's patch to make submission
3192 mode fix the return path as well as the Sender: header line, and to
3193 add a /name= option so that you can make the user's friendly name appear
3196 TF/03 Added the control = fakedefer ACL modifier.
3198 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3199 Mark Lowes for thorough testing.
3201 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3203 TK/03 Merged latest SRS patch from Miles Wilton.
3205 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3206 with the definition in sysexits.h (which is #included earlier).
3207 Fortunately, Exim does not actually use EX_OK. The code used to try to
3208 preserve the sysexits.h value, by assumimg that macro definitions were
3209 scanned for macro replacements. I have been disabused of this notion,
3210 so now the code just undefines EX_OK before #including unistd.h.
3212 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3213 in the smtp transport. When a block could not be written in a single
3214 write() function, the timeout was being re-applied to each part-write.
3215 This seems wrong - if the receiver was accepting one byte at a time it
3216 would take for ever. The timeout is now adjusted when this happens. It
3217 doesn't have to be particularly precise.
3219 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3220 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3222 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3223 without a domain if the check to <postmaster@domain> fails.
3225 SC/01 Eximstats: added -xls and the ability to specify output files
3226 (patch written by Frank Heydlauf).
3228 SC/02 Eximstats: use FileHandles for outputing results.
3230 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3232 SC/04 Eximstats: fixed display of large numbers with -nvr option
3234 SC/05 Eximstats: fixed merging of reports with empty tables.
3236 SC/06 Eximstats: added the -include_original_destination flag
3238 SC/07 Eximstats: removed tabs and trailing whitespace.
3240 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3242 TK/06 MBOX spool code: Add real "From " MBOX separator line
3243 so the .eml file is really in mbox format (even though
3244 most programs do not really care). Patch from Alex Miller.
3246 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3247 The latter is generated from $received_to and is only set if the
3248 message has one envelope recipient. SA can use these headers,
3249 obviously out-of-the-box. Patch from Alex Miller.
3251 PH/08 The ${def test on a variable was returning false if the variable's
3252 value was "0", contrary to what the specification has always said!
3253 The result should be true unless the variable is empty.
3255 PH/09 The syntax error of a character other than { following "${if
3256 def:variable_name" (after optional whitespace) was not being diagnosed.
3257 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3258 accidental colon was present, for example, could give incorrect results.
3260 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3261 result is used (not including appendfile, where other changes are about
3264 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3265 This involved changing a lot of size variables from int to off_t. It
3266 should work with maildirs and everything.
3268 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3269 spamd dying while we are connected to it.
3271 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3274 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3275 function for BDB 4.3.
3277 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3279 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3280 the sources for PCRE have been re-arranged and more files are now
3283 PH/15 The code I had for printing potentially long long variables in PH/11
3284 above was not the best (it lost precision). The length of off_t variables
3285 is now inspected at build time, and an appropriate printing format (%ld
3286 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3287 to be "long long int" or "long int". This is needed for the internal
3288 formatting function string_vformat().
3290 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3291 the configuration file to be ":syslog", then the script "guesses" where
3292 the logs files are, rather than using the compiled in default. In our
3293 case the guess is not the same as the compiled default, so the script
3294 suddenly stopped working when I started to use syslog. The patch checks
3295 to see if log_file_path is "". If so, it attempts to read it from exim
3296 with no configuration file to get the compiled in version, before it
3297 falls back to the previous guessing code."
3299 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3300 implementing BATV in an Exim configuration. See NewStuff for the gory
3303 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3304 Makefile that are specific to HP-UX.
3306 PH/18 If the "use_postmaster" option was set for a recipient callout together
3307 with the "random" option, the postmaster address was used as the MAIL
3308 FROM address for the random test, but not for the subsequent recipient
3309 test. It is now used for both.
3311 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3312 patch removes a few documentation additions to RFC 3028, because the
3313 latest draft now contains them. It adds the new en;ascii-case comparator
3314 and a new error check for 8bit text in MIME parts. Comparator and
3315 require names are now matched exactly. I enabled the subaddress
3316 extension, but it is not well tested yet (read: it works for me)."
3318 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3319 rework some of the code of TK/09 above to avoid the hardwired use of
3320 "%lld" and "long long". Replaced the call to snprintf() with a call to
3323 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3324 records point to non-existent hosts", "retry timeout exceeded", and
3325 "retry time not reached for any host after a long failure period".
3327 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3328 experimental DomainKeys support:
3330 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3331 (2) On an error such as an illegally used "control", the wrong name for
3332 the control was given.
3334 These problems did NOT occur unless DomainKeys support was compiled.
3336 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3338 PH/24 Added ${if match_ip condition.
3340 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3341 databases so that it will be absolutely obvious if a crash occurs in the
3342 DB library. This is a regular occurrence (often caused by mis-matched
3345 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3346 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3347 picked up on a user's system that detects such things. There doesn't seem
3348 to be a gcc warning option for this - only an attribute that has to be
3349 put on the function's prototype. It seems that in Fedora Core 4 they have
3350 set this on a number of new functions. No doubt there will be more in due
3353 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3354 specify a transport. However, if an address that was verified by such a
3355 router was the subject of a callout, Exim crashed because it tried to
3356 read the rcpt_include_affixes from the non-existent transport. Now it
3357 just assumes that the setting of that option is false. This bug was
3358 introduced by 4.51/PH/31.
3360 PH/28 Changed -d+all to exclude +memory, because that information is very
3361 rarely of interest, but it makes the output a lot bigger. People tend to
3362 do -d+all out of habit.
3364 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3365 code in os-type was giving problems when libc.so lives in lib64, like on
3368 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3369 aren't the modern standard, and it seems that some systems' include files
3370 don't always have them. Exim was already checking for some of the newer
3371 ones like T_AAAA, and defining it itself. I've added checks for all the
3372 record types that Exim uses.
3374 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3375 not automatically generating a new one, as it is supposed to. This
3376 prevented TLS from working. If the file did exist, but contained invalid
3377 data, a new version was generated, as expected. It was only the case of a
3378 non-existent file that was broken.
3380 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3381 with a change in libdomainkeys > 0.64.
3383 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3384 from DNS. If the selector record carries the flag, it now has
3385 precedence over the domain-wide flag.
3387 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3389 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3390 the use of a port name that isn't defined in /etc/services) occurred, the
3391 message was deferred as in a normal delivery, and thus remained on the
3392 spool, instead of being failed because of the mua_wrapper setting. This
3393 is now fixed, and I tidied up some of the mua_wrapper messages at the
3396 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3397 the messages in a hash of arrays rather than using individual hashes.
3398 This is a bit cleaner and results in dramatic memory savings, albeit
3399 at a slight CPU cost.
3401 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3402 as requested by Marc Sherman.
3404 SC/10 Eximstats: added histograms for user specified patterns as requested
3407 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3409 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3410 fopen() in the content-scanning modules that did not already have it.
3416 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3417 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3419 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3421 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3423 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3424 to test Sieve filters that use "vacation".
3426 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3427 that changes the way the GnuTLS parameters are stored in the cache file.
3428 The new format can be generated externally. For backward compatibility,
3429 if the data in the cache doesn't make sense, Exim assumes it has read an
3430 old-format file, and it generates new data and writes a new file. This
3431 means that you can't go back to an older release without removing the
3434 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3435 work if there are any delivery delays because "one_time" forces the
3436 parent to be marked "delivered", so its unseen clone is never tried
3437 again. For this reason, Exim now forbids the simultaneous setting of
3440 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3441 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3442 entry if you want to know the details. The fix, however, neglected to
3443 consider the case when local delivery batching is involved. The test for
3444 "previously delivered" was not happening when checking to see if an
3445 address could be batched with a previous (undelivered) one; under
3446 certain circumstances this could lead to multiple deliveries to the same
3449 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3450 in its include files, and this causes problems building Exim.
3452 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3453 header_syntax) but Exim was just ignoring anything given after a slash.
3454 In particular, this caused confusion with an attempt to use "verify =
3455 reverse_host_lookup/defer_ok". An error is now given when options are
3456 supplied for verify items that do not have them. (Maybe reverse_host_
3457 lookup should have a defer_ok option, but that's a different point.)
3459 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3460 defined by RFC 821) to 2048, because there were problems with some AUTH
3461 commands, and RFC 1869 says the size should be increased for extended
3462 SMTP commands that take arguments.
3464 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3467 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3468 "unknown" error; now it says that the functionality isn't in the binary.
3470 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3471 an address' error message when a string expansion fails (syntax or
3472 whatever). Otherwise the password may appear in the log. Following change
3473 PH/42 below, there is no longer a chance of it appearing in a bounce
3476 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3478 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3479 of its addresses was ever tried. (Bugzilla bug #2.)
3481 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3482 the result incorrectly in the debug output. (It correctly added a newline
3483 to what was transported.)
3485 TF/01 Added $received_time.
3487 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3488 commented out examples of how to interface to a virus scanner and to
3489 SpamAssassin. Also added commented examples of av_scanner and
3490 spamd_address settings.
3492 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3493 and controls are allowed in which ACLs. There were a couple of minor
3494 errors. Some of the entries in the conditions table (which is a table of
3495 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3496 negation of where the condition IS allowed.
3498 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3500 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3501 header file does not have a version number, so I've had to invent a new
3502 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3503 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3504 radiusclient), but it was contributed by a Radius user.
3506 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3507 files or directories, for OpenSSL.
3509 PH/19 When an Exim process that is running as root has to create an Exim log
3510 file, it does so in a subprocess that runs as exim:exim so as to get the
3511 ownership right at creation (otherwise, other Exim processes might see
3512 the file with the wrong ownership). There was no test for failure of this
3513 fork() call, which would lead to the process getting stuck as it waited
3514 for a non-existent subprocess. Forks do occasionally fail when resources
3515 run out. I reviewed all the other calls to fork(); they all seem to check
3518 PH/20 When checking for unexpected SMTP input at connect time (before writing
3519 the banner), Exim was not dealing correctly with a non-positive return
3520 from the read() function. If the client had disconnected by this time,
3521 the result was a log entry for a synchronization error with an empty
3522 string after "input=" when read() returned zero. If read() returned -1
3523 (an event I could not check), uninitialized data bytes were printed.
3524 There were reports of junk text (parts of files, etc) appearing after
3527 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3529 PH/22 Added support for macro redefinition, and (re)definition in between
3530 driver and ACL definitions.
3532 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3533 forgetting to use the resulting value; it was using the unexpanded value.
3535 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3536 hadn't been configured. The fix is from Juergen Kreileder, who
3537 understands it better than I do:
3539 "Here's what I see happening with three configured cyrus_sasl
3540 authenticators configured (plain, login, cram-md5):
3542 On startup auth_cyrus_sasl_init() gets called for each of these.
3543 This means three calls to sasl_listmech() without a specified mech_list.
3544 => SASL tests which mechs of all available mechs actually work
3545 => three warnings about OTP not working
3546 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3548 With the patch, sasl_listmech() also gets called three times. But now
3549 SASL's mech_list option is set to the server_mech specified in the the
3550 authenticator. Or in other words, the answer from sasl_listmech()
3551 gets limited to just the mech you're testing for (which is different
3553 => the return list contains just 'plain' or 'login', 'cram-md5' or
3554 nothing depending on the value of ob->server_mech.
3556 I've just tested the patch: Authentication still works fine,
3557 unavailable mechs specified in the exim configuration are still
3558 caught, and the auth.log warnings about OTP are gone."
3560 PH/25 When debugging is enabled, the contents of the command line are added
3561 to the debugging output, even when log_selector=+arguments is not
3564 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3565 answer is "GNU", and only if the return is "GNU/something" is the answer
3568 PH/27 $acl_verify_message is now set immediately after the failure of a
3569 verification in an ACL, and so is available in subsequent modifiers. In
3570 particular, the message can be preserved by coding like this:
3572 warn !verify = sender
3573 set acl_m0 = $acl_verify_message
3575 Previously, $acl_verify_message was set only while expanding "message"
3576 and "log_message" when a very denied access.
3578 PH/28 Modified OS/os.c-Linux with
3580 -#ifndef OS_LOAD_AVERAGE
3581 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3583 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3584 nomenclature these days.)
3586 PH/29 Installed patch from the Sieve maintainer that adds the options
3587 sieve_useraddress and sieve_subaddress to the redirect router.
3589 PH/30 In these circumstances:
3590 . Two addresses routed to the same list of hosts;
3591 . First host does not offer TLS;
3592 . First host accepts first address;
3593 . First host gives temporary error to second address;
3594 . Second host offers TLS and a TLS session is established;
3595 . Second host accepts second address.
3596 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3597 and peerdn, if requested) that were in fact used only for the second
3600 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3601 attention to any local part prefix or suffix that was matched by the
3602 router that accepted the address. It now behaves in the same way as it
3603 does for delivery: the affixes are removed from the local part unless
3604 rcpt_include_affixes is set on the transport.
3606 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3607 timeout during the DATA phase of an incoming message.
3609 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3610 applied a patch sanctioned by the Sieve maintainer.
3612 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3613 the uid or gid is negative. A case of a negative gid caused this to be
3614 noticed. The fix allows for either to be negative.
3616 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3617 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3620 PH/36 The change PH/12 above was broken. Fixed it.
3622 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3623 the grounds that routing the same address twice would always produce the
3624 same answer. This might have been true once, but it is certainly no
3625 longer true now. Routing a child address may depend on the previous
3626 routing that produced that child. Some complicated redirection strategies
3627 went wrong when messages had multiple recipients, and made Exim's
3628 behaviour dependent on the order in which the addresses were given.
3630 I have moved the duplicate checking until after the routing is complete.
3631 Exim scans the addresses that are assigned to local and remote
3632 transports, and removes any duplicates. This means that more work will be
3633 done, as duplicates will always all be routed, but duplicates are
3634 presumably rare, so I don't expect this is of any significance.
3636 For deliveries to pipes, files, and autoreplies, the duplicate checking
3637 still happens during the routing process, since they are not going to be
3640 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3641 It corrects a timeout issue with spamd. This is Ian's comment: "The
3642 background is that sometimes spamd either never reads data from a
3643 connection it has accepted, or it never writes response data. The exiscan
3644 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3645 blindly assumes that writes won't block so it may never time out."
3647 PH/39 Allow G after quota size as well as K and M.
3649 PH/40 The value set for $authenticated_id in an authenticator may not contain
3650 binary zeroes or newlines because the value is written to log lines and
3651 to spool files. There was no check on this. Now the value is run through
3652 the string_printing() function so that such characters are converted to
3653 printable escape sequences.
3655 PH/41 $message_linecount is a new variable that contains the total number of
3656 lines in the message. Compare $body_linecount, which is the count for the
3659 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3660 bounce and delay warning messages, except in certain special cases, which
3663 (a) An SMTP error message from a remote host;
3664 (b) A message specified in a :fail: redirection;
3665 (c) A message specified in a "fail" command in a system filter;
3666 (d) A message specified in a FAIL return from the queryprogram router;
3667 (e) A message specified by the cannot_route_message router option.
3669 In these cases only, Exim does include the error details in bounce and
3670 warning messages. There are also a few cases where bland messages such
3671 as "unrouteable address" or "local delivery error" are given.
3673 PH/43 $value is now also set for the "else" part of a ${run expansion.
3675 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3676 being worked on, but at least Exim now implements the latest version to
3679 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3680 process to complete was treated as a delivery failure, a timeout while
3681 writing the message to the pipe was logged, but erroneously treated as a
3682 successful delivery. Such timeouts include transport filter timeouts. For
3683 consistency with the overall process timeout, these timeouts are now
3684 treated as errors, giving rise to delivery failures by default. However,
3685 there is now a new Boolean option for the pipe transport called
3686 timeout_defer, which, if set TRUE, converts the failures into defers for
3687 both kinds of timeout. A transport filter timeout is now identified in
3690 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3691 systems where "make" and "gmake" are different, calling "gmake" at top
3692 level broke things. I've arranged for the value of $(MAKE) to be passed
3693 from the Makefile to this script so that it can call the same version of
3697 A note about Exim versions 4.44 and 4.50
3698 ----------------------------------------
3700 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3701 changes of various kinds. As a consequence, a big documentation update was
3702 needed. This delayed the release for rather longer than seemed good, especially
3703 in the light of a couple of (minor) security issues. Therefore, the changes
3704 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3705 release. So 4.44 and 4.50 are in effect two different branches that both start
3708 I have left the 4.50 change log unchanged; it contains all the changes since
3709 4.43. The change log for 4.44 is below; many of its items are identical to
3710 those for 4.50. This seems to be the most sensible way to preserve the
3711 historical information.
3717 1. Minor wording change to the doc/README.SIEVE file.
3719 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3720 computation of the current number of files was incorrect.
3722 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3723 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3726 4. Give more explanation in the error message when the command for a transport
3727 filter fails to execute.
3729 5. There are several places where Exim runs a non-Exim command in a
3730 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3731 was being done only for the command run by the queryprogram router. It is
3732 now done for all such subprocesses. The other cases are: ${run, transport
3733 filters, and the commands run by the lmtp and pipe transports.
3735 6. Added CONFIGURE_GROUP build-time option.
3737 7. Some older OS have a limit of 256 on the maximum number of file
3738 descriptors. Exim was using setrlimit() to set 1000 as a large value
3739 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3740 systems. I've change it so that if it can't get 1000, it tries for 256.
3742 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3743 was an oversight, and furthermore, ever since the addition of extra
3744 controls (e.g. 4.43/32), the checks on when to allow different forms of
3745 "control" were broken. There should now be diagnostics for all cases when a
3746 control that does not make sense is encountered.
3748 9. Added the /retain_sender option to "control=submission".
3750 10. $recipients is now available in the predata ACL (oversight).
3752 11. Tidy the search cache before the fork to do a delivery from a message
3753 received from the command line. Otherwise the child will trigger a lookup
3754 failure and thereby defer the delivery if it tries to use (for example) a
3755 cached ldap connection that the parent has called unbind on.
3757 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3758 of $address_data from the recipient verification was clobbered by the
3759 sender verification.
3761 13. The value of address_data from a sender verification is now available in
3762 $sender_address_data in subsequent conditions in the ACL statement.
3764 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3766 15. Added a new option "connect=<time>" to callout options, to set a different
3769 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3770 was its contents. (It was OK if the option was not defined at all.)
3772 17. A "Completed" log line is now written for messages that are removed from
3773 the spool by the -Mrm option.
3775 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3776 information about exactly what failed.
3778 19. Added -dd to debug only the daemon process.
3780 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3781 handles timeouts, both on the server side and network timeouts. Renamed the
3782 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3784 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3785 It is now set to "smtps".
3787 22. $host_address is now set to the target address during the checking of
3788 ignore_target_hosts.
3790 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3791 being passed; this would have caused $sender_host_name to have been used if
3792 matching the list had actually called for a host name (not very likely,
3793 since this list is usually IP addresses). A host name is now passed as
3796 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3797 code that specifies a non-restarting handler (typically sigaction() in
3798 modern systems) in an attempt to fix a rare and obscure crash bug.
3800 25. Narrowed the window for a race in the daemon that could cause it to ignore
3801 SIGCHLD signals. This is not a major problem, because they are used only to
3802 wake it up if nothing else does.
3804 26. A malformed maildirsize file could cause Exim to calculate negative values
3805 for the mailbox size or file count. Odd effects could occur as a result.
3806 The maildirsize information is now recalculated if the size or filecount
3809 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3810 support for a long time. Removed HAVE_SYS_VFS_H.
3812 28. Installed the latest version of exipick from John Jetmore.
3814 29. In an address list, if the pattern was not a regular expression, an empty
3815 subject address (from a bounce message) matched only if the pattern was an
3816 empty string. Non-empty patterns were not even tested. This was the wrong
3817 because it is perfectly reasonable to use an empty address as part of a
3818 database query. An empty address is now tested by patterns that are
3819 lookups. However, all the other forms of pattern expect the subject to
3820 contain a local part and a domain, and therefore, for them, an empty
3821 address still always fails if the pattern is not itself empty.
3823 30. Exim went into a mad DNS loop when attempting to do a callout where the
3824 host was specified on an smtp transport, and looking it up yielded more
3825 than one IP address.
3827 31. Re-factored the code for checking spool and log partition space into a
3828 function that finds that data and another that does the check. The former
3829 is then used to implement four new variables: $spool_space, $log_space,
3830 $spool_inodes, and $log_inodes.
3832 32. The RFC2047 encoding function was originally intended for short strings
3833 such as real names; it was not keeping to the 75-character limit for
3834 encoded words that the RFC imposes. It now respects the limit, and
3835 generates multiple encoded words if necessary. To be on the safe side, I
3836 have increased the buffer size for the ${rfc2047: expansion operator from
3839 33. It is now permitted to omit both strings after an "if" condition; if the
3840 condition is true, the result is "true". As before, when the second string
3841 is omitted, a false condition yields an empty string. This makes it less
3842 cumbersome to write custom ACL and router conditions.
3844 34. Failure to deliver a bounce message always caused it to be frozen, even if
3845 there was an errors_to setting on the router. The errors_to setting is now
3848 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3849 canonical form (fully expanded) before being placed in
3850 $sender_host_address.
3852 36. The table in the code that translates DNS record types into text (T_A to
3853 "A" for instance) was missing entries for NS and CNAME. It is just possible
3854 that this could have caused confusion if both these types were looked up
3855 for the same domain, because the text type is used as part of Exim's
3856 per-process caching. But the chance of anyone hitting this buglet seems
3859 37. The dnsdb lookup has been extended in a number of ways.
3861 (1) There is a new type, "zns", which walks up the domain tree until it
3862 finds some nameserver records. It should be used with care.
3864 (2) There is a new type, "mxh", which is like "mx" except that it returns
3865 just the host names, not the priorities.
3867 (3) It is now possible to give a list of domains (or IP addresses) to be
3868 looked up. The behaviour when one of the lookups defers can be
3869 controlled by a keyword.
3871 (4) It is now possible to specify the separator character for use when
3872 multiple records are returned.
3874 38. The dnslists ACL condition has been extended: it is now possible to supply
3875 a list of IP addresses and/or domains to be looked up in a particular DNS
3878 39. Added log_selector=+queue_time_overall.
3880 40. When running the queue in the test harness, wait just a tad after forking a
3881 delivery process, to get repeatability of debugging output.
3883 41. Include certificate and key file names in error message when GnuTLS fails
3884 to set them up, because the GnuTLS error message doesn't include the name
3885 of the failing file when there is a problem reading it.
3887 42. Allow both -bf and -bF in the same test run.
3889 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3891 44. The "Exiscan patch" is now merged into the mainline Exim source.
3893 45. Sometimes the final signoff response after QUIT could fail to get
3894 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3895 < 0 before doing a fflush(). This bug looks as though it goes back to the
3896 introduction of TLS in release 3.20, but "sometimes" must have been rare
3897 because the tests only now provoked it.
3899 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3900 (this can affect the format of dates).
3902 47. exim_tidydb, when checking for the continued existence of a message for
3903 which it has found a message-specific retry record, was not finding
3904 messages that were in split spool directories. Consequently, it was
3905 deleting retry records that should have stayed in existence.
3907 48. Steve fixed some bugs in eximstats.
3909 49. The SPA authentication driver was not abandoning authentication and moving
3910 on to the next authenticator when an expansion was forced to fail,
3911 contradicting the general specification for all authenticators. Instead it
3912 was generating a temporary error. It now behaves as specified.
3914 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3915 (the order specifies the preference for clients). The order is now AES256,
3916 AES128, 3DES, ARCFOUR128.
3918 51. Small patch to Sieve code - explicitly set From: when generating an
3921 52. Exim crashed if a remote delivery caused a very long error message to be
3922 recorded - for instance if somebody sent an entire SpamAssassin report back
3923 as a large number of 550 error lines. This bug was coincidentally fixed by
3924 increasing the size of one of Exim's internal buffers (big_buffer) that
3925 happened as part of the Exiscan merge. However, to be on the safe side, I
3926 have made the code more robust (and fixed the comments that describe what
3929 53. Now that there can be additional text after "Completed" in log lines (if
3930 the queue_time_overall log selector is set), a one-byte patch to exigrep
3931 was needed to allow it to recognize "Completed" as not the last thing in
3934 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3935 patch that reportedly fixes this has been added. I am not expert enough to
3936 create a test for it. This is what the patch creator wrote:
3938 "I found a little strange behaviour of ldap code when working with
3939 Windows 2003 AD Domain, where users was placed in more than one
3940 Organization Units. When I tried to give exim partial DN, the exit code
3941 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3942 But simultaneously result of request was absolutely normal ldap result,
3943 so I produce this patch..."
3945 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3946 REFERENCE, so I have modified the code to exclude the patch when that macro
3949 55. Some experimental protocols are using DNS PTR records for new purposes. The
3950 keys for these records are domain names, not reversed IP addresses. The
3951 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3952 leaves it alone. Component reversal etc. now happens only for IP addresses.
3955 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3957 57. Double the size of the debug message buffer (to 2048) so that more of very
3958 long debug lines gets shown.
3960 58. The exicyclog utility now does better if the number of log files to keep
3961 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3963 59. Two changes related to the smtp_active_hostname option:
3965 (1) $smtp_active_hostname is now available as a variable.
3966 (2) The default for smtp_banner uses $smtp_active_hostname instead
3967 of $primary_hostname.
3969 60. The host_aton() function is supposed to be passed a string that is known
3970 to be a valid IP address. However, in the case of IPv6 addresses, it was
3971 not checking this. This is a hostage to fortune. Exim now panics and dies
3972 if the condition is not met. A case was found where this could be provoked
3973 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3974 components; fortuitously, this particular loophole had already been fixed
3975 by change 4.50/55 above.
3977 If there are any other similar loopholes, the new check in host_aton()
3978 itself should stop them being exploited. The report I received stated that
3979 data on the command line could provoke the exploit when Exim was running as
3980 exim, but did not say which command line option was involved. All I could
3981 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3982 running as the user.
3985 61. There was a buffer overflow vulnerability in the SPA authentication code
3986 (which came originally from the Samba project). I have added a test to the
3987 spa_base64_to_bits() function which I hope fixes it.
3990 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3991 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3993 63. The daemon start-up calls getloadavg() while still root for those OS that
3994 need the first call to be done as root, but it missed one case: when
3995 deliver_queue_load_max is set with deliver_drop_privilege. This is
3996 necessary for the benefit of the queue runner, because there is no re-exec
3997 when deliver_drop_privilege is set.
3999 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4000 This has been fixed.
4002 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4003 was in use, was not putting the data itself into the right store pool;
4004 consequently, it could be overwritten for a subsequent message in the same
4005 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4008 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4010 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4011 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4012 the function were treating the return as a boolean value, which happened to
4013 work because 0=false and not-0=true, but is not correct code.
4015 68. The host_aton() function was not handling scoped IPv6 addresses (those
4016 with, for example, "%eth0" on the end) correctly.
4018 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4019 negated items (that is, ~something) in unsigned ints. Some compilers
4020 apparently mutter when there is no cast.
4022 70. If an address verification called from an ACL failed, and did not produce a
4023 user-specific message (i.e. there was only a "system" message), nothing was
4024 put in $acl_verify_message. In this situation, it now puts the system
4027 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4028 session; change 4.31/43 added the unwanted input to the log line - except
4029 that it did not do this in the start of session case. It now does.
4031 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4032 This is wrong and can cause the other end to generate a synchronization
4033 error if it is another Exim or anything else that does the synchronization
4034 check. A QUIT command is no longer sent after a timeout.
4036 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4037 during host lookups.
4039 74. The defer_ok option of callout verification was not working if it was used
4040 when verifying addresses in header lines, that is, for this case:
4042 verify = header_sender/callout=defer_ok
4044 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4045 those file descriptors could be used for SMTP connections. If anything
4046 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4047 could be sent to the SMTP client, causing chaos. The daemon now opens
4048 stdin, stdout, and stderr to /dev/null when it puts itself into the
4051 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4052 log by default. The user can override this with suitable Perl magic.
4054 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4055 the log message when discard triggers, was not working for the DATA ACL or
4056 for the non-SMTP ACL.
4058 78. Error message wording change in sieve.c.
4060 79. If smtp_accept_max_per_host was set, the number of connections could be
4061 restricted to fewer than expected, because the daemon was trying to set up
4062 a new connection before checking whether the processes handling previous
4063 connections had finished. The check for completed processes is now done
4064 earlier. On busy systems, this bug wouldn't be noticed because something
4065 else would have woken the daemon, and it would have reaped the completed
4068 80. If a message was submitted locally by a user whose login name contained one
4069 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4070 It caused a spool format error. I have fixed the spool reading code. A
4071 related problem was that the "from" clause in the Received: line became
4072 illegal because of the space(s). It is now covered by ${quote_local_part.
4074 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4077 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4079 83. Patch from Sieve maintainer to fix unterminated string problem in
4080 "vacation" handling.
4082 84. Some minor changes to the Linux configuration files to help with other
4083 OS variants using glibc.
4085 85. One more patch for Sieve to update vacation handling to latest spec.
4088 ----------------------------------------------------
4089 See the note above about the 4.44 and 4.50 releases.
4090 ----------------------------------------------------
4096 1. Change 4.43/35 introduced a bug that caused file counts to be
4097 incorrectly computed when quota_filecount was set in an appendfile
4100 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4101 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4104 3. Give more explanation in the error message when the command for a transport
4105 filter fails to execute.
4107 4. There are several places where Exim runs a non-Exim command in a
4108 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4109 was being done only for the command run by the queryprogram router. It is
4110 now done for all such subprocesses. The other cases are: ${run, transport
4111 filters, and the commands run by the lmtp and pipe transports.
4113 5. Some older OS have a limit of 256 on the maximum number of file
4114 descriptors. Exim was using setrlimit() to set 1000 as a large value
4115 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4116 systems. I've change it so that if it can't get 1000, it tries for 256.
4118 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4119 was an oversight, and furthermore, ever since the addition of extra
4120 controls (e.g. 4.43/32), the checks on when to allow different forms of
4121 "control" were broken. There should now be diagnostics for all cases when a
4122 control that does not make sense is encountered.
4124 7. $recipients is now available in the predata ACL (oversight).
4126 8. Tidy the search cache before the fork to do a delivery from a message
4127 received from the command line. Otherwise the child will trigger a lookup
4128 failure and thereby defer the delivery if it tries to use (for example) a
4129 cached ldap connection that the parent has called unbind on.
4131 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4132 of $address_data from the recipient verification was clobbered by the
4133 sender verification.
4135 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4136 was its contents. (It was OK if the option was not defined at all.)
4138 11. A "Completed" log line is now written for messages that are removed from
4139 the spool by the -Mrm option.
4141 12. $host_address is now set to the target address during the checking of
4142 ignore_target_hosts.
4144 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4145 being passed; this would have caused $sender_host_name to have been used if
4146 matching the list had actually called for a host name (not very likely,
4147 since this list is usually IP addresses). A host name is now passed as
4150 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4151 code that specifies a non-restarting handler (typically sigaction() in
4152 modern systems) in an attempt to fix a rare and obscure crash bug.
4154 15. Narrowed the window for a race in the daemon that could cause it to ignore
4155 SIGCHLD signals. This is not a major problem, because they are used only to
4156 wake it up if nothing else does.
4158 16. A malformed maildirsize file could cause Exim to calculate negative values
4159 for the mailbox size or file count. Odd effects could occur as a result.
4160 The maildirsize information is now recalculated if the size or filecount
4163 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4164 support for a long time. Removed HAVE_SYS_VFS_H.
4166 18. Updated exipick to current release from John Jetmore.
4168 19. Allow an empty sender to be matched against a lookup in an address list.
4169 Previously the only cases considered were a regular expression, or an
4172 20. Exim went into a mad DNS lookup loop when doing a callout where the
4173 host was specified on the transport, if the DNS lookup yielded more than
4176 21. The RFC2047 encoding function was originally intended for short strings
4177 such as real names; it was not keeping to the 75-character limit for
4178 encoded words that the RFC imposes. It now respects the limit, and
4179 generates multiple encoded words if necessary. To be on the safe side, I
4180 have increased the buffer size for the ${rfc2047: expansion operator from
4183 22. Failure to deliver a bounce message always caused it to be frozen, even if
4184 there was an errors_to setting on the router. The errors_to setting is now
4187 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4188 canonical form (fully expanded) before being placed in
4189 $sender_host_address.
4191 24. Updated eximstats to version 1.33
4193 25. Include certificate and key file names in error message when GnuTLS fails
4194 to set them up, because the GnuTLS error message doesn't include the name
4195 of the failing file when there is a problem reading it.
4197 26. Expand error message when OpenSSL has problems setting up cert/key files.
4200 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4201 (this can affect the format of dates).
4203 28. exim_tidydb, when checking for the continued existence of a message for
4204 which it has found a message-specific retry record, was not finding
4205 messages that were in split spool directories. Consequently, it was
4206 deleting retry records that should have stayed in existence.
4208 29. eximstats updated to version 1.35
4209 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4210 1.35 - bugfix such that pie charts by volume are generated correctly
4212 30. The SPA authentication driver was not abandoning authentication and moving
4213 on to the next authenticator when an expansion was forced to fail,
4214 contradicting the general specification for all authenticators. Instead it
4215 was generating a temporary error. It now behaves as specified.
4217 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4218 (the order specifies the preference for clients). The order is now AES256,
4219 AES128, 3DES, ARCFOUR128.
4221 31. Small patch to Sieve code - explicitly set From: when generating an
4224 32. Exim crashed if a remote delivery caused a very long error message to be
4225 recorded - for instance if somebody sent an entire SpamAssassin report back
4226 as a large number of 550 error lines. This bug was coincidentally fixed by
4227 increasing the size of one of Exim's internal buffers (big_buffer) that
4228 happened as part of the Exiscan merge. However, to be on the safe side, I
4229 have made the code more robust (and fixed the comments that describe what
4232 33. Some experimental protocols are using DNS PTR records for new purposes. The
4233 keys for these records are domain names, not reversed IP addresses. The
4234 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4235 leaves it alone. Component reversal etc. now happens only for IP addresses.
4238 34. The host_aton() function is supposed to be passed a string that is known
4239 to be a valid IP address. However, in the case of IPv6 addresses, it was
4240 not checking this. This is a hostage to fortune. Exim now panics and dies
4241 if the condition is not met. A case was found where this could be provoked
4242 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4243 components; fortuitously, this particular loophole had already been fixed
4244 by change 4.50/55 or 4.44/33 above.
4246 If there are any other similar loopholes, the new check in host_aton()
4247 itself should stop them being exploited. The report I received stated that
4248 data on the command line could provoke the exploit when Exim was running as
4249 exim, but did not say which command line option was involved. All I could
4250 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4251 running as the user.
4254 35. There was a buffer overflow vulnerability in the SPA authentication code
4255 (which came originally from the Samba project). I have added a test to the
4256 spa_base64_to_bits() function which I hope fixes it.
4259 36. The daemon start-up calls getloadavg() while still root for those OS that
4260 need the first call to be done as root, but it missed one case: when
4261 deliver_queue_load_max is set with deliver_drop_privilege. This is
4262 necessary for the benefit of the queue runner, because there is no re-exec
4263 when deliver_drop_privilege is set.
4265 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4266 was in use, was not putting the data itself into the right store pool;
4267 consequently, it could be overwritten for a subsequent message in the same
4268 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4271 38. Sometimes the final signoff response after QUIT could fail to get
4272 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4273 < 0 before doing a fflush(). This bug looks as though it goes back to the
4274 introduction of TLS in release 3.20, but "sometimes" must have been rare
4275 because the tests only now provoked it.
4281 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4282 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4283 Now it returns an integer. A number of calls were still expecting a T/F
4284 return. Fortuitously, in all cases, the tests worked in OK situations,
4285 which is the norm. However, things would have gone wrong on any write
4286 failures on the smtp file descriptor. This function is used when sending
4287 messages over SMTP and also when doing verify callouts.
4289 2. When Exim is called to do synchronous delivery of a locally submitted
4290 message (the -odf or -odi options), it no longer closes stderr before doing
4293 3. Implemented the mua_wrapper option.
4295 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4297 5. Implemented the functions header_remove(), header_testname(),
4298 header_add_at_position(), and receive_remove_recipient(), and exported them
4301 6. If an ACL "warn" statement specified the addition of headers, Exim already
4302 inserted X-ACL-Warn: at the start if there was no header name. However, it
4303 was not making this test for the second and subsequent header lines if
4304 there were newlines in the string. This meant that an invalid header could
4305 be inserted if Exim was badly configured.
4307 7. Allow an ACL "warn" statement to add header lines at the start or after all
4308 the Received: headers, as well as at the end.
4310 8. Added the rcpt_4xx retry error code.
4312 9. Added postmaster_mailfrom=xxx to callout verification option.
4314 10. Added mailfrom=xxxx to the callout verification option, for verify=
4317 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4318 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4320 12. Inserted some casts to stop certain compilers warning when using pointer
4321 differences as field lengths or precisions in printf-type calls (mostly
4322 affecting debugging statements).
4324 13. Added optional readline() support for -be (dynamically loaded).
4326 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4327 same clock tick as a message's arrival, so that its received time was the
4328 same as the "first fail" time on the retry record, and that message
4329 remained on the queue past the ultimate address timeout, every queue runner
4330 would try a delivery (because it was past the ultimate address timeout) but
4331 after another failure, the ultimate address timeout, which should have then
4332 bounced the address, did not kick in. This was a "< instead of <=" error;
4333 in most cases the first failure would have been in the next clock tick
4334 after the received time, and all would be well.
4336 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4337 being recognized when the domain list was tested by the match_domain
4338 condition in an expansion string.
4340 16. Added the ${str2b64: operator.
4342 17. Exim was always calling setrlimit() to set a large limit for the number of
4343 processes, without checking whether the existing limit was already
4344 adequate. (It did check for the limit on file descriptors.) Furthermore,
4345 errors from getrlimit() and setrlimit() were being ignored. Now they are
4346 logged to the main and panic logs, but Exim does carry on, to try to do its
4347 job under whatever limits there are.
4349 18. Imported PCRE 5.0.
4351 19. Trivial typo in log message " temporarily refused connection" (the leading
4354 20. If the log selector return_path_on_delivery was set and an address was
4355 redirected to /dev/null, the delivery process crashed because it assumed
4356 that a return path would always be set for a "successful" delivery. In this
4357 case, the whole delivery is bypassed as an optimization, and therefore no
4360 21. Internal re-arrangement: the function for sending a challenge and reading
4361 a response while authentication was assuming a zero-terminated challenge
4362 string. It's now changed to take a pointer and a length, to allow for
4363 binary data in such strings.
4365 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4367 23. Exim was not respecting finduser_retries when seeking the login of the
4368 uid under which it was called; it was always trying 10 times. (The default
4369 setting of finduser_retries is zero.) Also, it was sleeping after the final
4370 failure, which is pointless.
4372 24. Implemented tls_on_connect_ports.
4374 25. Implemented acl_smtp_predata.
4376 26. If the domain in control=submission is set empty, Exim assumes that the
4377 authenticated id is a complete email address when it generates From: or
4378 Sender: header lines.
4380 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4381 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4382 chown and chgrp in /bin and hostname in /usr/bin.
4384 28. Exim was keeping the "process log" file open after each use, just as it
4385 does for the main log. This opens the possibility of it remaining open for
4386 long periods when the USR1 signal hits a daemon. Occasional processlog
4387 errors were reported, that could have been caused by this. Anyway, it seems
4388 much more sensible not to leave this file open at all, so that is what now
4391 29. The long-running daemon process does not normally write to the log once it
4392 has entered its main loop, and it closes the log before doing so. This is
4393 so that log files can straightforwardly be renamed and moved. However,
4394 there are a couple of unusual error situations where the daemon does write
4395 log entries, and I had neglected to close the log afterwards.
4397 30. The text of an SMTP error response that was received during a remote
4398 delivery was being truncated at 512 bytes. This is too short for some of
4399 the long messages that one sometimes sees. I've increased the limit to
4402 31. It is now possible to make retry rules that apply only when a message has a
4403 specific sender, in particular, an empty sender.
4405 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4406 it possible to be selective about when SMTP synchronization is enforced.
4408 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4410 32. Implemented hosts_connection_nolog.
4412 33. Added an ACL for QUIT.
4414 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4417 35. Added mailbox_size and mailbox_filecount to appendfile.
4419 36. Added control = no_multiline_responses to ACLs.
4421 37. There was a bug in the logic of the code that waits for the clock to tick
4422 in the case where the clock went backwards by a substantial amount such
4423 that the microsecond fraction of "now" was more than the microsecond
4424 fraction of "then" (but the whole seconds number was less).
4426 38. Added support for the libradius Radius client library this is found on
4427 FreeBSD (previously only the radiusclient library was supported).
4433 1. When certain lookups returned multiple values in the form name=value, the
4434 quoting of the values was not always being done properly. Specifically:
4435 (a) If the value started with a double quote, but contained no whitespace,
4437 (b) If the value contained whitespace other than a space character (i.e.
4438 tabs or newlines or carriage returns) it was not quoted.
4439 This fix has been applied to the mysql and pgsql lookups by writing a
4440 separate quoting function and calling it from the lookup code. The fix
4441 should probably also be applied to nisplus, ibase and oracle lookups, but
4442 since I cannot test any of those, I have not disturbed their existing code.
4444 2. A hit in the callout cache for a specific address caused a log line with no
4445 reason for rejecting RCPT. Now it says "Previous (cached) callout
4446 verification failure".
4448 3. There was an off-by-one bug in the queryprogram router. An over-long
4449 return line was truncated at 256 instead of 255 characters, thereby
4450 overflowing its buffer with the terminating zero. As well as fixing this, I
4451 have increased the buffer size to 1024 (and made a note to document this).
4453 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4454 when Exim is waiting for an SMTP response from a remote server, Exim
4455 restarts its select() call on the socket, thereby resetting its timeout.
4456 This is not a problem when such interrupts are rare. Somebody set up a cron
4457 job to run exiwhat every 2 minutes, which is less than the normal select()
4458 timeout (5 or 10 minutes). This meant that the select() timeout never
4459 kicked in because it was always reset. I have fixed this by comparing the
4460 time when an interrupt arrives with the time at the start of the first call
4461 to select(). If more time than the timeout has elapsed, the interrupt is
4462 treated as a timeout.
4464 5. Some internal re-factoring in preparation for the addition of Sieve
4465 extensions (by MH). In particular, the "personal" test is moved to a
4466 separate function, and given an option for scanning Cc: and Bcc: (which is
4467 not set for Exim filters).
4469 6. When Exim created an email address using the login of the caller as the
4470 local part (e.g. when creating a From: or Sender: header line), it was not
4471 quoting the local part when it contained special characters such as @.
4473 7. Installed new OpenBSD configuration files.
4475 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4476 try to make them clearer.
4478 9. Callout options, other than the timeout value, were being ignored when
4479 verifying sender addresses in header lines. For example, when using
4481 verify = header_sender/callout=no_cache
4483 the cache was (incorrectly) being used.
4485 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4486 only the Cygwin environment.
4488 11. When return_path_on_delivery was set as a log selector, if different remote
4489 addresses in the same message used different return paths and parallel
4490 remote delivery occurred, the wrong values would sometimes be logged.
4491 (Whenever a remote delivery process finished, the return path value from
4492 the most recently started remote delivery process was logged.)
4494 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4495 lines when AUTH and/or TLS are in use. This is the "received protocol"
4496 field. Exim used to use "asmtp" for authenticated SMTP, without any
4497 indication (in the protocol name) for TLS use. Now it follows the RFC and
4498 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4499 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4500 names appear in log lines as well as in Received: header lines.
4502 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4503 extensions, and comparison tests, and to fix some bugs.
4505 14. Changes to the "personal" filter test:
4507 (1) The test was buggy in that it was just doing the equivalent of
4508 "contains" tests on header lines. For example, if a user's address was
4509 anne@some.where, the "personal" test would incorrectly be true for
4511 To: susanne@some.where
4513 This test is now done by extracting each address from the header in turn,
4514 and checking the entire address. Other tests that are part of "personal"
4515 are now done using regular expressions (for example, to check local parts
4516 of addresses in From: header lines).
4518 (2) The list of non-personal local parts in From: addresses has been
4519 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4520 taken from the Sieve specification recommendations.
4522 (3) If the message contains any header line starting with "List-" it is
4523 treated as non-personal.
4525 (4) The test for "circular" in the Subject: header line has been removed
4526 because it now seems ill-conceived.
4528 15. Minor typos in src/EDITME comments corrected.
4530 16. Installed latest exipick from John Jetmore.
4532 17. If headers_add on a router specified a text string that was too long for
4533 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4534 of string_sprintf() is now avoided.
4536 18. $message_body_size was not set (it was always zero) when running the DATA
4537 ACL and the local_scan() function.
4539 19. For the "mail" command in an Exim filter, no default was being set for
4540 the once_repeat time, causing a random time value to be used if "once" was
4541 specified. (If the value happened to be <= 0, no repeat happened.) The
4542 default is now 0s, meaning "never repeat". The "vacation" command was OK
4543 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4544 (I found it when inspecting the code).
4546 20. There is now an overall timeout for performing a callout verification. It
4547 defaults to 4 times the callout timeout, which applies to individual SMTP
4548 commands during the callout. The overall timeout applies when there is more
4549 than one host that can be tried. The timeout is checked before trying the
4550 next host. This prevents very long delays if there are a large number of
4551 hosts and all are timing out (e.g. when the network connections are timing
4552 out). The value of the overall timeout can be changed by specifying an
4553 additional sub-option for "callout", called "maxwait". For example:
4555 verify = sender/callout=5s,maxwait=20s
4557 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4558 to the end before writing, but this should make it even safer).
4560 22. Exim was forgetting that it had advertised PIPELINING for the second and
4561 subsequent messages on an SMTP connection. It was also not resetting its
4562 memory on STARTTLS and an internal HELO.
4564 23. When Exim logs an SMTP synchronization error within a session, it now
4565 records whether PIPELINING has been advertised or not.
4567 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4568 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4569 rather than long int.
4571 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4573 26. Added the never_mail option to autoreply.
4579 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4580 crash if the getsockname() call failed; this can happen if a connection is
4581 closed very soon after it is established. The problem was simply in the
4582 order in which certain operations were done, causing Exim to try to write
4583 to the SMTP stream before it had set up the file descriptor. The bug has
4584 been fixed by making things happen in the correct order.
4590 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4591 before the connection was closed, thus losing the rejection response.
4593 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4594 some early Solaris releases, but causes trouble in current releases where
4595 socklen_t is defined.
4597 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4600 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4603 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4604 that releasing the top of it at the end releases what was used for sub-
4605 expansions (unless the block got too big). However, discard this block if
4606 the first thing is a variable or header, so that we can use its block when
4607 it is dynamic (useful for very large $message_headers, for example).
4609 6. Lookups now cache *every* query, not just the most recent. A new, separate
4610 store pool is used for this. It can be recovered when all lookup caches are
4611 flushed. Lookups now release memory at the end of their result strings.
4612 This has involved some general refactoring of the lookup sources.
4614 7. Some code has been added to the store_xxx() functions to reduce the amount
4615 of flapping under certain conditions.
4617 8. log_incoming_interface used to affect only the <= reception log lines. Now
4618 it causes the local interface and port to be added to several more SMTP log
4619 lines, for example "SMTP connection from", and rejection lines.
4621 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4623 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4625 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4626 could be overwritten at the end of the current message (or the start of a
4627 new message if it was set in a HELO ACL). The value is now preserved for
4628 the duration of the SMTP connection.
4630 12. If a transport had a headers_rewrite setting, and a matching header line
4631 contained an unqualified address, that address was qualified, even if it
4632 did not match any rewriting rules. The underlying bug was that the values
4633 of the flags that permit the existence of unqualified sender and recipient
4634 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4635 non-local messages, and by -bnq for local messages) were not being
4636 preserved with the message after it was received.
4638 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4639 "next input=" as part of the text comprising the host identity instead of
4640 the correct text. The code was using the same buffer for two different
4641 strings. However, depending on which order the printing function evaluated
4642 its arguments, the bug did not always show up. Under Linux, for example, my
4643 test suite worked just fine.
4645 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4646 doesn't work with some older versions of Perl. It has been changed to "my",
4647 which in any case is probably the better facility to use.
4649 15. A really picky compiler found some instances of statements for creating
4650 error messages that either had too many or two few arguments for the format
4653 16. The size of the buffer for calls to the DNS resolver has been increased
4654 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4655 for addresses that have a lot of PTR records. This alleviates a problem; it
4656 does not fully solve it.
4658 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4659 buffer now truncates the list and logs the incident, which is the same
4660 action as happens when Exim is looking up a host name and its aliases.
4661 Previously in this situation something unpredictable would happen;
4662 sometimes it was "internal error: store_reset failed".
4664 18. If a server dropped the connection unexpectedly when an Exim client was
4665 using GnuTLS and trying to read a response, the client delivery process
4666 crashed while trying to generate an error log message.
4668 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4669 string, for example:
4671 warn message = H1: something\nH2: something
4673 the text was added as a single header line from Exim's point of view
4674 though it ended up OK in the delivered message. However, searching for the
4675 second and subsequent header lines using $h_h2: did not work. This has been
4676 fixed. Similarly, if a system filter added multiple headers in this way,
4677 the routers could not see them.
4679 20. Expanded the error message when iplsearch is called with an invalid key to
4680 suggest using net-iplsearch in a host list.
4682 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4683 ACLs are no longer actually imposed (and a message to that effect is
4686 22. If a "gecos" field in a passwd entry contained escaped characters, in
4687 particular, if it contained a \" sequence, Exim got it wrong when building
4688 a From: or a Sender: header from that name. A second bug also caused
4689 incorrect handling when an unquoted " was present following a character
4690 that needed quoting.
4692 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4693 was not being matched caselessly.
4695 24. Arranged for all hyphens in the exim.8 source to be escaped with
4698 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4699 properly. Recipient callout cache records were still being keyed to include
4700 the sender, even when use_sender was set false. This led to far more
4701 callouts that were necessary. The sender is no longer included in the key
4702 when use_sender is false.
4704 26. Added "control = submission" modifier to ACLs.
4706 27. Added the ${base62d: operator to decode base 62 numbers.
4708 28. dnsdb lookups can now access SRV records.
4710 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4711 the configuration file.
4713 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4714 (-v) mode. This makes the output for a verbose queue run more intelligible.
4716 31. Added a use_postmaster feature to recipient callouts.
4718 32. Added the $body_zerocount variable, containing the number of binary zero
4719 bytes in the message body.
4721 33. The time of last modification of the "new" subdirectory is now used as the
4722 "mailbox time last read" when there is a quota error for a maildir
4725 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4727 35. Added +ignore_unknown as a special item in host lists.
4729 36. Code for decoding IPv6 addresses in host lists is now included, even if
4730 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4731 address was recognized as an IP address, but was then not correctly decoded
4732 into binary, causing unexpected and incorrect effects when compared with
4739 1. Very minor rewording of debugging text in manualroute to say "list of
4740 hosts" instead of "hostlist".
4742 2. If verify=header_syntax was set, and a header line with an unqualified
4743 address (no domain) and a large number of spaces between the end of the
4744 name and the colon was received, the reception process suffered a buffer
4745 overflow, and (when I tested it) crashed. This was caused by some obsolete
4746 code that should have been removed. The fix is to remove it!
4748 3. When running in the test harness, delay a bit after writing a bounce
4749 message to get a bit more predictability in the log output.
4751 4. Added a call to search_tidyup() just before forking a reception process. In
4752 theory, someone could use a lookup in the expansion of smtp_accept_max_
4753 per_host which, without the tidyup, could leave open a database connection.
4755 5. Added the variables $recipient_data and $sender_data which get set from a
4756 lookup success in an ACL "recipients" or "senders" condition, or a router
4757 "senders" option, similar to $domain_data and $local_part_data.
4759 6. Moved the writing of debug_print from before to after the "senders" test
4762 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4763 problems for message scanning, either using a data ACL, or using
4764 local_scan() because the Received: header was not generated till after they
4765 were called (in order to set the time as the time of reception completion).
4766 I have revised the way this works. The header is now generated after the
4767 body is received, but before the ACL or local_scan() are called. After they
4768 are run, the timestamp in the header is updated.
4774 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4775 before starting a queue runner without re-exec. This happened only when
4776 deliver_drop_privilege was set or when the Exim user was set to root. The
4777 effect of the bug was that timeouts during subsequent deliveries caused
4778 crashes instead of being properly handled. The handler is now left at its
4779 default (and expected) setting.
4781 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4782 message, again when deliver_drop_privilege is set or Exim is run as root.
4783 The bug described in (1) was not present in this case, but the tidying up
4784 of the other signals was missing. I have made the two cases consistent.
4786 3. The ignore_target_hosts setting on a manualroute router was being ignored
4787 for hosts that were looked up using the /MX notation.
4789 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4792 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4793 operated on the sender address. After changing the $sender_address to <>
4794 for the sender address verify, Exim was re-instated it as the original
4795 (before rewriting) address, but remembering that it had rewritten it, so it
4796 wasn't rewriting it again. This bug also had the effect of breaking the
4797 sender address verification caching when the sender address was rewritten.
4799 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4800 This has been changed so that if the ip literal address matches
4801 ignore_target_hosts, the router declines.
4803 7. Added expansion conditions match_domain, match_address, and match_local_
4804 part (NOT match_host).
4806 8. The placeholder for the Received: header didn't have a length field set.
4808 9. Added code to Exim itself and to exim_lock to test for a specific race
4809 condition that could lead to file corruption when using MBX delivery. The
4810 issue is with the lockfile that is created in /tmp. If this file is removed
4811 after a process has opened it but before that process has acquired a lock,
4812 there is the potential for a second process to recreate the file and also
4813 acquire a lock. This could lead to two Exim processes writing to the file
4814 at the same time. The added code performs the same test as UW imapd; it
4815 checks after acquiring the lock that its file descriptor still refers to
4816 the same named file.
4818 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4819 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4822 11. Added the smtp_active_hostname option. If used, this will typically be made
4823 to depend on the incoming interface address. Because $interface_address is
4824 not set up until the daemon has forked a reception process, error responses
4825 that can happen earlier (such as "too many connections") no longer contain
4828 12. If an expansion in a condition on a "warn" statement fails because a lookup
4829 defers, the "warn" statement is abandoned, and the next ACL statement is
4830 processed. Previously this caused the whole ACL to be aborted.
4832 13. Added the iplsearch lookup type.
4834 14. Added ident_timeout as a log selector.
4836 15. Added tls_certificate_verified as a log selector.
4838 16. Added a global option tls_require_ciphers (compare the smtp transport
4839 option of the same name). This controls incoming TLS connections.
4841 17. I finally figured out how to make tls_require_ciphers do a similar thing
4842 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4843 before starting the TLS session.
4845 18. Tabs are now shown as \t in -bP output.
4847 19. If the log selector return_path_on_delivery was set, Exim crashed when
4848 bouncing a message because it had too many Received: header lines.
4850 20. If two routers both had headers_remove settings, and the first one included
4851 a superfluous trailing colon, the final name in the first list and the
4852 first name in the second list were incorrectly joined into one item (with a
4853 colon in the middle).
4859 1. Added -C and -D options to the exinext utility, mainly to make it easier
4860 to include in the automated testing, but these could be helpful when
4861 multiple configurations are in use.
4863 2. The exinext utility was not formatting the output nicely when there was
4864 an alternate port involved in the retry record key, nor when there was a
4865 message id as well (for retries that were specific to a specific message
4866 and a specific host). It was also confused by IPv6 addresses, because of
4867 the additional colons they contain. I have fixed the IPv4 problem, and
4868 patched it up to do a reasonable job for IPv6.
4870 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4871 delivery, the log line now contains "pipelined" if PIPELINING was used.
4873 4. An SMTP transport process used to panic and die if the bind() call to set
4874 an explicit outgoing interface failed. This has been changed; it is now
4875 treated in the same way as a connect() failure.
4877 5. A reference to $sender_host_name in the part of a conditional expansion
4878 that was being skipped was still causing a DNS lookup. This no longer
4881 6. The def: expansion condition was not recognizing references to header lines
4882 that used bh_ and bheader_.
4884 7. Added the _cache feature to named lists.
4886 8. The code for checking quota_filecount in the appendfile transport was
4887 allowing one more file than it should have been.
4889 9. For compatibility with Sendmail, the command line option
4897 and sets the incoming protocol and host name (for trusted callers). The
4898 host name and its colon can be omitted when only the protocol is to be set.
4899 Note the Exim already has two private options, -pd and -ps, that refer to
4900 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4901 "s", but I don't think that's a major issue.
4903 10. A number of refactoring changes to the code, none of which should affect
4906 (a) The number of logging options was getting close to filling up the
4907 32-bit word that was used as a bit map. I have split them into two classes:
4908 those that are passed in the argument to log_write(), and those that are
4909 only ever tested independently outside of that function. These are now in
4910 separate 32-bit words, so there is plenty of room for expansion again.
4911 There is no change in the user interface or the logging behaviour.
4913 (b) When building, for example, log lines, the code previously used a
4914 macro that called string_cat() twice, in order to add two strings. This is
4915 not really sufficiently general. Furthermore, there was one instance where
4916 it was actually wrong because one of the argument was used twice, and in
4917 one call a function was used. (As it happened, calling the function twice
4918 did not affect the overall behaviour.) The macro has been replaced by a
4919 function that can join an arbitrary number of extra strings onto a growing
4922 (c) The code for expansion conditions now uses a table and a binary chop
4923 instead of a serial search (which was left over from when there were very
4924 few conditions). Also, it now recognizes conditions like "pam" even when
4925 the relevant support is not compiled in: a suitably worded error message is
4926 given if an attempt is made to use such a condition.
4928 11. Added ${time_interval:xxxxx}.
4930 12. A bug was causing one of the ddress fields not to be passed back correctly
4931 from remote delivery subprocesses. The field in question was not being
4932 subsequently used, so this caused to problems in practice.
4934 13. Added new log selectors queue_time and deliver_time.
4936 14. Might have fixed a bug in maildirsizefile handling that threw up
4937 "unexpected character" debug warnings, and recalculated the data
4938 unnecessarily. In any case, I expanded the warning message to give more
4941 15. Added the message "Restricted characters in address" to the statements in
4942 the default ACL that block characters like @ and % in local parts.
4944 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4945 Three changes have been made:
4947 (a) There was a serious bug; a negative response to MAIL caused the whole
4948 recipient domain to be cached as invalid, thereby blocking all messages
4949 to all local parts at the same domain, from all senders. This bug has
4950 been fixed. The domain is no longer cached after a negative response to
4951 MAIL if the sender used is not empty.
4953 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4956 (c) A new callout option, "use_sender" has been added for people who want
4957 the modified behaviour.
4963 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4966 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4967 indeed breaks things for older releases.
4969 3. Added additional logging to the case where there is a problem reading data
4970 from a filter that is running in a subprocess using a pipe, in order to
4971 try to track down a specific problem.
4973 4. Testing facility fudge: when running in the test harness and attempting
4974 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4975 getting "No route to host". Convert this to a timeout.
4977 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4980 6. Some OS don't have socklen_t but use size_t instead. This affects the
4981 fifth argument of getsockopt() amongst other things. This is now
4982 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4983 can be set for individual OS. I have set it for SunOS5, OSF1, and
4984 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4985 some earlier ones do not.
4987 7. Change 4.30/15 was not doing the test caselessly.
4989 8. The standard form for an IPv6 address literal was being rejected by address
4990 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4991 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4992 this, as well as the form without the "IPv6" on the front (but only when
4993 address literals are enabled, of course).
4995 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4997 10. Exim crashed if a message with an empty sender address specified by -f
4998 encountered a router with an errors_to setting. This could be provoked only
4999 by a command such as
5003 where an empty string was supplied; "<>" did not hit this bug.
5005 11. Installed PCRE release 4.5.
5007 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5008 remained set. It is now erased.
5010 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5011 times from message ids (which are base 36 rather than the normal 62).
5013 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5014 were being counted as actual protocol errors, and logged if the log
5015 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5016 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5017 and DATA following a set of rejected RCPTs do not count as protocol errors.
5018 In other words, Exim assumes they were pipelined, though this may not
5019 actually be the case. Of course, in all cases the client gets an
5020 appropriate error code.
5022 15. If a lookup fails in an ACL condition, a message about the failure may
5023 be available; it is used if testing the ACL cannot continue, because most
5024 such messages specify what the cause of the deferral is. However, some
5025 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5026 that caused an old message to be retained and used if a later statement
5027 caused a defer, replacing the real cause of the deferral.
5029 16. If an IP address had so many PTR records that the DNS lookup buffer
5030 was not large enough to hold them, Exim could crash while trying to process
5031 the truncated data. It now detects and logs this case.
5033 17. Further to 4.21/58, another change has been made: if (and only if) the
5034 first line of a message (the first header line) ends with CRLF, a bare LF
5035 in a subsequent header line has a space inserted after it, so as not to
5036 terminate the header.
5038 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5039 unnecessarily, used atoi() instead of strtol(), and didn't check the
5040 termination when getting file sizes from file names by regex.
5042 19. Completely re-implemented the support for maildirsize files, in the light
5043 of a number of problems with the previous contributed implementation
5044 (4.30/29). In particular:
5046 . If the quota is zero, the maildirsize file is maintained, but no quota is
5049 . If the maildir directory does not exist, it is created before any attempt
5050 to write a maildirsize file.
5052 . The quota value in the file is just a cache; if the quota is changed in
5053 the transport, the new value overrides.
5055 . A regular expression is available for excluding directories from the
5058 20. The autoreply transport checks the characters in options that define the
5059 message's headers; it allows continued headers, but it was checking with
5060 isspace() after an embedded newline instead of explicitly looking for a
5063 21. If all the "regular" hosts to which an address was routed had passed their
5064 expiry times, and had not reached their retry times, the address was
5065 bounced, even if fallback hosts were defined. Now Exim should go on to try
5068 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5069 equivalent code in the SMTP transport. Some hosts send humungous responses
5070 to HELO/EHLO, more than 1024 it seems.
5072 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5073 gives compiler warnings in some environments. I've now done it "properly",
5076 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5077 (because of problems with the built-in one) was declared to return uschar *
5078 instead of char *, causing compiler failure.
5080 25. Fixed a file descriptor leak when processing alias/forward files.
5082 26. Fixed a minor format string issue in dbfn.c.
5084 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5086 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5087 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5090 29. When a sender address is verified, it is cached, to save repeating the test
5091 when there is more than one recipient in a message. However, when the
5092 verification involves a callout, it is possible for different callout
5093 options to be set for different recipients. It is too complicated to keep
5094 track of this in the cache, so now Exim always runs a verification when a
5095 callout is required, relying on the callout cache for the optimization.
5096 The overhead is duplication of the address routing, but this should not be
5099 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5100 to be verified with callout=postmaster, and the main callout worked but the
5101 postmaster check failed, the verification correctly failed. However, if a
5102 subsequent RCPT command asked for sender verification *without* the
5103 postmaster check, incorrect caching caused this verification also to fail,
5106 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5107 it was not caching the DNS options (qualify_single, search_parents) that
5108 were used when the lookup failed. A subsequent lookup with different
5109 options therefore always gave the same answer, though there were cases
5110 where it should not have. (Example: a "domains = !$mx_any" option on a
5111 dnslookup router: the "domains" option is always processed without any
5112 widening, but the router might have qualify_single set.) Now Exim uses the
5113 cached value only when the same options are set.
5115 32. Added John Jetmore's "exipick" utility to the distribution.
5117 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5118 than a timeout (e.g. a certificate is required, and is not provided), an
5119 Exim server now closes the connection immediately. Previously it waited for
5120 the client to close - but if the client is SSL, it seems that they each
5121 wait for each other, leading to a delay before one of them times out.
5123 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5124 maintained 0.8.x compatibility because I don't think many are using it, and
5125 it is clearly obsolete.
5127 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5130 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5131 client certificate was expired. A simple patch fixes this, though I don't
5132 understand the full logic of why the verify callback is called multiple
5135 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5136 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5137 which causes problems with some clients (such as the Certicom SSL Plus
5138 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5139 disables the coutermeasure allowing Eudora to connect."
5141 38. Exim was not checking that a write() to a log file succeeded. This could
5142 lead to Bad Things if a log got too big, in particular if it hit a file
5143 size limit. Exim now panics and dies if it cannot write to a log file, just
5144 as it does if it cannot open a log file.
5146 39. Modified OS/Makefile-Linux so that it now contains
5148 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5150 The two -D definitions ensure that Exim is compiled with large file
5151 support, which makes it possible to handle log files that are bigger than
5154 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5155 instance) a domain was checked against a named list that involved a lookup,
5156 causing $domain_data to be set, then another domain was checked against the
5157 same list, then the first domain was re-checked, the value of $domain_data
5158 after the final check could be wrong. In particular, if the second check
5159 failed, it could be set empty. This bug probably also applied to
5162 41. The strip_trailing_dot option was not being applied to the address given
5163 with the -f command-line option.
5165 42. The code for reading a message's header from the spool was incrementing
5166 $received_count, but never initializing it. This meant that the value was
5167 incorrect (doubled) while delivering a message in the same process in which
5168 it was received. In the most common configuration of Exim, this never
5169 happens - a fresh exec is done - but it can happen when
5170 deliver_drop_privilege is set.
5172 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5173 it now includes up to 150 characters of the unexpected data in the log
5176 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5177 and building data strings. The size of both of these buffers was 10 000
5178 bytes - far larger than anybody would *ever* want, thought I. Needless to
5179 say, somebody hit the limit. I have increased the maximum line length to
5180 20 000 and the maximum data length of concatenated lines to 100 000. I have
5181 also fixed two bugs, because there was no checking on these buffers. Tsk,
5182 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5183 buffer is too small.
5185 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5186 lsearch lookups. Now it does.
5188 46. When parsing a route_list item in a manualroute router, a fixed-length
5189 buffer was used for the list of hosts. I made this 1024 bytes long,
5190 thinking that nobody would ever have a list of hosts that long. Wrong.
5191 Somebody had a whole pile of complicated expansion conditions, and the
5192 string was silently truncated, leading to an expansion error. It turns out
5193 that it is easier to change to an unlimited length (owing to other changes
5194 that have happened since this code was originally written) than to build
5195 structure for giving a limitation error. The length of the item that
5196 expands into the list of hosts is now unlimited.
5198 47. The lsearch lookup could not handle data where the length of text line was
5199 more than 4095 characters. Such lines were truncated, leading to shortened
5200 data being returned. It should now handle lines of any length.
5202 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5203 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5206 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5207 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5208 no longer happens in comment lines. A list of replacements is now placed
5209 at the head of all of the source files, except those whose only change is
5210 to replace PERL_COMMAND in the very first #! line.
5212 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5213 messages on the queue, with a bottom-up merge sort, using code contributed
5214 by Michael Haardt. This should make operations like -bp somewhat faster on
5215 large queues. It won't affect queue runners, except when queue_run_in_order
5218 51. Installed eximstats 1.31 in the distribution.
5220 52. Added support for SRV lookups to the dnslookup router.
5222 53. If an ACL referred to $message_body or $message_body_end, the value was not
5223 reset for any messages that followed in the same SMTP session.
5225 54. The store-handling optimization for building very long strings was not
5226 differentiating between the different store pools. I don't think this
5227 actually made any difference in practice, but I've tidied it.
5229 55. While running the routers to verify a sender address, $sender_address
5230 was still set to the sender address. This is wrong, because when routing to
5231 send a bounce to the sender, it would be empty. Therefore, I have changed
5232 it so that, while verifying a sender address, $sender_address is set to <>.
5233 (There is no change to what happens when verifying a recipient address.)
5235 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5236 target A or AAAA records (if not already returned) without resetting the
5237 qualify_single or search_parents options of the DNS resolver. These are
5238 inappropriate in this case because the targets of MX and SRV records must
5239 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5240 target that, when qualified, matched something in the local domain. These
5241 two options are now turned off when doing these lookups.
5243 57. It seems that at least some releases of Reiserfs (which does not have the
5244 concept of a fixed number of inodes) returns zero and not -1 for the
5245 number of available inodes. This interacted badly with check_spool_inodes,
5246 which assumed that -1 was the "no such thing" setting. What I have done is
5247 to check that the total number of inodes is greater than zero before doing
5248 the test of how many are available.
5250 58. When a "warn" ACL statement has a log_message modifier, the message is
5251 remembered, and not repeated. This is to avoid a lot of repetition when a
5252 message has many recipients that cause the same warning to be written.
5253 Howewer, Exim was preserving the list of already written lines for an
5254 entire SMTP session, which doesn't seem right. The memory is now reset if a
5255 new message is started.
5257 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5258 debugging output unless log_rewrite was also set.
5260 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5261 of (int)(handle) when we know that handle contains (void *)(-1).
5263 61. The Exim daemon panic-logs an error return when it closes the incoming
5264 connection. However "connection reset by peer" seems to be common, and
5265 isn't really an error worthy of noting specially, so that particular error
5268 62. When Exim is trying to find all the local interfaces, it used to panic and
5269 die if the ioctl to get the interface flags failed. However, it seems that
5270 on at least one OS (Solaris 9) it is possible to have an interface that is
5271 included in the list of interfaces, but for which you get a failure error
5272 for this call. This happens when the interface is not "plumbed" into a
5273 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5274 failure of the "get flags" call assumes that the interface is down.
5276 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5277 makes life easier for people who are doing arithmetic on fields extracted
5278 from dates, where you often get leading zeros that should not be
5279 interpreted as octal.
5281 64. Added qualify_domain to the redirect router, to override the global
5284 65. If a pathologically long header line contained very many addresses (the
5285 report of this problem mentioned 10 000) and each of them was rewritten,
5286 Exim could use up a very large amount of memory. (It kept on making new
5287 copies of the header line as it rewrote, and never released the old ones.)
5288 At the expense of a bit more processing, the header rewriting function has
5289 been changed so that it no longer eats memory in this way.
5291 66. The generation of the Received: header has been moved from the time that a
5292 message starts to be received, to the time that it finishes. The timestamp
5293 in the Received: header should now be very close to that of the <= log
5294 line. There are two side-effects of this change:
5296 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5297 logged header lines no longer include the local Received: line, because
5298 it has not yet been created. The same applies to a copy of the message
5299 that is returned to a non-SMTP sender when a message is rejected.
5301 (b) When a filter file is tested using -bf, no additional Received: header
5302 is added to the test message. After some thought, I decided that this
5305 This change does not affect the value of $received_for. It is still set
5306 after address rewriting, but before local_scan() is called.
5308 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5310 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5311 gave an unhelpful panic error message, and a defer error. I have managed to
5312 change this behaviour so that it now rejects any supplied certificate,
5313 which seems right, as the list of acceptable certificates is empty.
5315 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5316 gave an unhelpful defer error. I have not managed to make this reject any
5317 supplied certificates, but the error message it gives is "no certificate
5318 supplied", which is not helpful.
5320 70. exigrep's output now also includes lines that are not associated with any
5321 message, but which match the given pattern. Implemented by a patch from
5322 Martin Sluka, which also tidied up the Perl a bit.
5324 71. Recipient callout verification, like sender verification, was using <> in
5325 the MAIL FROM command. This isn't really the right thing, since the actual
5326 sender may affect whether the remote host accepts the recipient or not. I
5327 have changed it to use the actual sender in the callout; this means that
5328 the cache record is now keyed on a recipient/sender pair, not just the
5329 recipient address. There doesn't seem to be a real danger of callout loops,
5330 since a callout by the remote host to check the sender would use <>.
5331 [SEE ABOVE: changed after hitting problems.]
5333 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5334 temporary errors. However, in the case of such a code being given after
5335 the end of a data transmission (i.e. after ".") Exim was failing to write
5336 a retry record for the message. (Yes, there was some broken host that was
5337 actually sending 8xx at this point.)
5339 73. An unknown lookup type in a host list could cause Exim to panic-die when
5340 the list was checked. (An example that provoked this was putting <; in the
5341 middle of a list instead of at the start.) If this happened during a DATA
5342 ACL check, a -D file could be left lying around. This kind of configuration
5343 error no longer causes Exim to die; instead it causes a defer errror. The
5344 incident is still logged to the main and panic logs.
5346 74. Buglet left over from Exim 3 conversion. The message "too many messages
5347 in one connection" was written to the rejectlog but not the mainlog, except
5348 when address rewriting (yes!) was being logged.
5350 75. Added write_rejectlog option.
5352 76. When a system filter was run not as root (that is, when system_filter_user
5353 was set), the values of the $n variables were not being returned to the
5354 main process; thus, they were not subsequently available in the $sn
5357 77. Added +return_path_on_delivery log selector.
5359 78. A connection timeout was being treated differently from recipients deferred
5360 when testing hosts_max_try with a message that was older than the host's
5361 retry timeout. (The host should not be counted, thus allowing all hosts to
5362 be tried at least once before bouncing.) This may have been the cause of an
5363 occasionally reported bug whereby a message would remain on the queue
5364 longer than the retry timeout, but would be bounced if a delivery was
5365 forced. I say "may" because I never totally pinned down the problem;
5366 setting up timeout/retry tests is difficult. See also the next item.
5368 79. The ultimate address timeout was not being applied to errors that involved
5369 a combination of host plus message (for example, a timeout on a MAIL
5370 command). When an address resolved to a number of possible hosts, and they
5371 were not all tried for each delivery (e.g. because of hosts_max_try), a
5372 message could remain on the queue longer than the retry timeout.
5374 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5377 81. Fixed an obscure SMTP outgoing bug which required at least the following
5378 conditions: (a) there was another message waiting for the same server;
5379 (b) the server returned 5xx to all RCPT commands in the first message so
5380 that the message was not completed; (c) the server dropped the connection
5381 or gave a negative response to the RSET that Exim sends to abort the
5382 transaction. The observed case was a dropped connection after DATA that had
5383 been sent in pipelining mode. That is, the server had advertised PIPELINING
5384 but was not implementing it correctly. The effect of the bug was incorrect
5385 behaviour, such as trying another host, and this could lead to a crash.
5391 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5392 and daemon.c were passed as pointers to ints; they should have been
5393 pointers to socklen_t variables (which are typically unsigned ints).
5395 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5398 3. Fixed a really odd bug that affected only the testing scheme; patching a
5399 certain fixed string in the binary changed the value of another string that
5400 happened to be identical to the end of the original first string.
5402 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5403 name", it returns that address as the IP address. On some operating
5404 systems (e.g. Solaris), it also passes back the IP address string as the
5405 "host name". However, on others (e.g. Linux), it passes back an empty
5406 string. Exim wasn't checking for this, and was changing the host name to an
5407 empty string, assuming it had been canonicized.
5409 5. Although rare, it is permitted to have more than one PTR record for a given
5410 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5411 all the names associated with an address, because they do in Solaris.
5412 However, it seems that they do not in Linux for data that comes from the
5413 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5414 I found this out when I moved to a new Linux workstation and tried to run
5415 the Exim test suite.
5417 To get round this problem I have changed the code so that it now does its
5418 own call to the DNS to look up PTR records when searching for a host name.
5419 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5420 addresses that are only in /etc/hosts are still found.
5422 This behaviour is, however, controlled by an option called host_lookup_
5423 order, which defaults to "bydns:byaddr". If people want to use the other
5424 order, or indeed, just use one or the other means of lookup, they can
5425 specify it in this variable.
5427 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5428 some operating systems, this comes back from gethostbyaddr() as an empty
5429 string, and this is what Exim used to test for. However, it seems that in
5430 other systems, "." is yielded. Exim now tests for this case too.
5432 7. The values of check_spool_space and check_log_space are now held internally
5433 as a number of kilobytes instead of an absolute number of bytes. If a
5434 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5435 kilobyte. This means that much larger values can be stored.
5437 8. Exim monitor: an attempt to get the action menu when not actually pointing
5438 at a message produces an empty menu entitled "No message selected". This
5439 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5440 no entries in it ("Shell widget menu has zero width and/or height"). So I
5441 have added a single, blank menu entry in this case.
5443 9. Added ${quote_local_part.
5445 10. MIME decoding is now applied to the contents of Subject: header lines when
5448 11. Now that a reference to $sender_host_address automatically causes a reverse
5449 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5450 host lookup before query-style lookups in lists that might use this
5451 variable. This has therefore been abolished, and the "net-" prefix is no
5452 longer necessary for query-style lookups.
5454 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5455 have been a typo for LFLAGS, so it has been changed.
5457 13. The install script calls Exim with "-C /dev/null" in order to find the
5458 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5459 to be output. Howeve, since Exim outputs its version number before the
5460 error, it didn't break the script. It just looked ugly. I fixed this by
5461 always allowing "-C /dev/null" if the caller is root.
5463 14. Ignore overlarge ACL variable number when reading spool file - insurance
5464 against a later release with more variables having written the file.
5466 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5467 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5468 this, as well as the form without the "IPv6" on the front.
5470 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5471 OS/Makefile-Darwin file.
5473 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5474 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5476 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5477 variable. However, some OS use ints for inodes. I've added cast to long int
5478 to get rid of the compiler warning.
5480 19. I had forgotten to lock out "/../" in configuration file names when
5481 ALT_CONFIG_PREFIX was set.
5483 20. Routers used for verification do not need to specify transports. However,
5484 if such a router generated a host list, and callout was configured, Exim
5485 crashed, because it could not find a port number from the (non-existent)
5486 transport. It now assumes port 25 in this circumstance.
5488 21. Added the -t option to exigrep.
5490 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5491 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5492 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5493 them is set, LOOKUP_LSEARCH is forced.
5495 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5497 24. Added sender and host information to the "rejected by local_scan()" log
5498 line; previously there was no indication of these.
5500 25. Added .include_if_exists.
5502 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5503 Linux. It turns out that not all file systems support this. Apparently some
5504 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5505 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5506 not supported on the file descriptor, is now ignored when Exim is trying to
5507 sync a directory. This applies only to Linux.
5509 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5511 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5512 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5515 29. Added support for maildirsize files from supplied patch (modified a bit).
5517 30. The use of :fail: followed by an empty string could lead Exim to respond to
5518 sender verification failures with (e.g.):
5520 550 Verification failed for <xxx>
5521 550 Sender verify failed
5523 where the first response line was missing the '-' that indicates it is not
5524 the final line of the response.
5526 31. The loop for finding the name of the user that called Exim had a hardwired
5527 limit of 10; it now uses the value of finduser_retries, which is used for
5528 all other user lookups.
5530 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5533 33. Exim was neglecting to zero errno before one call of strtol() when
5534 expanding a string and expecting an integer value. On some systems this
5535 resulted in spurious "integer overflow" errors. Also, it was casting the
5536 result into an int without checking.
5538 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5539 did not work. The code looks as if it has *never* worked, though it appears
5540 to have been documented since at least releast 1.62. I have made it work.
5542 35. The "timeout_DNS" error in retry rules, also documented since at least
5543 1.62, also never worked. As it isn't clear exactly what this means, and
5544 clearly it isn't a major issue, I have abolished the feature by treating it
5545 as "timeout", and writing a warning to the main and panic logs.
5547 36. The display of retry rules for -brt wasn't always showing the error code
5550 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5551 timeout_connect_A, timeout_connect_MX.
5553 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5554 to the empty sender.
5556 39. The daemon was not analysing the content of -oX till after it had closed
5557 stderr and disconnected from the controlling terminal. This meant that any
5558 syntax errors were only noted on the panic log, and the return code from
5559 the command was 0. By re-arranging the code a little, I've made the
5560 decoding happen first, so such errors now appear on stderr, and the return
5561 code is 1. However, the actual setting up of the sockets still happens in
5562 the disconnected process, so errors there are still only recorded on the
5565 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5566 connections (as happens on some IP stacks) was logged at start up time as
5567 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5568 it from "IPv6 and IPv4", which means that two separate sockets are being
5571 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5572 says whether AF_INET or AF_INET6 was passed as an argument.
5574 42. Exiwhat output was messed up when time zones were included in log
5577 43. Exiwhat now gives more information about the daemon's listening ports,
5578 and whether -tls-on-connect was used.
5580 44. The "port" option of the smtp transport is now expanded.
5582 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5583 silently ignored. Now an error message is written to the main and panic
5586 46. There's a new ACL modifier called "logwrite" which writes to a log file
5587 as soon as it is encountered.
5589 47. Added $local_user_uid and $local_user_gid at routing time.
5591 48. Exim crashed when trying to verify a sender address that was being
5594 49. Exim was recognizing only a space character after ".include". It now also
5595 recognizes a tab character.
5597 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5598 extracting the relevant information from the specification. The man page no
5599 longer contains scrambled data for the -d option, and I've added a section
5600 at the front about calling Exim under different names.
5602 51. Added "extra_headers" argument to the "mail" command in filter files.
5604 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5607 53. Installed eximstats 1.29.
5609 54. Added transport_filter_timeout as a generic transport option.
5611 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5612 Cc: header lines. This was required by RFC 822, but it not required by RFC
5615 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5616 incoming messages that did not have them. Now it does so only if the
5617 message originates locally, that is, if there is no associated remote host
5618 address. When Resent- header lines are present, this applies to the Resent-
5619 lines rather than the non-Resent- lines.
5621 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5622 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5624 58. Messages for configuration errors now include the name of the main
5625 configuration file - useful now that there may be more than one file in a
5626 list (.included file names were always shown).
5628 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5629 for those rare installations that do not start the daemon as root or run it
5630 setuid root. I've cut out the call to initgroups() if the daemon is not
5633 60. The Exim user and group can now be bound into the binary as text strings
5634 that are looked up at the start of Exim's processing.
5636 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5638 62. Added $mailstore_basename variable.
5640 63. Installed patch to sieve.c from Michael Haardt.
5642 64. When Exim failed to open the panic log after failing to open the main log,
5643 the original message it was trying to log was written to stderr and debug
5644 output, but if they were not available (the usual case in production), it
5645 was lost. Now it is written to syslog before the two lines that record the
5646 failures to open the logs.
5648 65. Users' Exim filters run in subprocesses under the user's uid. It is
5649 possible for a "deliver" command or an alias in a "personal" command to
5650 provoke an address rewrite. If logging of address rewriting is configured,
5651 this fails because the process is not running as root or exim. There may be
5652 a better way of dealing with this, but for the moment (because 4.30 needs
5653 to be released), I have disabled address rewrite logging when running a
5654 filter in a non-root, non-exim process.
5660 1. The buildconfig auxiliary program wasn't quoting the value set for
5661 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5662 not defined. This bug was masked in 4.22 by the effect that was fixed in
5665 2. Some messages that were rejected after a message id was allocated were
5666 shown as "incomplete" by exigrep. It no longer does this for messages that
5667 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5669 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5670 have allow_domain_literals set, the ID did not get logged in the <= line.
5671 Domain literals are now always recognized in Message-ID: header lines.
5673 4. The first argument for a ${extract expansion item is the key name or field
5674 number. Leading and trailing spaces in this item were not being ignored,
5675 causing some misleading effects.
5677 5. When deliver_drop_privilege was set, single queue runner processes started
5678 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5679 same command in the process it spins off) were not dropping privilege.
5681 6. When the daemon running as "exim" started a queue runner, it always
5682 re-executed Exim in the spun-off process. This is a waste of effort when
5683 deliver_drop_privilege is set. The new process now just calls the
5684 queue-runner function directly.
5690 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5693 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5694 ignored. Though the use of -oP was forcing the writing of a pid file, it
5695 was always written to the default place.
5697 3. If the message "no IP address found for host xxxx" is generated during
5698 incoming verification, it is now followed by identification of the incoming
5699 connection (so you can more easily find what provoked it).
5701 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5703 5. Added some features to "harden" Exim a bit more against certain attacks:
5705 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5706 be put in Local/Makefile. This is like the never_users runtime option,
5707 but it cannot be overridden. The default setting is "root".
5709 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5710 prefix string with which any file named in a -C command line option
5713 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5714 is retained for -C and -D only if the caller of Exim is root. Without
5715 it, the exim user may also use -C and -D and retain privilege.
5717 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5718 command line option is disabled.
5720 6. Macro names set by the -D option must start with an upper case letter, just
5721 like macro names defined in the configuration file.
5723 7. Added "dereference=" facility to LDAP.
5725 8. Two instances of the typo "uknown" in the source files are fixed.
5727 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5728 the Configure-Makefile script screwed up while processing it.
5730 10. Incorporated PCRE 4.4.
5732 11. The SMTP synchronization check was not operating right at the start of an
5733 SMTP session. For example, it could not catch a HELO sent before the client
5734 waited for the greeting. There is now a check for outstanding input at the
5735 point when the greeting is written. Because of the duplex, asynchronous
5736 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5737 way, but not yet received, when the check is performed.
5739 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5740 on TCP/IP sockets, because this apparently causes some broken clients to
5743 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5744 unchanged) from the Cygwin maintainer.
5746 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5747 was supported instead of testing for mbx. Effectively a typo.
5749 15. The spa authenticator server code was not checking that the input it
5750 received was valid base64.
5752 16. The debug output line for the "set" modifier in ACLs was not showing the
5753 name of the variable that was being set.
5755 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5757 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5758 lookup on its own. Something else was needed to trigger the lookup. For
5759 example, a match in host_lookup or the need for a host name in a host list.
5760 Now, if $sender_host_name is referenced and the host name has not yet been
5761 looked up, a lookup is performed. If the lookup fails, the variable remains
5762 empty, and $host_lookup_failed is set to "1".
5764 19. Added "eqi" as a case-independent comparison operator.
5766 20. The saslauthd authentication condition could segfault if neither service
5767 nor realm was specified.
5769 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5770 error message that was logged was misleading, and incoming SMTP
5771 connections were dropped. The message is now more accurate, and temporary
5772 errors are given to SMTP connections.
5774 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5775 (except RSET) with a 421 error, until QUIT is received. However, it was
5776 failing to send a response to QUIT.
5778 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5779 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5780 ACL is now run after the helo_try_verify_hosts code.
5782 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5783 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5784 case-independent, so other case variants are also recognized). Apparently
5785 some systems use these upper case variants.
5787 25. If more than two messages were waiting for the same host, and a transport
5788 filter was specified for the transport, Exim sent two messages over the
5789 same TCP/IP connection, and then failed with "socket operation on non-
5790 socket" when it tried to send the third.
5792 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5794 27. The extern definition of crypt16() in expand.c was not being excluded when
5795 the OS had its own crypt16() function.
5797 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5798 as a preferred synonym for return_size_limit, both as an option and as an
5801 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5803 30. Changed the default configuration ACL to relax the local part checking rule
5804 for addresses that are not in any local domains. For these addresses,
5805 slashes and pipe symbols are allowed within local parts, but the sequence
5806 /../ is explicitly forbidden.
5808 31. SPA server authentication was not clearing the challenge buffer before
5811 32. log_message in a "warn" ACL statement was writing to the reject log as
5812 well as to the main log, which contradicts the documentation and doesn't
5813 seem right (because no rejection is happening). So I have stopped it.
5815 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5816 However, I am unable to do any testing of this.
5818 34. Fixed an infelicity in the appendfile transport. When checking directories
5819 for a mailbox, to see if any needed to be created, it was accidentally
5820 using path names with one or more superfluous leading slashes; tracing
5821 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5823 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5824 contents are added to the log line that is written for every discarded
5825 recipient. (Previously a log_message setting was ignored.)
5827 36. The ${quote: operator now quotes the string if it is empty.
5829 37. The install script runs exim in order to find its version number. If for
5830 some reason other than non-existence or emptiness, which it checks, it
5831 could not run './exim', it was installing it with an empty version number,
5832 i.e. as "exim-". This error state is now caught, and the installation is
5835 38. An argument was missing from the function that creates an error message
5836 when Exim fails to connect to the socket for saslauthd authentication.
5837 This could cause Exim to crash, or give a corrupted message.
5839 39. Added isip, isip4, and isip6 to ${if conditions.
5841 40. The ACL variables $acl_xx are now saved with the message, and can be
5842 accessed later in routers, transports, and filters.
5844 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5845 strings in the file are not string-expanded.
5847 42. If a MAIL command specified a SIZE value that was too large to fit into an
5848 int variable, the check against message_size_limit failed. Such values are
5849 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5850 day this will have to be increased, but I don't think I want to be around
5851 when emails are that large.
5858 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5859 iconv() is not standard in FreeBSD.
5861 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5862 IPv6 enabled. The observed symptom was a segmentation fault on return from
5863 the function os_common_find_running_interfaces() in src/os.c.
5865 3. In the check_special_case() function in daemon.c I had used "errno" as an
5866 argument name, which causes warnings on some systems. This was basically a
5867 typo, since it was named "eno" in the comments!
5869 4. The code that waits for the clock to tick (at a resolution of some fraction
5870 of a second) so as to ensure message-id uniqueness was always waiting for
5871 at least one whole tick, when it could have waited for less. [This is
5872 almost certainly not relevant at current processor speeds, where it is
5873 unlikely to ever wait at all. But we try to future-proof.]
5875 5. The function that sleeps for a time interval that includes fractions of a
5876 second contained a race. It did not block SIGALRM between setting the
5877 timer, and suspending (a couple of lines later). If the interval was short
5878 and the sigsuspend() was delayed until after it had expired, the suspension
5879 never ended. On busy systems this could lead to processes getting stuck for
5882 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5883 process, before it forks any delivery processes. The open lookup caching
5884 mechanism meant that the open file or database connection was passed into
5885 the delivery process. The problem was that delivery processes always tidy
5886 up cached lookup data. This could cause a problem for the next delivery
5887 process started by the queue runner, because the external queue runner
5888 process does not know about the closure. So the next delivery process
5889 still has data in the lookup cache. In the case of a file lookup, there was
5890 no problem because closing a file descriptor in a subprocess doesn't affect
5891 the parent. However, if the lookup was caching a connection to a database,
5892 the connection was closed, and the second delivery process was likely to
5893 see errors such as "PGSQL: query failed: server closed the connection
5894 unexpectedly". The problem has been fixed by closing all cached lookups
5895 in a queue runner before running a delivery process.
5897 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5898 seem to have the "const" qualifier which it has on other OS. I've
5901 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5902 *of the same type* (client or server) with the same public name that an
5903 error should be diagnosed.
5905 9. When Exim looked up a host name for an IP address, but failed to find the
5906 original IP address when looking up the host name (a safety check), it
5907 output the message "<ip address> does not match any IP for NULL", which was
5908 confusing, to say the least. The bug was that the host name should have
5909 appeared instead of "NULL".
5911 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5912 user that is built into the binary, and the -C or -D options is used, root
5913 privilege is dropped before the configuration file is read. In addition,
5914 logging is switched to stderr instead of the normal log files. If the
5915 configuration then re-defines the Exim user, the unprivileged environment
5916 is probably not what is expected, so Exim logs a panic warning message (but
5919 However, if deliver_drop_privilege is set, the unprivileged state may well
5920 be exactly what is intended, so the warning has been cut out in that case,
5921 and Exim is allowed to try to write to its normal log files.
5927 1. smtp_return_error_details was not giving details for temporary sender
5928 or receiver verification errors.
5930 2. Diagnose a configuration error if two authenticators have the same public
5933 3. Exim used not to create the message log file for a message until the first
5934 delivery attempt. This could be confusing when incoming messages were held
5935 for policy or load reasons. The message log file is now created at the time
5936 the message is received, and an initial "Received" line is written to it.
5938 4. The automatically generated man page for command line options had a minor
5939 bug that caused no ill effects; however, a more serious problem was that
5940 the procedure for building the man page automatically didn't always
5941 operate. Consequently, release 4.20 contains an out-of-date version. This
5942 shouldn't happen again.
5944 5. When building Exim with embedded Perl support, the script that builds the
5945 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5946 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5948 6. The freeze_tell option was not being used for messages that were frozen on
5949 arrival, either by an ACL or by local_scan().
5951 7. Added the smtp_incomplete_transaction log selector.
5953 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5954 was accepting AUTH without a new EHLO.
5956 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5957 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5960 10. Logging of TCP/IP connections (when configured) now happens in the main
5961 daemon process instead of the child process, so that the TCP/IP connection
5962 count is more accurate (but it can never be perfect).
5964 11. The use of "drop" in a nested ACL was not being handled correctly in the
5965 outer ACL. Now, if condition failure induced by the nested "drop" causes
5966 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5967 or "require"), the connection is dropped.
5969 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5970 that yield "discard" can now be used with an "accept" or a "discard" verb,
5971 but an error is generated for any others (because I can't see a useful way
5972 to define what should happen).
5974 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5975 are now processed in the same way as lines in the Exim configuration file.
5976 In particular, continuation lines are supported.
5978 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5980 15. Added -ti meaning -t -i.
5982 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5983 domains, and warn by logging if others are found.
5985 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5986 returned from GIFCONF when Exim is trying to find the list of interfaces on
5987 a host. The code in os.c has been modified to copy each ifreq to an aligned
5988 structure in all cases.
5990 Also, in some cases, the returned ifreq's were being copied to a 'struct
5991 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5992 means the last couple of bytes of an IPv6 address could be chopped if the
5993 ifreq contained only a normal sockaddr (14 bytes storage).
5995 18. Named domain lists were not supported in the hosts_treat_as_local option.
5996 An entry such as +xxxx was not recognized, and was treated as a literal
5999 19. Ensure that header lines added by a DATA ACL are included in the reject log
6000 if the ACL subsequently rejects the message.
6002 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6003 MD5 (which is deprecated).
6005 21. When testing a filter file using -bf, Exim was writing a message when it
6006 took the sender from a "From " line in the message, but it was not doing so
6007 when it took $return_path from a Return-Path: header line. It now does.
6009 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6010 with a valid header line field name (a series of printing characters
6011 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6013 23. Changed "disc" in the source to "disk" to conform to the documentation and
6014 the book and for uniformity.
6016 24. Ignore Sendmail's -Ooption=value command line item.
6018 25. When execve() failed while trying to run a command in a pipe transport,
6019 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6020 could be confused with a return value of 69 from the command itself. This
6021 has been changed to 127, the value the shell returns if it is asked to run
6022 a non-existent command. The wording for the related log line suggests a
6023 non-existent command as the problem.
6025 26. If received_header_text expands to an empty string, do not add a Received:
6026 header line to the message. (Well, it adds a token one on the spool, but
6027 marks it "old" so that it doesn't get used or transmitted.)
6029 27. Installed eximstats 1.28 (addition of -nt option).
6031 28. There was no check for failure on the call to getsockname() in the daemon
6032 code. This can fail if there is a shortage of resources on the system, with
6033 ENOMEM, for example. A temporary error is now given on failure.
6035 29. Contrary to the C standard, it seems that in some environments, the
6036 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6037 program. Exim now does this explicitly; it affects the formatting of
6038 timestamps using strftime().
6040 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6041 complaints. I've now initialized all the variables, to avoid this.
6043 32. Header lines added by a system filter were not being "seen" during
6044 transport-time rewrites.
6046 33. The info_callback() function passed to OpenSSL is set up with type void
6047 (*)(SSL *, int, int), as described somewhere. However, when calling the
6048 function (actually a macro) that sets it up, the type void(*)() is
6049 expected. I've put in a cast to prevent warnings from picky compilers.
6051 34. If a DNS black list lookup found a CNAME record, but there were no A
6052 records associated with the domain it pointed at, Exim crashed.
6054 35. If a DNS black list lookup returned more than one A record, Exim ignored
6055 all but the first. It now scans all returned addresses if a particular IP
6056 value is being sought. In this situation, the contents of the
6057 $dnslist_value variable are a list of all the addresses, separated by a
6060 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6061 to accept a host name and all its aliases if the forward lookup for any of
6062 them yielded the IP address of the incoming connection. Now it accepts only
6063 those names whose forward lookup yields the correct IP address. Any other
6064 names are discarded. This closes a loophole whereby a rogue DNS
6065 administrator could create reverse DNS records to break through a
6066 wildcarded host restriction in an ACL.
6068 37. If a user filter or a system filter that ran in a subprocess used any of
6069 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6070 the wrong values were passed to the pipe command ($thisaddress had the
6071 value of $0, $0 had the value of $1, etc). This bug was introduced by
6072 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6074 38. Improved the line breaking for long SMTP error messages from ACLs.
6075 Previously, if there was no break point between 40 and 75 characters, Exim
6076 left the rest of the message alone. Two changes have been made: (a) I've
6077 reduced the minimum length to 35 characters; (b) if it can't find a break
6078 point between 35 and 75 characters, it looks ahead and uses the first one
6079 that it finds. This may give the occasional overlong line, but at least the
6080 remaining text gets split now.
6082 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6083 file descriptors might be low, and that setting 1000 would always raise it.
6084 It turns out that in some environments, the limit is already over 1000 and
6085 that lowering it causes trouble. So now Exim takes care not to decrease it.
6087 40. When delivering a message, the value of $return_path is set to $sender_
6088 address at the start of routing (routers may change the value). By an
6089 oversight, this default was not being set up when an address was tested by
6090 -bt or -bv, which affected the outcome if any router or filter referred to
6093 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6094 log only when "message" or "log_message" are set. However, if one of the
6095 conditions was an address verification, or a call to a nested ACL, the
6096 messages generated by the underlying test were being passed through. This
6097 no longer happens. The underlying message is available in $acl_verify_
6098 message for both "message" and "log_message" expansions, so it can be
6099 passed through if needed.
6101 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6102 new expansion $bh_ to give the encoded byte string without charset
6103 translation. Translation happens only if iconv() is available; HAVE_ICONV
6104 indicates this at build time. HEADERS_CHARSET gives the charset to
6105 translate to; headers_charset can change it in the configuration, and
6106 "headers charset" can change it in an individual filter file.
6108 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6109 that creates RFC 2047 encoded "words" labels them as that charset instead
6110 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6111 expansion operator; (ii) when Exim creates a From: line for a local
6112 message; (iii) when a header line is rewritten to include a "phrase" part.
6114 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6115 buggy, causing it to skip the first lines of messages whose message ID
6116 ended in 'D'. This would not have bitten before Exim release 4.14, because
6117 message IDs were unlikely to end in 'D' before then. The effect was to have
6118 incorrect size information for certain domains.
6120 45. #include "config.h" was missing at the start of the crypt16.c module. This
6121 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6124 46. If there was a timeout during a "random" callout check, Exim treated it as
6125 a failure of the random address, and carried on sending RSET and the real
6126 address. If the delay was just some slowness somewhere, the response to the
6127 original RCPT would be taken as a response to RSET and so on, causing
6128 mayhem of various kinds.
6130 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6131 when I implemented it. It didn't allow for the fact that some option values
6132 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6133 the right test for positive values.
6135 48. Domain names in DNS records are case-independent. Exim always looks them up
6136 in lower case. Some resolvers return domain names in exactly the case they
6137 appear in the zone file, that is, they may contain uppercase letters. Not
6138 all resolvers do this - some return always lower case. Exim was treating a
6139 change of case by a resolver as a change of domain, similar to a widening
6140 of a domain abbreviation. This triggered its re-routing code and so it was
6141 trying to route what was effectively the same domain again. This normally
6142 caused routing to fail (because the router wouldn't handle the domain
6143 twice). Now Exim checks for this case specially, and just changes the
6144 casing of the domain that it ultimately uses when it transmits the message
6147 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6150 50. If a filter generated a file delivery with a non-absolute name (possible if
6151 no home directory exists for the router), the forbid_file option was not
6154 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6155 the existing equality matching.
6157 52. Exim was using ints instead of ino_t variables in some places where it was
6158 dealing with inode numbers.
6160 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6161 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6162 called TMPDIR, and if it finds it is different, it changes its value.
6164 54. The smtp_printf() function is now made available to local_scan() so
6165 additional output lines can be written before returning. There is also an
6166 smtp_fflush() function to enable the detection of a dropped connection.
6167 The variables smtp_input and smtp_batched_input are exported to
6170 55. Changed the default runtime configuration: the message "Unknown user"
6171 has been removed from the ACL, and instead placed on the localuser router,
6172 using the cannot_route_message feature. This means that any verification
6173 failures that generate their own messages won't get overridden. Similarly,
6174 the "Unrouteable address" message that was in the ACL for unverifiable
6175 relay addresses has also been removed.
6177 56. Added hosts_avoid_esmtp to the smtp transport.
6179 57. The exicyclog script was not checking for the esoteric option
6180 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6181 will work only if exicyclog is run under the appropriate euid.
6183 58. Following a discussion on the list, the rules by which Exim recognises line
6184 endings on incoming messages have been changed. The -dropcr and drop_cr
6185 options are now no-ops, retained only for backwards compatibility. The
6186 following line terminators are recognized: LF CRLF CR. However, special
6187 processing applies to CR:
6189 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6190 nor a local message in the state where . is a terminator.
6192 (ii) If a bare CR is encountered in a header line, an extra space is added
6193 after the line terminator so as not to end the header. The reasoning
6194 behind this is that bare CRs in header lines are most likely either
6195 to be mistakes, or people trying to play silly games.
6197 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6198 was being incorrectly given as 18 bytes larger than it should have been.
6199 This is a VOB (very old bug).
6201 60. This may never have affected anything current, but just in case it has:
6202 When the local host is found other than at the start of a list of hosts,
6203 the local host, those with the same MX, and any that follow, are discarded.
6204 When the list in question was part of a longer list of hosts, the following
6205 hosts (not currently being processed) were also being discarded. This no
6206 longer happens. I'm not sure if this situation could ever has previously
6209 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6212 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6213 header. This is something that is recommended in a new Internet Draft, and
6214 is something that is documented as being done by Sendmail. There are two
6215 possible values. For messages generated by the autoreply transport, Exim
6218 Auto-Submitted: auto-replied
6220 whereas for all other generated messages (e.g. bounces) it adds
6222 Auto-Submitted: auto-generated
6224 63. The "personal" condition in filters now includes a test for the
6225 Auto-Submitted: header. If it contains the string "auto-" the message it
6226 not considered personal.
6228 64. Added rcpt_include_affixes as a generic transport option.
6230 65. Added queue_only_override (default true).
6232 66. Added the syslog_duplication option.
6234 67. If what should have been the first header line of a message consisted of
6235 a space followed by a colon, Exim was mis-interpreting it as a header line.
6236 It isn't of course - it is syntactically invalid and should therefore be
6237 treated as the start of the message body. The misbehaviour could have
6238 caused a number of strange effects, including loss of data in subsequent
6239 header lines, and spool format errors.
6241 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6242 client host had authenticated. This control can now be exercised by an ACL
6243 for more flexibility.
6245 69. By default, callouts do not happen when testing with -bh. There is now a
6246 variant, -bhc, which does actually run the callout code, including
6247 consulting and updating the callout cache.
6249 70. Added support for saslauthd authentication, courtesy of Alexander
6252 71. If statvfs() failed on the spool or log directories while checking their
6253 size for availability, Exim confusingly gave the error "space shortage".
6254 Furthermore, in debugging mode it crashed with a floating point exception.
6255 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6256 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6257 serious problem, Exim now writes to the main and panic logs when this
6258 happens, with details of the failure. It then refuses to accept the
6259 incoming message, giving the message "spool directory problem" or "log
6260 directory problem" with a 421 code for SMTP messages.
6262 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6263 0, 1, and 2 exist, because some OS complain for execs without them (see
6264 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6265 descriptors. However, the code omitted to check that the open succeeded,
6266 causing mysterious errors if for some reason the permissions on /dev/null
6267 got screwed. Now Exim writes a message to the main and panic logs, and
6268 bombs out if it can't open /dev/null.
6270 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6271 interact so that it is all more flexible. It is supposed to remain
6272 backwards compatible. Also added extra_local_interfaces.
6274 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6275 to bomb out with an assertion failure - to the client this appears as a
6276 connection drop. This problem occurs in the part of the code that was taken
6277 from the Samba project. Fortunately, the assertion is in a very simple
6278 function, so I have fixed this by reproducing the function inline in the
6279 one place where it is called, and arranging for authentication to fail
6280 instead of killing the process with assert().
6282 75. The SPA client code was not working when the server requested OEM rather
6283 than Unicode encoding.
6285 76. Added code to make require_files with a specific uid setting more usable in
6286 the case where statting the file as root fails - usually a non-root-mounted
6287 NFS file system. When this happens and the failure is EACCES, Exim now
6288 forks a subprocess and does the per-uid checking as the relevant uid.
6290 77. Added process_log_path.
6292 78. If log_file_path was not explicitly set, a setting of check_log_space or
6293 check_log_inodes was ignored.
6295 79. If a space check for the spool or log partitions fails, the incident is now
6296 logged. Of course, in the latter case the data may get lost...
6298 80. Added the %p formatting code to string_format() so that it can be used to
6299 print addresses in debug_print(). Adjusted all the address printing in the
6300 debugging in store.c to use %p rather than %d.
6302 81. There was a concern that a line of code in smtp_in.c could overflow a
6303 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6304 initially expressed, the concern was not well-founded, because trailing
6305 spaces are removed early. However, if the trailing spaces were followed by
6306 a NULL, they did not get removed, so the overflow was possible. Two fixes
6309 (a) I re-wrote the offending code in a cleaner fashion.
6310 (b) If an incoming SMTP command contains a NULL character, it is rejected
6313 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6314 runs initgroups(), so that if the Exim user is in any additional groups,
6315 they will be used during message reception.
6321 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob