1 Change log file for Exim from version 4.21
2 -------------------------------------------
5 HS/04 Add support for keep_environment and add_environment options.
8 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
11 JH/02 The smtp transport option "multi_domain" is now expanded.
13 JH/03 The smtp transport now requests PRDR by default, if the server offers
16 JH/04 Certificate name checking on server certificates, when exim is a client,
17 is now done by default. The transport option tls_verify_cert_hostnames
18 can be used to disable this per-host. The build option
19 EXPERIMENTAL_CERTNAMES is withdrawn.
21 JH/05 The value of the tls_verify_certificates smtp transport and main options
22 default to the word "system" to access the system default CA bundle.
23 For GnuTLS, only version 3.0.20 or later.
25 JH/06 Verification of the server certificate for a TLS connection is now tried
26 (but not required) by default. The verification status is now logged by
27 default, for both outbound TLS and client-certificate supplying inbound
30 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
33 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
34 Status Notification (bounce) messages are now MIME format per RFC 3464.
35 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
36 under the control of the dsn_advertise_hosts option, and routers may
37 have a dsn_lasthop option.
39 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
40 default, modifiable by a malware= option. The list separator for
41 the options can now be changed in the usual way. Bug 68.
43 JH/10 The smtp_receive_timeout main option is now expanded before use.
45 JH/11 The incoming_interface log option now also enables logging of the
46 local interface on delivery outgoing connections.
48 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
49 if the interface and destination host and port all match.
51 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
54 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
55 Patch from Andrew Lewis.
57 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
58 now supports optional time-restrictions, weighting, and priority
59 modifiers per server. Patch originally by <rommer@active.by>.
61 JH/16 The spamd_address main option now supports a mixed list of local
62 and remote servers. Remote servers can be IPv6 addresses, and
65 JH/17 Bug 68: The spamd_address main option now supports an optional
66 timeout value per server.
68 JH/18 Bug 1581: Router and transport options headers_add/remove can
69 now have the list separator specified.
71 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
74 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
77 JH/21 Support for the A6 type of dns record is withdrawn.
79 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
80 rather than the verbs used.
82 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
83 from 255 to 1024 chars.
85 JH/24 Verification callouts now attempt to use TLS by default.
87 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
88 are generic router options now. The defaults didn't change.
90 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
91 Original patch from Alexander Shikoff, worked over by JH.
93 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
94 files if ZCAT_COMMAND is not executable.
96 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
98 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
100 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
101 Normally benign, it bites when the pair was led to by a CNAME;
102 modern usage is to not canoicalize the domain to a CNAME target
103 (and we were inconsistent anyway for A-only vs AAAA+A).
105 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
107 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
108 when evaluating $sender_host_dnssec.
110 JH/31 Check the HELO verification lookup for DNSSEC, adding new
111 $sender_helo_dnssec variable.
113 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
115 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
117 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
119 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
120 documented as working, but never had. Support all but $spam_report.
122 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
123 added for tls authenticator.
128 TL/01 When running the test suite, the README says that variables such as
129 no_msglog_check are global and can be placed anywhere in a specific
130 test's script, however it was observed that placement needed to be near
131 the beginning for it to behave that way. Changed the runtest perl
132 script to read through the entire script once to detect and set these
133 variables, reset to the beginning of the script, and then run through
134 the script parsing/test process like normal.
136 TL/02 The BSD's have an arc4random API. One of the functions to induce
137 adding randomness was arc4random_stir(), but it has been removed in
138 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
139 function when detected.
141 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
142 cause callback expansion.
144 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
145 syntax errors in an expansion can be treated as a string instead of
146 logging or causing an error, due to the internal use of bool_lax
147 instead of bool when processing it.
149 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
150 server certificates when making smtp deliveries.
152 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
154 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
156 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
158 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
159 Merged patch from Sebastian Wiedenroth.
161 JH/05 Fix results-pipe from transport process. Several recipients, combined
162 with certificate use, exposed issues where response data items split
163 over buffer boundaries were not parsed properly. This eventually
164 resulted in duplicates being sent. This issue only became common enough
165 to notice due to the introduction of conection certificate information,
166 the item size being so much larger. Found and fixed by Wolfgang Breyha.
168 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
169 size buffer was used, resulting in syntax errors when an expansion
172 JH/07 Add support for directories of certificates when compiled with a GnuTLS
173 version 3.3.6 or later.
175 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
176 is EXPERIMENTAL_EVENT, the main-configuration and transport options
177 both become "event_action", the variables become $event_name, $event_data
178 and $event_defer_errno. There is a new variable $verify_mode, usable in
179 routers, transports and related events. The tls:cert event is now also
180 raised for inbound connections, if the main configuration event_action
183 TL/06 In test suite, disable OCSP for old versions of openssl which contained
184 early OCSP support, but no stapling (appears to be less than 1.0.0).
186 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
187 server certificate names available under the smtp transport option
188 "tls_verify_cert_hostname" now do not permit multi-component wildcard
191 JH/10 Time-related extraction expansions from certificates now use the main
192 option "timezone" setting for output formatting, and are consistent
193 between OpenSSL and GnuTLS compilations. Bug 1541.
195 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
196 encoded parameter in the incoming message. Bug 1558.
198 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
199 include certificate info, eximon was claiming there were spoolfile
202 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
204 JH/14 Log delivery-related information more consistently, using the sequence
205 "H=<name> [<ip>]" wherever possible.
207 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
208 are problematic for Debian distribution, omit them from the release
211 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
213 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
216 JH/17 Fix a null-indirection in certextract expansions when a nondefault
217 output list separator was used.
222 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
223 checkers that were complaining about end of non-void function with no
226 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
227 This was a regression intruduced in 4.83 by another bugfix.
229 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
231 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
232 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
238 TF/01 Correctly close the server side of TLS when forking for delivery.
240 When a message was received over SMTP with TLS, Exim failed to clear up
241 the incoming connection properly after forking off the child process to
242 deliver the message. In some situations the subsequent outgoing
243 delivery connection happened to have the same fd number as the incoming
244 connection previously had. Exim would try to use TLS and fail, logging
245 a "Bad file descriptor" error.
247 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
248 utilities have not been installed.
250 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
251 temporary space as the ACL may create new global variables.
253 TL/01 LDAP support uses per connection or global context settings, depending
254 upon the detected version of the libraries at build time.
256 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
257 to extract and use the src ip:port in logging and expansions as if it
258 were a direct connection from the outside internet. PPv2 support was
259 updated based on HAProxy spec change in May 2014.
261 JH/02 Add ${listextract {number}{list}{success}{fail}}.
263 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
264 Properly escape header and check for NULL return.
266 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
269 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
271 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
272 characters in header names, implemented as a verify condition.
273 Contributed by Michael Fischer v. Mollard.
275 TL/05 Rename SPF condition results err_perm and err_temp to standardized
276 results permerror and temperror. Previous values are deprecated but
277 still accepted. In a future release, err_perm and err_temp will be
278 completely removed, which will be a backward incompatibility if the
279 ACL tests for either of these two old results. Patch contributed by
280 user bes-internal on the mailing list.
282 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
284 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
285 selectors, in both main and reject logs.
287 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
290 JH/07 Add malware type "sock" for talking to simple daemon.
292 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
294 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
295 routers/transports under cutthrough routing.
297 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
298 numbers. Touch up "bool" conditional to keep the same definition.
300 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
302 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
304 JH/12 Expand items in router/transport headers_add or headers_remove lists
305 individually rather than the list as a whole. Bug 1452.
307 Required for reasonable handling of multiple headers_ options when
308 they may be empty; requires that headers_remove items with embedded
309 colons must have them doubled (or the list-separator changed).
311 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
312 view the policy declared in the DMARC record. Currently, $dmarc_status
313 is a combined value of both the record presence and the result of the
316 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
318 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
319 dnslookup router and the smtp transport (applying to the forward
322 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
323 of ldap servers used for a specific lookup. Patch provided by Heiko
326 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
327 New variable $lookup_dnssec_authenticated for observability.
329 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
330 Patch submitted by Lars Timman.
332 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
334 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
335 Requires trusted mode and valid format message id, aborts otherwise.
336 Patch contributed by Heiko Schlichting.
338 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
339 certextract with support for various fields. Bug 1358.
341 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
342 is requested by default, modifiable by smtp transport option
345 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
346 operate on certificate variables to give certificate fingerprints
347 Also new ${sha256:cert_variable}.
349 JH/23 The PRDR feature is moved from being Experimental into the mainline.
351 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
352 Christian Aistleitner.
354 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
356 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
357 file. Patch from Wolfgang Breyha.
359 JH/25 Expand the coverage of the delivery $host and $host_address to
360 client authenticators run in verify callout. Bug 1476.
362 JH/26 Port service names are now accepted for tls_on_connect_ports, to
363 align with daemon_smtp_ports. Bug 72.
365 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
366 support and error reporting did not work properly.
368 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
369 and is readable. Patch from Andrew Colin Kissa.
371 TL/14 Enhance documentation of ${run expansion and how it parses the
372 commandline after expansion, particularly in the case when an
373 unquoted variable expansion results in an empty value.
375 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
377 PP/02 Fix internal collision of T_APL on systems which support RFC3123
378 by renaming away from it. Addresses GH issue 15, reported by
381 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
383 TL/15 SECURITY: prevent double expansion in math comparison functions
384 (can expand unsanitized data). Not remotely exploitable.
391 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
393 PP/02 Make -n do something, by making it not do something.
394 When combined with -bP, the name of an option is not output.
396 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
399 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
400 $sender_host_name and config options to manage this, and basic check
403 PP/05 DSCP support for outbound connections and control modifier for inbound.
405 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
406 (Only plugin which currently uses this is kerberos4, which nobody should
407 be using, but we should make it available and other future plugins might
408 conceivably use it, even though it would break NAT; stuff *should* be
409 using channel bindings instead).
411 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
412 name; added for Sendmail compatibility; requires admin caller.
413 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
414 just ignore it); requires trusted caller.
415 Also parse but ignore: -Ac -Am -X<logfile>
418 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
420 TL/02 Add +smtp_confirmation as a default logging option.
422 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
423 Patch by Magnus Holmgren from 2007-02-20.
425 TL/04 Bugzilla 1281 - Spec typo.
426 Bugzilla 1283 - Spec typo.
427 Bugzilla 1290 - Spec grammar fixes.
429 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
431 TL/06 Add Experimental DMARC support using libopendmarc libraries.
433 TL/07 Fix an out of order global option causing a segfault. Reported to dev
434 mailing list by by Dmitry Isaikin.
436 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
438 JH/02 Support "G" suffix to numbers in ${if comparisons.
440 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
442 NM/01 Bugzilla 1197 - Spec typo
443 Bugzilla 1196 - Spec examples corrections
445 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
447 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
448 gnutls_enable_pkcs11, but renamed to more accurately indicate its
451 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
452 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
454 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
455 "acl {{name}{arg}...}", and optional args on acl condition
458 JH/05 Permit multiple router/transport headers_add/remove lines.
460 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
462 JH/07 Avoid using a waiting database for a single-message-only transport.
463 Performance patch from Paul Fisher. Bugzilla 1262.
465 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
468 JH/09 Add $headers_added variable, with content from use of ACL modifier
469 add_header (but not yet added to the message). Bugzilla 199.
471 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
472 Pulled from Bugzilla 817 by Wolfgang Breyha.
474 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
476 (nb: this is the same fix as in Exim 4.80.1)
478 JH/11 Add A= logging on delivery lines, and a client_set_id option on
481 JH/12 Add optional authenticated_sender logging to A= and a log_selector
484 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
486 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
487 advertise SMTP AUTH mechanism to us, instead of a generic
488 protocol violation error. Also, make Exim more robust to bad
489 data from the Dovecot auth socket.
491 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
493 When a queue runner is handling a message, Exim first routes the
494 recipient addresses, during which it prunes them based on the retry
495 hints database. After that it attempts to deliver the message to
496 any remaining recipients. It then updates the hints database using
499 So if a recipient address works intermittently, it can get repeatedly
500 deferred at routing time. The retry hints record remains fresh so the
501 address never reaches the final cutoff time.
503 This is a fairly common occurrence when a user is bumping up against
504 their storage quota. Exim had some logic in its local delivery code
505 to deal with this. However it did not apply to per-recipient defers
506 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
508 This change adds a proper retry rule check during routing so that the
509 final cutoff time is checked against the message's age. We only do
510 this check if there is an address retry record and there is not a
511 domain retry record; this implies that previous attempts to handle
512 the address had the retry_use_local_parts option turned on. We use
513 this as an approximation for the destination being like a local
514 delivery, as in LMTP.
516 I suspect this new check makes the old local delivery cutoff check
517 redundant, but I have not verified this so I left the code in place.
519 TF/02 Correct gecos expansion when From: is a prefix of the username.
521 Test 0254 submits a message to Exim with the header
525 When I ran the test suite under the user fanf2, Exim expanded
526 the header to contain my full name, whereas it should have added
527 a Resent-Sender: header. It erroneously treats any prefix of the
528 username as equal to the username.
530 This change corrects that bug.
532 GF/01 DCC debug and logging tidyup
533 Error conditions log to paniclog rather than rejectlog.
534 Debug lines prefixed by "DCC: " to remove any ambiguity.
536 TF/03 Avoid unnecessary rebuilds of lookup-related code.
538 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
539 Bug spotted by Jeremy Harris; was flawed since initial commit.
540 Would have resulted in OCSP responses post-SNI triggering an Exim
541 NULL dereference and crash.
543 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
545 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
546 Bug detection, analysis and fix by Samuel Thibault.
547 Bugzilla 1331, Debian bug #698092.
549 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
551 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
552 Server implementation by Todd Lyons, client by JH.
553 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
554 config variable "prdr_enable" controls whether the server
555 advertises the facility. If the client requests PRDR a new
556 acl_data_smtp_prdr ACL is called once for each recipient, after
557 the body content is received and before the acl_smtp_data ACL.
558 The client is controlled by bolth of: a hosts_try_prdr option
559 on the smtp transport, and the server advertisement.
560 Default client logging of deliveries and rejections involving
561 PRDR are flagged with the string "PRDR".
563 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
564 fclose(). Diagnosis by Todd Lyons.
566 PP/17 Update configure.default to handle IPv6 localhost better.
567 Patch by Alain Williams (plus minor tweaks).
570 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
571 This is now consistent with GnuTLS, and is now documented: the
572 previous undocumented portable approach to treating the option as
573 unset was to force an expansion failure. That still works, and
574 an empty string is now equivalent.
576 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
577 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
578 not performing validation itself.
580 PP/20 Added force_command boolean option to pipe transport.
581 Patch from Nick Koston, of cPanel Inc.
583 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
586 TF/04 Added udpsend ACL modifer and hexquote expansion operator
588 PP/21 Fix eximon continuous updating with timestamped log-files.
589 Broken in a format-string cleanup in 4.80, missed when I repaired the
590 other false fix of the same issue.
591 Report and fix from Heiko Schlichting.
594 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
595 Report from Prashanth Katuri.
597 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
598 It's SecureTransport, so affects any MacOS clients which use the
599 system-integrated TLS libraries, including email clients.
601 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
602 using a MIME ACL for non-SMTP local injection.
603 Report and assistance in diagnosis by Warren Baker.
605 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
607 JH/16 Fix comparisons for 64b. Bugzilla 1385.
609 TL/09 Add expansion variable $authenticated_fail_id to keep track of
610 last id that failed so it may be referenced in subsequent ACL's.
612 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
615 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
616 ldap library initialization, allowing self-signed CA's to be
617 used. Also properly sets require_cert option later in code by
618 using NULL (global ldap config) instead of ldap handle (per
619 session). Bug diagnosis and testing by alxgomz.
621 TL/12 Enhanced documentation in the ratelimit.pl script provided in
622 the src/util/ subdirectory.
624 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
625 renamed to Transport Post Delivery Action by Jeremy Harris, as
628 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
629 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
630 redis_servers = needs to be configured which will be used by the redis
631 lookup. Patch from Warren Baker, of The Packet Hub.
633 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
635 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
636 hostname or reverse DNS when processing a host list. Used suggestions
637 from multiple comments on this bug.
639 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
641 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
642 Missed a few lines, added it to make the runtest require no keyboard
645 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
646 contains upper case chars. Make router use caseful_local_part.
648 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
649 support when GnuTLS has been built with p11-kit.
655 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
657 This, or similar/improved, will also be change PP/11 of 4.82.
663 PP/01 Handle short writes when writing local log-files.
664 In practice, only affects FreeBSD (8 onwards).
665 Bugzilla 1053, with thanks to Dmitry Isaikin.
667 NM/01 Bugzilla 949 - Documentation tweak
669 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
672 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
674 PP/02 Implemented gsasl authenticator.
676 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
678 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
679 `pkg-config foo` for cflags/libs.
681 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
682 with rest of GSASL and with heimdal_gssapi.
684 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
685 `pkg-config foo` for cflags/libs for the TLS implementation.
687 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
688 properties get this fed in as external SSF. A number of robustness
689 and debugging improvements to the cyrus_sasl authenticator.
691 PP/08 cyrus_sasl server now expands the server_realm option.
693 PP/09 Bugzilla 1214 - Log authentication information in reject log.
694 Patch by Jeremy Harris.
696 PP/10 Added dbmjz lookup type.
698 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
700 PP/12 MAIL args handles TAB as well as SP, for better interop with
701 non-compliant senders.
702 Analysis and variant patch by Todd Lyons.
704 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
705 Bug report from Lars Müller <lars@samba.org> (via SUSE),
706 Patch from Dirk Mueller <dmueller@suse.com>
708 PP/13 tls_peerdn now print-escaped for spool files.
709 Observed some $tls_peerdn in wild which contained \n, which resulted
710 in spool file corruption.
712 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
713 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
714 or write after TLS renegotiation, which otherwise led to messages
717 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
718 as a tracking header (ie: a signed header comes before the signature).
719 Patch from Wolfgang Breyha.
721 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
722 comma-sep list; embedded commas doubled.
724 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
726 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
728 Report and patch from Dmitry Banschikov.
730 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
731 Removed SSL_clear() after SSL_new() which led to protocol negotiation
732 failures. We appear to now support TLS1.1+ with Exim.
734 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
735 lets Exim select keys and certificates based upon TLS SNI from client.
736 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
737 before an outbound SMTP session. New log_selector, +tls_sni.
739 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
740 NULL dereference. Report and patch from Alun Jones.
742 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
743 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
744 Not seeing resolver debug output on NetBSD, but suspect this is a
745 resolver implementation change.
747 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
748 Left warnings. Added "eximon gdb" invocation mode.
750 PP/21 Defaulting "accept_8bitmime" to true, not false.
752 PP/22 Added -bw for inetd wait mode support.
754 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
755 locate the relevant includes and libraries. Made this the default.
757 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
758 Bugzilla 1246, report and most of solution from Tomasz Kusy.
760 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
761 This may cause build issues on older platforms.
763 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
764 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
765 gnutls_require_mac & gnutls_require_protocols (no longer supported).
766 Added SNI support via GnuTLS too.
767 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
769 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
771 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
772 Applied second patch from Janne, implementing suggestion to default
773 multiple-strings-in-record handling to match SPF spec.
775 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
777 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
778 read-only, out of scope).
779 Patch from Wolfgang Breyha, report from Stuart Northfield.
781 PP/29 Fix three issues highlighted by clang analyser static analysis.
782 Only crash-plausible issue would require the Cambridge-specific
783 iplookup router and a misconfiguration.
784 Report from Marcin Mirosław.
786 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
788 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
789 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
790 As part of this, removing so much warning spew let me fix some minor
791 real issues in debug logging.
793 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
794 assignment on my part. Fixed.
796 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
797 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
798 Janne Snabb (who went above and beyond: thank you).
800 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
801 string otherwise requires a connection and a bunch more work and it's
802 relatively easy to get wrong. Should also expose TLS library linkage
805 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
806 64-bit ${eval} (JH/03).
808 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
809 GNU libc to support some of the 64-bit stuff, should not lead to
810 conflicts. Defined before os.h is pulled in, so if a given platform
811 needs to override this, it can.
813 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
814 protection layer was required, which is not implemented.
815 Bugzilla 1254, patch from Wolfgang Breyha.
817 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
818 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
819 tls_dhparam take prime identifiers. Also unbreak combination of
820 OpenSSL+DH_params+TLSSNI.
822 PP/39 Disable SSLv2 by default in OpenSSL support.
828 PP/01 Solaris build fix for Oracle's LDAP libraries.
829 Bugzilla 1109, patch from Stephen Usher.
831 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
833 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
836 TF/02 Fix a couple more cases where we did not log the error message
837 when unlink() failed. See also change 4.74-TF/03.
839 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
840 lock up or crash if it happened to be inside a call to libc when it
841 got a SIGUSR1 from exiwhat.
843 The SIGUSR1 handler appends the current process status to the process
844 log which is later printed by exiwhat. It used to use the general
845 purpose logging code to do this, but several functions it calls are
846 not safe for signals.
848 The new output code in the SIGUSR1 handler is specific to the process
849 log, and simple enough that it's easy to inspect for signal safety.
850 Removing some special cases also simplifies the general logging code.
851 Removing the spurious timestamps from the process log simplifies
854 TF/04 Improved ratelimit ACL condition.
856 The /noupdate option has been deprecated in favour of /readonly which
857 has clearer semantics. The /leaky, /strict, and /readonly update modes
858 are mutually exclusive. The update mode is no longer included in the
859 database key; it just determines when the database is updated. (This
860 means that when you upgrde Exim will forget old rate measurements.)
862 Exim now checks that the per_* options are used with an update mode that
863 makes sense for the current ACL. For example, when Exim is processing a
864 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
865 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
866 must specify per_mail/readonly. If you omit the update mode it defaults to
867 /leaky where that makes sense (as before) or /readonly where required.
869 The /noupdate option is now undocumented but still supported for
870 backwards compatibility. It is equivalent to /readonly except that in
871 ACLs where /readonly is required you may specify /leaky/noupdate or
872 /strict/noupdate which are treated the same as /readonly.
874 A useful new feature is the /count= option. This is a generalization
875 of the per_byte option, so that you can measure the throughput of other
876 aggregate values. For example, the per_byte option is now equivalent
877 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
879 The per_rcpt option has been generalized using the /count= mechanism
880 (though it's more complicated than the per_byte equivalence). When it is
881 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
882 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
883 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
884 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
885 Note that using per_rcpt with a non-readonly update mode in more than
886 one ACL will cause the recipients to be double-counted. (The per_mail
887 and per_byte options don't have this problem.)
889 The handling of very low rates has changed slightly. If the computed rate
890 is less than the event's count (usually one) then this event is the first
891 after a long gap. In this case the rate is set to the same as this event's
892 count, so that the first message of a spam run is counted properly.
894 The major new feature is a mechanism for counting the rate of unique
895 events. The new per_addr option counts the number of different
896 recipients that someone has sent messages to in the last time period. It
897 behaves like per_rcpt if all the recipient addresses are different, but
898 duplicate recipient addresses do not increase the measured rate. Like
899 the /count= option this is a general mechanism, so the per_addr option
900 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
901 example, measure the rate that a client uses different sender addresses
902 with the options per_mail/unique=$sender_address. There are further
903 details in the main documentation.
905 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
907 TF/06 Removed a few PCRE remnants.
909 TF/07 Automatically extract Exim's version number from tags in the git
910 repository when doing development or release builds.
912 PP/02 Raise smtp_cmd_buffer_size to 16kB.
913 Bugzilla 879. Patch from Paul Fisher.
915 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
916 Heavily based on revision 40f9a89a from Simon Arlott's tree.
919 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
921 PP/05 Variable $av_failed, true if the AV scanner deferred.
922 Bugzilla 1078. Patch from John Horne.
924 PP/06 Stop make process more reliably on build failure.
925 Bugzilla 1087. Patch from Heiko Schlittermann.
927 PP/07 Make maildir_use_size_file an _expandable_ boolean.
928 Bugzilla 1089. Patch from Heiko Schlittermann.
930 PP/08 Handle ${run} returning more data than OS pipe buffer size.
931 Bugzilla 1131. Patch from Holger Weiß.
933 PP/09 Handle IPv6 addresses with SPF.
934 Bugzilla 860. Patch from Wolfgang Breyha.
936 PP/10 GnuTLS: support TLS 1.2 & 1.1.
938 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
941 PP/11 match_* no longer expand right-hand-side by default.
942 New compile-time build option, EXPAND_LISTMATCH_RHS.
943 New expansion conditions, "inlist", "inlisti".
945 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
947 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
949 PP/14 fix log_write() format string regression from TF/03.
950 Bugzilla 1152. Patch from Dmitry Isaikin.
956 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
958 PP/02 Harmonised TLS library version reporting; only show if debugging.
959 Layout now matches that introduced for other libraries in 4.74 PP/03.
961 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
963 PP/04 New "dns_use_edns0" global option.
965 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
968 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
969 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
971 TK/01 Updated PolarSSL code to 0.14.2.
972 Bugzilla 1097. Patch from Andreas Metzler.
974 PP/07 Catch divide-by-zero in ${eval:...}.
977 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
980 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
981 format-string attack -- SECURITY: remote arbitrary code execution.
983 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
984 time unintentionally subject to list matching rules, letting the header
985 cause arbitrary Exim lookups (of items which can occur in lists, *not*
986 arbitrary string expansion). This allowed for information disclosure.
988 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
989 INT_MIN/-1 -- value coerced to INT_MAX.
995 NM/01 Workround for PCRE version dependency in version reporting
998 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
999 This fixes portability to compilers other than gcc, notably
1000 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1002 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1003 makefiles for portability to HP-UX and POSIX correctness.
1005 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1006 Also via indented variable definition in the Makefile.
1007 (Debugging by Oliver Heesakkers).
1009 PP/02 Restore caching of spamd results with expanded spamd_address.
1010 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1012 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1013 Improves build reliability. Fix from: Frank Elsner
1015 NM/02 Fix wide character breakage in the rfc2047 coding
1016 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1018 NM/03 Allow underscore in dnslist lookups
1019 Fixes bug 1026. Patch from Graeme Fowler
1021 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1022 Code patches from Adam Ciarcinski of NetBSD.
1024 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1027 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1028 is logged, to avoid truncation. Patch from John Horne.
1030 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1031 Patch from Jakob Hirsch.
1033 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1034 SQL string expansion failure details.
1035 Patch from Andrey Oktyabrski.
1037 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1038 Patch from Simon Arlott.
1040 PP/09 New lookups functionality failed to compile on old gcc which rejects
1041 extern declarations in function scope.
1042 Patch from Oliver Fleischmann
1044 PP/10 Use sig_atomic_t for flags set from signal handlers.
1045 Check getgroups() return and improve debugging.
1046 Fixed developed for diagnosis in bug 927 (which turned out to be
1049 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1050 Patch from Mark Zealey.
1052 PP/12 Bugzilla 1056: Improved spamd server selection.
1053 Patch from Mark Zealey.
1055 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1056 Based on patch from Heiko Schlittermann.
1058 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1059 Patch from Uwe Doering, sign-off by Michael Haardt.
1061 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1062 variable declaration deep within a block. Bug and patch from
1065 PP/15 lookups-Makefile IRIX compatibilty coercion.
1067 PP/16 Make DISABLE_DKIM build knob functional.
1069 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1070 Patch by Simon Arlott
1072 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1073 variable argument macros. Our copy now differs from upstream.
1079 TF/01 Failure to get a lock on a hints database can have serious
1080 consequences so log it to the panic log.
1082 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1083 controlled using the smtp_confirmation log selector.
1085 TF/03 Include the error message when we fail to unlink a spool file.
1087 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1088 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1089 for maintaining out-of-tree patches for some time.
1091 PP/01 Bugzilla 139: Documentation and portability issues.
1092 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1093 Handle per-OS dynamic-module compilation flags.
1095 PP/02 Let /dev/null have normal permissions.
1096 The 4.73 fixes were a little too stringent and complained about the
1097 permissions on /dev/null. Exempt it from some checks.
1098 Reported by Andreas M. Kirchwitz.
1100 PP/03 Report version information for many libraries, including
1101 Exim version information for dynamically loaded libraries. Created
1102 version.h, now support a version extension string for distributors
1103 who patch heavily. Dynamic module ABI change.
1105 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1106 privilege escalation vulnerability whereby the Exim run-time user
1107 can cause root to append content of the attacker's choosing to
1110 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1113 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1114 If dropping privileges for untrusted macros, we disabled normal logging
1115 on the basis that it would fail; for the Exim run-time user, this is not
1116 the case, and it resulted in successful deliveries going unlogged.
1117 Fixed. Reported by Andreas Metzler.
1123 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1124 only prepend for the Resent-* case. Fixes regression introduced in
1125 Exim 4.70 by NM/22 for Bugzilla 607.
1127 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1128 increasing numbers of administrators be bitten by this.
1130 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1132 PP/03 Bugzilla 994: added openssl_options main configuration option.
1134 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1136 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1138 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1140 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1141 without a peer certificate, leading to a segfault because of an
1142 assumption that peers always have certificates. Be a little more
1143 paranoid. Problem reported by Martin Tscholak.
1145 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1146 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1147 NB: ClamAV planning to remove STREAM in "middle of 2010".
1148 CL also introduces -bmalware, various -d+acl logging additions and
1149 more caution in buffer sizes.
1151 PP/09 Implemented reverse_ip expansion operator.
1153 PP/10 Bugzilla 937: provide a "debug" ACL control.
1155 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1157 PP/12 Bugzilla 973: Implement --version.
1159 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1161 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1163 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1165 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1166 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1167 ignore trailing whitespace.
1169 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1171 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1174 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1175 Notification from Dr Andrew Aitchison.
1177 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1178 ExtendedDetectionInfo response format.
1179 Notification from John Horne.
1181 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1184 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1185 XSL and documented dependency on system catalogs, with examples of how
1188 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1191 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1192 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1193 configuration file which is writeable by the Exim user or group.
1195 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1196 of configuration files to cover files specified with the -C option if
1197 they are going to be used with root privileges, not just the default
1200 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1201 option (effectively making it always true).
1203 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1204 files to be used while preserving root privileges.
1206 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1207 that rogue child processes cannot use them.
1209 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1210 run-time user, instead of root.
1212 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1213 Exim run-time user without dropping privileges.
1215 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1216 result string, instead of calling string_vformat() twice with the same
1219 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1220 for other users. Others should always drop root privileges if they use
1221 -C on the command line, even for a whitelisted configure file.
1223 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1225 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1231 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1232 $data_path, and $header_path variables; fixed documentation bugs and
1235 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1236 exipick to access non-standard spools, including the "frozen" queue
1239 NM/01 Bugzilla 965: Support mysql stored procedures.
1240 Patch from Alain Williams
1242 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1244 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1245 Patch from Andreas Metzler
1247 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1248 Patch from Kirill Miazine
1250 NM/05 Bugzilla 671: Added umask to procmail example.
1252 JJ/03 installed exipick 20100323.0, fixing doc bug
1254 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1255 directory. Notification and patch from Dan Rosenberg.
1257 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1259 TK/02 Improve log output when DKIM signing operation fails.
1261 MH/01 Treat the transport option dkim_domain as a colon separated
1262 list, not as a single string, and sign the message with each element,
1263 omitting multiple occurences of the same signer.
1265 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1266 Bugzilla 985, 986. Patch by Simon Arlott
1268 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1269 Patch by Simon Arlott
1271 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1272 MBX locking. Notification from Dan Rosenberg.
1278 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1280 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1282 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1284 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1286 NM/04 Bugzilla 915: Flag broken perl installation during build.
1292 TK/01 Added patch by Johannes Berg that expands the main option
1293 "spamd_address" if it starts with a dollar sign.
1295 TK/02 Write list of recipients to X-Envelope-Sender header when building
1296 the mbox-format spool file for content scanning (suggested by Jakob
1299 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1300 (http://www.dcc-servers.net/) support via dccifd. Activated by
1301 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1303 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1304 by Mark Daniel Reidel <mr@df.eu>.
1306 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1307 When building exim an external PCRE library is now needed -
1308 PCRE is a system library on the majority of modern systems.
1309 See entry on PCRE_LIBS in EDITME file.
1311 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1312 conversation. Added nologin parameter to request.
1313 Patch contributed by Kirill Miazine.
1315 TF/01 Do not log submission mode rewrites if they do not change the address.
1317 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1319 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1320 log files in place. Contributed by Roberto Lima.
1322 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1324 TF/03 Bugzilla 615: When checking the local_parts router precondition
1325 after a local_part_suffix or local_part_prefix option, Exim now
1326 does not use the address's named list lookup cache, since this
1327 contains cached lookups for the whole local part.
1329 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1330 Robert Millan. Documentation is in experimental-spec.txt.
1332 TF/04 Bugzilla 668: Fix parallel build (make -j).
1334 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1336 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1337 Patch provided by Jan Srzednicki.
1339 TF/05 Leading white space used to be stripped from $spam_report which
1340 wrecked the formatting. Now it is preserved.
1342 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1343 that they are available at delivery time.
1345 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1347 TF/08 TLS error reporting now respects the incoming_interface and
1348 incoming_port log selectors.
1350 TF/09 Produce a more useful error message if an SMTP transport's hosts
1351 setting expands to an empty string.
1353 NM/06 Bugzilla 744: EXPN did not work under TLS.
1354 Patch provided by Phil Pennock.
1356 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1357 Patch provided by Richard Godbee.
1359 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1360 acl_smtp_notquit, added index entry.
1362 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1363 Patch provided by Eugene Bujak.
1365 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1366 accept(). Patch provided by Maxim Dounin.
1368 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1369 Patch provided by Phil Pennock.
1371 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1373 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1374 Patch provided by Brad "anomie" Jorsch.
1376 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1377 Patch provided by Dean Brooks.
1379 TK/05 Add native DKIM support (does not depend on external libraries).
1381 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1382 Patch provided by Graeme Fowler.
1384 NM/16 Bugzilla 851: Documentation example syntax fix.
1386 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1388 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1391 NM/19 Bugzilla 745: TLS version reporting.
1392 Patch provided by Phil Pennock.
1394 NM/20 Bugzilla 167: bool: condition support.
1395 Patch provided by Phil Pennock.
1397 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1398 clients. Patch provided by Phil Pennock.
1400 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1401 Patch provided by Brad "anomie" Jorsch.
1403 NM/23 Bugzilla 687: Fix misparses in eximstats.
1404 Patch provided by Heiko Schlittermann.
1406 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1407 Patch provided by Heiko Schlittermann.
1409 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1410 plus update to original patch.
1412 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1414 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1415 Patch provided by David Brownlee.
1417 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1419 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1421 NM/30 Bugzilla 888: TLS documentation bugfixes.
1423 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1425 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1426 Unlike the original bugzilla I have changed all shell scripts in src tree.
1428 NM/33 Bugzilla 898: Transport filter timeout fix.
1429 Patch by Todd Rinaldo.
1431 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1432 Patch by Serge Demonchaux.
1434 NM/35 Bugzilla 39: Base64 decode bug fixes.
1435 Patch by Jakob Hirsch.
1437 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1439 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1441 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1443 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1449 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1450 ALT-N's libdkim that I have put here:
1451 http://duncanthrax.net/exim-experimental/
1453 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1454 were called 'true' and it seems that C99 defines that as a reserved
1455 keyword to be used with 'bool' variable types. That means you could
1456 not include C99-style headers which use bools without triggering
1457 build errors in sieve.c.
1459 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1460 as mailq or other aliases. Changed the --help handling significantly
1461 to do whats expected. exim_usage() emits usage/help information.
1463 SC/01 Added the -bylocaldomain option to eximstats.
1465 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1467 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1469 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1475 PH/01 Another patch from the Sieve maintainer.
1477 PH/02 When an IPv6 address is converted to a string for single-key lookup
1478 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1479 dots are used instead of colons so that keys in lsearch files need not
1480 contain colons. This was done some time before quoting was made available
1481 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1482 (notated using the quote facility) so as to distinguish them from IPv4
1483 keys. This meant that lookups for IP addresses in host lists did not work
1484 for iplsearch lookups.
1486 This has been fixed by arranging for IPv6 addresses to be expressed with
1487 colons if the lookup type is iplsearch. This is not incompatible, because
1488 previously such lookups could never work.
1490 The situation is now rather anomolous, since one *can* have colons in
1491 ordinary lsearch keys. However, making the change in all cases is
1492 incompatible and would probably break a number of configurations.
1494 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1497 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1498 conversion specification without a maximum field width, thereby enabling
1499 a rogue spamd server to cause a buffer overflow. While nobody in their
1500 right mind would setup Exim to query an untrusted spamd server, an
1501 attacker that gains access to a server running spamd could potentially
1502 exploit this vulnerability to run arbitrary code as the Exim user.
1504 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1505 $primary_hostname instead of what libspf2 thinks the hosts name is.
1507 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1508 a directory entry by the name of the lookup key. Previously, if a
1509 symlink pointed to a non-existing file or a file in a directory that
1510 Exim lacked permissions to read, a lookup for a key matching that
1511 symlink would fail. Now it is enough that a matching directory entry
1512 exists, symlink or not. (Bugzilla 503.)
1514 PH/03 The body_linecount and body_zerocount variables are now exported in the
1517 PH/04 Added the $dnslist_matched variable.
1519 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1520 This means they are set thereafter only if the connection becomes
1523 PH/06 Added the client_condition to authenticators so that some can be skipped
1524 by clients under certain conditions.
1526 PH/07 The error message for a badly-placed control=no_multiline_responses left
1527 "_responses" off the end of the name.
1529 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1531 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1532 (without spaces) instead of just copying the configuration text.
1534 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1536 PH/11 Added $max_received_linelength.
1538 PH/12 Added +ignore_defer and +include_defer to host lists.
1540 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1541 way in which PCRE > 7.0 is built.
1543 PH/14 Implemented queue_only_load_latch.
1545 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1546 MAIL command. The effect was to mangle the value on 64-bit systems.
1548 PH/16 Another patch from the Sieve maintainer.
1550 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1552 PH/18 If a system quota error occurred while trying to create the file for
1553 a maildir delivery, the message "Mailbox is full" was not appended to the
1554 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1555 only to a quota excession during the actual writing of the file.
1557 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1558 characters?) which causes problems in log lines. The DN values are now
1559 passed through string_printing() before being added to log lines.
1561 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1562 and InterBase are left for another time.)
1564 PH/21 Added message_body_newlines option.
1566 PH/22 Guard against possible overflow in moan_check_errorcopy().
1568 PH/23 POSIX allows open() to be a macro; guard against that.
1570 PH/24 If the recipient of an error message contained an @ in the local part
1571 (suitably quoted, of course), incorrect values were put in $domain and
1572 $local_part during the evaluation of errors_copy.
1578 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1579 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1582 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1583 issue a MAIL command.
1585 PH/02 In an ACL statement such as
1587 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1589 if a client was not listed at all, or was listed with a value other than
1590 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1591 the condition was not true (as it should be), so access was not denied.
1592 The bug was that the ! inversion was incorrectly passed on to the second
1593 item. This has been fixed.
1595 PH/03 Added additional dnslists conditions == and =& which are different from
1596 = and & when the dns lookup returns more than one IP address.
1598 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1599 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1601 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1602 FSYNC, which compiles an option called disable_fsync that allows for
1603 bypassing fsync(). The documentation is heavily laced with warnings.
1605 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1607 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1608 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1609 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1610 including adding "make clean"; (3) Added -fPIC when compiling the test
1611 dynamically loaded module, to get rid of a warning.
1613 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1614 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1615 = 0s. The bug is otherwise harmless.
1617 PH/07 There was a bug in the dovecot authenticator such that the value of
1618 $auth1 could be overwritten, and so not correctly preserved, after a
1619 successful authentication. This usually meant that the value preserved by
1620 the server_setid option was incorrect.
1622 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1624 PH/09 Installed PCRE release 7.0.
1626 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1627 run for batched SMTP input. It is now run at the start of every message
1628 in the batch. While fixing this I discovered that the process information
1629 (output by running exiwhat) was not always getting set for -bs and -bS
1630 input. This is fixed, and it now also says "batched" for BSMTP.
1632 PH/11 Added control=no_pipelining.
1634 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1635 patch, slightly modified), and move the expansion of helo_data till after
1636 the connection is made in the smtp transport (so it can use these
1639 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1641 PH/14 Added log_selector = +pid.
1643 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1645 PH/16 Add ${if forany and ${if forall.
1647 PH/17 Added dsn_from option to vary the From: line in DSNs.
1649 PH/18 Flush SMTP output before performing a callout, unless control =
1650 no_callout_flush is set.
1652 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1653 was true (the default) a successful delivery failed to delete the retry
1654 item, thus causing premature timeout of the address. The bug is now
1657 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1659 PH/21 Long custom messages for fakedefer and fakereject are now split up
1660 into multiline reponses in the same way that messages for "deny" and
1661 other ACL rejections are.
1663 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1664 with slight modification.
1666 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1667 draft, changing the syntax and factoring some duplicate code".
1669 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1670 for deliveries of the second and subsequent messages over the same SMTP
1673 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1674 ${reduce, with only minor "tidies".
1676 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1678 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1679 expansion side effects.
1681 PH/27 When a message times out after an over-quota error from an Exim-imposed
1682 quota, the bounce message says "mailbox is full". This message was not
1683 being given when it was a system quota that was exceeded. It now should
1686 MH/03 Made $recipients available in local_scan(). local_scan() already has
1687 better access to the recipient list through recipients_list[], but
1688 $recipients can be useful in postmaster-provided expansion strings.
1690 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1691 in the case of a MAIL command with additional options following the
1692 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1693 were accidentally chopped off.
1695 PH/29 SMTP synchronization checks are implemented when a command is read -
1696 there is a check that no more input is waiting when there shouldn't be
1697 any. However, for some commands, a delay in an ACL can mean that it is
1698 some time before the response is written. In this time, more input might
1699 arrive, invalidly. So now there are extra checks after an ACL has run for
1700 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1701 pipelining has not been advertised.
1703 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1705 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1706 returned in the SMTP response when smtp_return_error_details was false.
1707 This has been fixed.
1709 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1710 instead of the C I/O that was originally supplied, because problems were
1711 reported on Solaris.
1713 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1714 Exim which did not show up earlier: it was assuming that a call to
1715 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1716 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1717 was a macro that became an assignment, so it seemed to work. This has
1718 changed to a proper function call with a void return, hence the compile
1719 error. Exim's code has been fixed.
1721 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1724 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1726 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1728 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1729 because it turns out to be too big or there is a timeout) there is no
1730 "Completed" line in the log. When some messages of this type were
1731 selected by exigrep, they were listed as "not completed". Others were
1732 picked up by some special patterns. I have improved the selection
1733 criteria to be more general.
1735 PH/38 The host_find_failed option in the manualroute router can now be set
1736 to "ignore", to completely ignore a host whose IP address cannot be
1737 found. If all hosts are ignored, the behaviour is controlled by the new
1738 host_all_ignored option.
1740 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1741 homing or because of multiple MX records with /mx) generated more than
1742 one IP address, and the following item turned out to be the local host,
1743 all the secondary addresses of the first item were incorrectly removed
1744 from the list, along with the local host and any following hosts (which
1745 is what is supposed to happen).
1747 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1748 whoever called Exim into the -H file. In the case of the daemon it was
1749 behaving confusingly. When first started, it used values for whoever
1750 started the daemon, but after a SIGHUP it used the Exim user (because it
1751 calls itself on a restart). I have changed the code so that it now always
1754 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1755 message are rejected with the same error (e.g. no authentication or bad
1756 sender address), and a DATA command is nevertheless sent (as can happen
1757 with PIPELINING or a stupid MUA), the error message that was given to the
1758 RCPT commands is included in the rejection of the DATA command. This is
1759 intended to be helpful for MUAs that show only the final error to their
1762 PH/42 Another patch from the Sieve maintainer.
1764 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1765 Eximstats - Fixed some broken HTML links and added missing column headers
1767 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1768 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1770 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1772 PH/43 Yet another patch from the Sieve maintainer.
1774 PH/44 I found a way to check for a TCP/IP connection going away before sending
1775 the response to the final '.' that terminates a message, but only in the
1776 case where the client has not sent further data following the '.'
1777 (unfortunately, this is allowed). However, in many cases there won't be
1778 any further data because there won't be any more messages to send. A call
1779 to select() can be used: if it shows that the input is "ready", there is
1780 either input waiting, or the socket has been closed. An attempt to read
1781 the next input character can distinguish the two cases. Previously, Exim
1782 would have sent an OK response which the client would never have see.
1783 This could lead to message repetition. This fix should cure that, at
1784 least in a lot of common cases.
1786 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1787 advertised in response to EHLO.
1793 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1794 fixed by 4.65/MH/01 (is this a record?) are fixed:
1796 (i) An empty string was always treated as zero by the numeric comparison
1797 operators. This behaviour has been restored.
1799 (ii) It is documented that the numeric comparison operators always treat
1800 their arguments as decimal numbers. This was broken in that numbers
1801 starting with 0 were being interpreted as octal.
1803 While fixing these problems I realized that there was another issue that
1804 hadn't been noticed. Values of message_size_limit (both the global option
1805 and the transport option) were treated as octal if they started with 0.
1806 The documentation was vague. These values are now always treated as
1807 decimal, and I will make that clear in the documentation.
1813 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1814 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1817 MH/01 Don't check that the operands of numeric comparison operators are
1818 integers when their expansion is in "skipping" mode (fixes bug
1819 introduced by 4.64-PH/07).
1821 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1822 child addresses, Exim now panics and dies. Previously, because the count
1823 is held in a short int, deliveries were likely to be lost. As such a
1824 large number of recipients for a single message is ridiculous
1825 (performance will be very, very poor), I have chosen to impose a limit
1826 rather than extend the field.
1832 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1833 leftover -K file (the existence of which was triggered by #402).
1834 While we were at it, introduced process PID as part of the -K
1835 filename. This should rule out race conditions when creating
1838 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1839 processing considerably. Previous code took too long for large mails,
1840 triggering a timeout which in turn triggers #401.
1842 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1843 in the DK code in transports.c. sendfile() is not really portable,
1844 hence the _LINUX specificness.
1846 TF/01 In the add_headers option to the mail command in an Exim filter,
1847 there was a bug that Exim would claim a syntax error in any
1848 header after the first one which had an odd number of characters
1851 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1852 callout verification, Exim cached a "reject" for the entire domain. This
1853 is correct for most verifications, but it is not correct for a recipient
1854 verification with use_sender or use_postmaster set, because in that case
1855 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1856 case of MAIL FROM:<> rejection from other early rejections (e.g.
1857 rejection of HELO). When verifying a recipient using a non-null MAIL
1858 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1859 Whatever the result of the callout, the value of the domain cache is
1860 left unchanged (for any other kind of callout, getting as far as trying
1861 RCPT means that the domain itself is ok).
1863 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1866 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1867 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1868 session, the daemon ignores ECONNECT errors and logs others; it now
1869 ignores EPIPE as well.
1871 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1872 (quoted-printable decoding).
1874 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1875 later the small subsequent patch to fix an introduced bug.
1877 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1879 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1881 PH/08 An error is now given if message_size_limit is specified negative.
1883 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1884 to be given (somewhat) arbitrary names.
1886 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1889 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1890 miscellaneous code fixes
1892 PH/10 Added the log_reject_target ACL modifier to specify where to log
1895 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1896 hostname. This is wrong, because it relates to the incoming message (and
1897 probably the interface on which it is arriving) and not to the outgoing
1898 callout (which could be using a different interface). This has been
1899 changed to use the value of the helo_data option from the smtp transport
1900 instead - this is what is used when a message is actually being sent. If
1901 there is no remote transport (possible with a router that sets up host
1902 addresses), $smtp_active_hostname is used.
1904 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1905 tweaks were necessary in order to get it to work (see also 21 below):
1906 (a) The code assumed that strncpy() returns a negative number on buffer
1907 overflow, which isn't the case. Replaced with Exim's string_format()
1909 (b) There were several signed/unsigned issues. I just did the minimum
1910 hacking in of casts. There is scope for a larger refactoring.
1911 (c) The code used strcasecmp() which is not a standard C function.
1912 Replaced with Exim's strcmpic() function.
1913 (d) The code set only $1; it now sets $auth1 as well.
1914 (e) A simple test gave the error "authentication client didn't specify
1915 service in request". It would seem that Dovecot has changed its
1916 interface. Fortunately there's a specification; I followed it and
1917 changed what the client sends and it appears to be working now.
1919 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1922 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1923 address A is aliased to B and C, where B exists and C does not. Without
1924 -v the output is "A verified" because verification stops after a
1925 successful redirection if more than one address is generated. However,
1926 with -v the child addresses are also verified. Exim was outputting "A
1927 failed to verify" and then showing the successful verification for C,
1928 with its parentage. It now outputs "B failed to verify", showing B's
1929 parentage before showing the successful verification of C.
1931 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1932 look up a TXT record in a specific list after matching in a combined
1935 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1936 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1937 they consult the DNS. I had assumed they would set it the way they
1938 wanted; and indeed my experiments on Linux seem to show that in some
1939 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1940 To be on the safe side, however, I have now made the interface to
1941 host_find_byname() similar to host_find_bydns(), with an argument
1942 containing the DNS resolver options. The host_find_byname() function now
1943 sets these options at its start, just as host_find_bydns() does. The smtp
1944 transport options dns_qualify_single and dns_search_parents are passed to
1945 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1946 of host_find_byname() use the default settings of RES_DEFNAMES
1947 (qualify_single) but not RES_DNSRCH (search_parents).
1949 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1950 spool_read_header() do less string testing, by means of a preliminary
1951 switch on the second character of optional "-foo" lines. (This is
1952 overdue, caused by the large number of possibilities that now exist.
1953 Originally there were few.) While I was there, I also converted the
1954 str(n)cmp tests so they don't re-test the leading "-" and the first
1955 character, in the hope this might squeeze out yet more improvement.
1957 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1958 flag allowing group syntax was set by the header_syntax check but not
1959 turned off, possible causing trouble later; (2) The flag was not being
1960 set at all for the header_verify test, causing "group"-style headers to
1961 be rejected. I have now set it in this case, and also caused header_
1962 verify to ignore an empty address taken from a group. While doing this, I
1963 came across some other cases where the code for allowing group syntax
1964 while scanning a header line wasn't quite right (mostly, not resetting
1965 the flag correctly in the right place). These bugs could have caused
1966 trouble for malformed header lines. I hope it is now all correct.
1968 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1969 with the "reply" argument non-NULL. The code, however (which originally
1970 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1971 but it didn't always do it. This confused somebody who was copying the
1972 code for some other use. I have removed all the tests.
1974 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1975 feature that was used to support insecure browsers during the U.S. crypto
1976 embargo. It requires special client support, and Exim is probably the
1977 only MTA that supported it -- and would never use it because real RSA is
1978 always available. This code has been removed, because it had the bad
1979 effect of slowing Exim down by computing (never used) parameters for the
1980 RSA_EXPORT functionality.
1982 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1983 authenticator to fail if there's a tab character in the incoming data
1984 (there should never be unless someone is messing about, as it's supposed
1985 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1986 if the connection is using TLS or if the remote IP is the same as the
1987 local IP, and the "valid-client-cert option" if a client certificate has
1990 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1991 authenticators. This can be used for authorization after authentication
1992 succeeds. (In the case of plaintext, it servers for both authentication
1995 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1996 if any retry times were supplied.
1998 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1999 connection, where there is no client IP address to check. In this
2000 situation, the verify now always succeeds.
2002 PH/25 Applied John Jetmore's -Mset patch.
2004 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2006 PH/27 In a string expansion for a processed (not raw) header when multiple
2007 headers of the same name were present, leading whitespace was being
2008 removed from all of them, but trailing whitespace was being removed only
2009 from the last one. Now trailing whitespace is removed from each header
2010 before concatenation. Completely empty headers in a concatenation (as
2011 before) are ignored.
2013 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2014 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2016 PH/29 [Removed. This was a change that I later backed out, and forgot to
2017 correct the ChangeLog entry (that I had efficiently created) before
2018 committing the later change.]
2020 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2021 address errors (4xx response to RCPT) over the same connection as other
2022 messages routed to the same hosts. Such deliveries are always "forced",
2023 so retry times are not inspected. This resulted in far too many retries
2024 for the affected addresses. The effect occurred only when there were more
2025 hosts than the hosts_max_try setting in the smtp transport when it had
2026 the 4xx errors. Those hosts that it had tried were not added to the list
2027 of hosts for which the message was waiting, so if all were tried, there
2028 was no problem. Two fixes have been applied:
2030 (i) If there are any address or message errors in an SMTP delivery, none
2031 of the hosts (tried or untried) are now added to the list of hosts
2032 for which the message is waiting, so the message should not be a
2033 candidate for sending over the same connection that was used for a
2034 successful delivery of some other message. This seems entirely
2035 reasonable: after all the message is NOT "waiting for some host".
2036 This is so "obvious" that I'm not sure why it wasn't done
2037 previously. Hope I haven't missed anything, but it can't do any
2038 harm, as the worst effect is to miss an optimization.
2040 (ii) If, despite (i), such a delivery is accidentally attempted, the
2041 routing retry time is respected, so at least it doesn't keep
2042 hammering the server.
2044 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2045 in ${readsocket because some servers need this prod.
2047 PH/32 Added some extra debug output when updating a wait-xxx database.
2049 PH/33 The hint "could be header name not terminated by colon", which has been
2050 given for certain expansion errors for a long time, was not being given
2051 for the ${if def:h_colon_omitted{... case.
2053 PH/34 The spec says: "With one important exception, whenever a domain list is
2054 being scanned, $domain contains the subject domain." There was at least
2055 one case where this was not true.
2057 PH/35 The error "getsockname() failed: connection reset by peer" was being
2058 written to the panic log as well as the main log, but it isn't really
2059 panic-worthy as it just means the connection died rather early on. I have
2060 removed the panic log writing for the ECONNRESET error when getsockname()
2063 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2064 runs only) independently of the message's sender address. This meant
2065 that, if the 4xx error was in fact related to the sender, a different
2066 message to the same recipient with a different sender could confuse
2067 things. In particualar, this can happen when sending to a greylisting
2068 server, but other circumstances could also provoke similar problems.
2069 I have changed the default so that the retry time for these errors is now
2070 based a combination of the sender and recipient addresses. This change
2071 can be overridden by setting address_retry_include_sender=false in the
2074 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2075 remote server are returned as part of bounce messages. This was not
2076 happening for LMTP over a pipe (the lmtp transport), but now it is the
2077 same for both kinds of LMTP.
2079 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2080 in header lines that were in fact CNAMEs. This is no longer the case.
2082 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2083 and queue runs started by the daemon processed all messages. This has
2084 been fixed so that -R and -S can now usefully be given with -q<time>.
2086 PH/40 Import PCRE release 6.7 (fixes some bugs).
2088 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2090 PH/42 Give an error if -q is specified more than once.
2092 PH/43 Renamed the variables $interface_address and $interface_port as
2093 $received_ip_address and $received_port, to make it clear that these
2094 values apply to message reception, and not to the outgoing interface when
2095 a message is delivered. (The old names remain recognized, of course.)
2097 PH/44 There was no timeout on the connect() call when using a Unix domain
2098 socket in the ${readsocket expansion. There now is.
2100 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2101 be meaningful with "accept".
2103 SC/01 Eximstats V1.43
2104 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2106 SC/02 Eximstats V1.44
2107 Use a glob alias rather than an array ref in the generated
2108 parser. This improves both readability and performance.
2110 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2111 Collect SpamAssassin and rejection statistics.
2112 Don't display local sender or destination tables unless
2113 there is data to show.
2114 Added average volumes into the top table text output.
2116 SC/04 Eximstats V1.46
2117 Collect data on the number of addresses (recipients)
2118 as well as the number of messages.
2120 SC/05 Eximstats V1.47
2121 Added 'Message too big' to the list of mail rejection
2122 reasons (thanks to Marco Gaiarin).
2124 SC/06 Eximstats V1.48
2125 Mainlog lines which have GMT offsets and are too short to
2126 have a flag are now skipped.
2128 SC/07 Eximstats V1.49 (Alain Williams)
2129 Added the -emptyok flag.
2131 SC/08 Eximstats V1.50
2132 Fixes for obtaining the IP address from reject messages.
2134 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2135 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2136 whitesspace changes from 4.64-PH/27
2138 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2141 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2142 are found, allow negative numbers in numeric criteria)
2144 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2146 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2147 to match changes made in 4.64-PH/43
2149 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2151 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2152 "domain" is actually an IP address, and force a failure. This locks out
2153 those revolvers/nameservers that support "A-for-A" lookups, in
2154 contravention of the specifications.
2156 PH/48 When a host name was looked up from an IP address, and the subsequent
2157 forward lookup of the name timed out, the host name was left in
2158 $sender_host_name, contrary to the specification.
2160 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2161 restricted to single-key lookups, Exim was not diagnosing an error if
2162 * or *@ was used with a query-style lookup.
2164 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2166 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2167 long ago, but noone interested enough thought of it. Let's just say that
2168 the "1.1" means that there are some new functions that weren't there at
2169 some point in the past.
2171 PH/51 Error processing for expansion failure of helo_data from an smtp
2172 transport during callout processing was broken.
2174 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2175 tested/used via the -bh/-bhc/-bs options.
2177 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2178 bug, fixed in subsequent PCRE releases).
2180 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2181 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2183 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2189 SC/01 Use a glob alias rather than an array ref in eximstats generated
2190 parser. This improves both readability and performance.
2192 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2193 Don't display local sender or destination tables in eximstats unless
2194 there is data to show.
2195 Added average volumes into the eximstats top table text output.
2197 SC/03 Collect data on the number of addresses (recipients) as well
2198 as the number of messages in eximstats.
2200 TF/01 Correct an error in the documentation for the redirect router. Exim
2201 does (usually) call initgroups() when daemonizing.
2203 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2204 with consistent privilege compared to when running as a daemon.
2206 TF/03 Note in the spec that $authenticated_id is not set for local
2207 submissions from trusted users.
2209 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2210 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2212 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2213 by adding some example configuration directives to the default
2214 configuration file. A little bit of work is required to uncomment the
2215 directives and define how usernames and passwords are checked, but
2216 there is now a framework to start from.
2218 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2219 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2220 without this. I don't know how relevant this is to other LDAP libraries.
2222 PH/02 Add the verb name to the "unknown ACL verb" error.
2224 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2226 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2228 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2229 directory not expanded when it should be if an expanded home directory
2230 was set for the address (which is overridden by the transport).
2232 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2235 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2236 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2237 because it is too late at that time, and has no effect.
2239 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2240 security issue with \' (bugzilla #107). I could not use the
2241 PQescapeStringConn() function, because it needs a PGconn value as one of
2244 PH/08 When testing addresses using -bt, indicate those final addresses that
2245 are duplicates that would not cause an additional delivery. At least one
2246 person was confused, thinking that -bt output corresponded to deliveries.
2247 (Suppressing duplicates isn't a good idea as you lose the information
2248 about possibly different redirections that led to the duplicates.)
2250 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2251 systems where poll() doesn't work, in particular OS X.
2253 PH/10 Added more information to debugging output for retry time not reached.
2255 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2256 operations in malware.c.
2258 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2261 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2262 syslog_duplication set false, log lines that would normally be written
2263 both the the main log and to the reject log were not written to syslog at
2266 PH/14 In the default configuration, change the use of "message" in ACL warn
2267 statements to "add_header".
2269 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2270 not followed by a command (e.g. "seen endif").
2272 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2273 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2276 PH/17 Added extra conditions to the default value of delay_warning_condition
2280 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2281 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2282 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2285 The Auto-Submitted: and various List- headers are standardised, whereas I
2286 don't think Precedence: ever was.
2288 PH/18 Refactored debugging code in route_finduser() to show more information,
2289 in particular, the error code if getpwnam() issues one.
2291 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2292 This is apparently needed in addition to the PH/07 change above to avoid
2293 any possible encoding problems.
2295 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2296 but not after initializing Perl.
2298 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2299 output them only if debugging. By default they are written stderr,
2300 apparently, which is not desirable.
2302 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2305 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2308 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2310 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2311 authenticated or an ident call has been made. Suppress the default
2312 values for $authenticated_id and $authenticated_sender (but permit -oMai
2313 and -oMas) when testing with -bh.
2315 PH/24 Re-jigged the order of the tests in the default configuration so that the
2316 tests for valid domains and recipients precede the DNS black list and CSA
2317 tests, on the grounds that those ones are more expensive.
2319 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2320 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2321 This bug exists in every version of Exim that I still have, right back to
2324 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2325 However, an attempt to turn on case-sensitivity in a regex key by
2326 including (?-i) didn't work because the subject string was already
2327 lowercased, and the effects were non-intuitive. It turns out that a
2328 one-line patch can be used to allow (?-i) to work as expected.
2334 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2335 other effects) broke the use of negated acl sub-conditions.
2337 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2340 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2341 "Deny" causes Exim to reject the incoming connection with a 554 error.
2342 Unfortunately, if there is a major crisis, such as a disk failure,
2343 tcp-wrappers gives "deny", whereas what one would like would be some
2344 kind of temporary error. A kludge has been added to help with this.
2345 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2346 554 error is used if errno is still zero or contains ENOENT (which occurs
2347 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2350 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2352 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2353 errors. Otherwise a message that provokes a temporary error (when other
2354 messages do not) can cause a whole host to time out.
2356 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2357 addresses were routed directly to files or pipes from a redirect router.
2358 File deliveries just didn't batch; pipe deliveries might have suffered
2361 PH/06 A failure to get a lock for a hints database would erroneously always say
2362 "Failed to get write lock", even when it was really a read lock.
2364 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2365 of 0600. This has been changed to use the value of the lockfile_mode
2366 option (which defaults to 0600).
2368 PH/08 Applied small patch from the Sieve maintainer.
2370 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2371 folder from quota calculations, a direct delivery into this folder messed
2372 up the contents of the maildirsize file. This was because the regex was
2373 used only to exclude .Trash (or whatever) when the size of the mailbox
2374 was calculated. There was no check that a delivery was happening into an
2375 excluded directory. This bug has been fixed by ignoring all quota
2376 processing for deliveries into excluded directories.
2378 PH/10 Added the maildirfolder_create_regex option to appendfile.
2384 PH/01 The code for finding all the local interface addresses on a FreeBSD
2385 system running IPv6 was broken. This may well have applied to all BSD
2386 systems, as well as to others that have similar system calls. The broken
2387 code found IPv4 interfaces correctly, but gave incorrect values for the
2388 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2389 that it would not match correctly against @[] and not recognize the IPv6
2392 PH/02 The ipliteral router was not recognizing addresses of the form user@
2393 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2395 PH/03 Added disable_ipv6.
2397 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2398 decoded form, because it is most often used to construct To: headers
2399 lines in autoreplies, and the decoded form may well be syntactically
2400 invalid. However, $reply_address has leading white space removed, and all
2401 newlines turned into spaces so that the autoreply transport does not
2404 PH/05 If group was specified without a user on a router, and no group or user
2405 was specified on a transport, the group from the router was ignored.
2407 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2408 for visible compile-time settings that can be used to change these
2409 numbers, for those that want even more. Backwards compatibility with old
2410 spool files has been maintained. However, going back to a previous Exim
2411 release will lost any variables that are in spool files.
2413 PH/07 Two small changes when running in the test harness: increase delay when
2414 passing a TCP/IP connection to a new process, in case the original
2415 process has to generate a bounce, and remove special handling of
2416 127.0.0.2 (sic), which is no longer necessary.
2418 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2419 be the same on different OS.
2421 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2424 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2425 whether --show-vars was specified or not
2427 JJ/02 exipick: Added support for new ACL variable spool format introduced
2430 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2431 syntactically invalid From: or Reply-to: line, and a filter used this to
2432 generate an autoreply, and therefore failed to obtain an address for the
2433 autoreply, Exim could try to deliver to a non-existent relative file
2434 name, causing unrelated and misleading errors. What now happens is that
2435 it logs this as a hard delivery error, but does not attempt to create a
2438 PH/11 The exinext utility has a -C option for testing purposes, but although
2439 the given file was scanned by exinext itself; it wasn't being passed on
2440 when Exim was called.
2442 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2443 an end-of-file indication when reading a command response.
2445 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2446 compiled. In many other places in Exim, IPv6 addresses are always
2447 recognized, so I have changed this. It also means that IPv4 domain
2448 literals of the form [IPV4:n.n.n.n] are now always recognized.
2450 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2451 used if the router is not running as root, for example, when verifying at
2452 ACL time, or when using -bh. The debugging output from this situation was
2453 non-existent - all you got was a failure to exec. I have made two
2456 (a) Failures to set uid/gid, the current directory, or a process leader
2457 in a subprocess such as that created by queryprogram now generate
2458 suitable debugging ouput when -d is set.
2460 (b) The queryprogram router detects when it is not running as root,
2461 outputs suitable debugging information if -d is set, and then runs
2462 the subprocess without attempting to change uid/gid.
2464 PH/15 Minor change to Makefile for building test_host (undocumented testing
2467 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2468 additional section of a DNS packet that returns MX or SRV records.
2469 Instead, it always explicitly searches for A/AAAA records. This avoids
2470 major problems that occur when a DNS server includes only records of one
2471 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2472 fixed another bug: if SRV records were looked up and the corresponding
2473 address records were *not* found in the additional section, the port
2474 values from the SRV records were lost.
2476 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2477 using the correct key (the original address) when searching the retry
2478 rules in order to find which one to use for generating the retry hint.
2480 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2481 adding the default one. Similarly, if it contains a Reply-To: header, the
2482 errors_reply_to option, if set, is not used.
2484 PH/19 When calculating a retry time, Exim used to measure the "time since
2485 failure" by looking at the "first failed" field in the retry record. Now
2486 it does not use this if it is later than than the arrival time of the
2487 message. Instead it uses the arrival time. This makes for better
2488 behaviour in cases where some deliveries succeed, thus re-setting the
2489 "first failed" field. An example is a quota failure for a huge message
2490 when small messages continue to be delivered. Without this change, the
2491 "time since failure" will always be short, possible causing more frequent
2492 delivery attempts for the huge message than are intended.
2493 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2495 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2496 $1, $2, $3) because the numerical variables can be reset during some
2497 expansion items (e.g. "match"), thereby losing the authentication data.
2499 PH/21 Make -bV show the size of off_t variables so that the test suite can
2500 decide whether to run tests for quotas > 2G.
2502 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2503 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2504 filecount value is greater than 2G or if a quota value is greater than 2G
2505 on a system where the size of off_t is not greater than 4, a panic error
2508 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2509 never match. The debug and -bh output now contains an explicit error
2510 message indicating a malformed IPv4 address or mask.
2512 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2513 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2514 PH/23 above applies.
2516 PH/25 Do not write to syslog when running in the test harness. The only
2517 occasion when this arises is a failure to open the main or panic logs
2518 (for which there is an explicit test).
2520 PH/26 Added the /no_tell option to "control=freeze".
2522 PH/27 If a host name lookup failed very early in a connection, for example, if
2523 the IP address matched host_lookup and the reverse lookup yielded a name
2524 that did not have a forward lookup, an error message of the form "no IP
2525 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2526 could be logged. Now it outputs the IP address instead of "NULL".
2528 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2529 allows the sender and the authenticated sender to be set when
2530 submitting a message from within Exim. Since child_open_exim() is
2531 documented for local_scan(), the new function should be too.
2533 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2534 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2535 results in an empty string is now treated as unset.
2537 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2539 PH/31 Added sender_verify_fail logging option.
2541 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2542 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2543 tidied the source and removed it altogether.
2545 PH/33 When a queue run was abandoned because the load average was too high, a
2546 log line was always written; now it is written only if the queue_run log
2547 selector is set. In addition, the log line for abandonment now contains
2548 information about the queue run such as the pid. This is always present
2549 in "start" and "stop" lines but was omitted from the "abandon" line.
2551 PH/34 Omit spaces between a header name and the colon in the error message that
2552 is given when verify = headers_syntax fails (if there are lots of them,
2553 the message gets confusing).
2555 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2556 names, as there are now some PTR records that contain slashes. This check
2557 is only to protect against broken name servers that fall over on strange
2558 characters, so the fact that it applies to all lookups doesn't matter.
2560 PH/36 Now that the new test suite is complete, we can remove some of the
2561 special code in Exim that was needed for the old test suite. For example,
2562 sorting DNS records because real resolvers return them in an arbitrary
2563 order. The new test suite's fake resolver always returns records in the
2566 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2567 bounces) except when queue_only is set, to avoid logging races between
2568 the different processes.
2570 PH/38 Panic-die if .include specifies a non-absolute path.
2572 PH/39 A tweak to the "H" retry rule from its user.
2574 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2575 a label. They prevented compilation on older perls.
2577 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2578 a warning to be raised on newish perls.
2580 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2581 on queue. Changes to match documented behaviour of showing count of
2582 messages matching specified criteria.
2584 PH/40 Changed the default ident timeout from 30s to 5s.
2586 PH/41 Added support for the use of login_cap features, on those BSD systems
2587 that have them, for controlling the resources used by pipe deliveries.
2589 PH/42 The content-scanning code uses fopen() to create files in which to put
2590 message data. Previously it was not paying any attention to the mode of
2591 the files. Exim runs with umask(0) because the rest of the code creates
2592 files with open(), and sets the required mode explicitly. Thus, these
2593 files were ending up world-writeable. This was not a big issue, because,
2594 being within the spool directory, they were not world-accessible. I have
2595 created a function called modefopen, which takes an additional mode
2596 argument. It sets umask(777), creates the file, chmods it to the required
2597 mode, then resets the umask. All the relevant calls to fopen() in the
2598 content scanning code have been changed to use this function.
2600 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2601 to 24 hours. This avoids potential overflow problems when processing G
2602 and H retry rules. I suspect nobody ever tinkers with this value.
2604 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2606 PH/45 When the plaintext authenticator is running as a client, the server's
2607 challenges are checked to ensure they are valid base64 strings. By
2608 default, the authentication attempt is cancelled if an invalid string is
2609 received. Setting client_ignore_invalid_base64 true ignores these errors.
2610 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2611 they are received. Thus, the responses can be made to depend on the
2612 challenges. If an invalid string is ignored, an empty string is placed in
2615 PH/46 Messages that are created by the autoreply transport now contains a
2616 References: header, in accordance with RFCs 2822 and 3834.
2618 PH/47 Added authenticated_sender_force to the smtp transport.
2620 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2622 PH/49 Installed latest patch from the Sieve maintainer.
2624 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2625 was also set, the appendfile transport was unnecessarily scanning a
2626 directory of message files (e.g. for maildir delivery) to find the count
2627 of files (along with the size), even though it did not need this
2628 information. It now does the scan only if it needs to find either the
2629 size of the count of files.
2631 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2633 PH/52 Two bugs concerned with error handling when the smtp transport is
2636 (i) Exim was not creating retry information for temporary errors given
2637 for individual recipients after the DATA command when the smtp transport
2638 was used in LMTP mode. This meant that they could be retried too
2639 frequently, and not timed out correctly.
2641 (ii) Exim was setting the flag that allows error details to be returned
2642 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2643 recipients that were returned after the DATA command.
2645 PH/53 This is related to PH/52, but is more general: for any failing address,
2646 when detailed error information was permitted to be returned to the
2647 sender, but the error was temporary, then after the final timeout, only
2648 "retry timeout exceeded" was returned. Now it returns the full error as
2649 well as "retry timeout exceeded".
2651 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2652 do this, and (what is worse) MTAs that accept it.
2654 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2655 will now be deprecated.
2657 PH/56 New os.c-cygwin from the Cygwin maintainer.
2659 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2660 formats (previously only available in exim formats via -bpr, -bpru,
2661 and -bpra. Now also available in native and exiqgrep formats)
2663 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2664 with very large, slow to parse queues
2666 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2668 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2670 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2671 responses to authentication challenges, though it was showing the
2672 challenges; (ii) I've removed the CR characters from the debug output for
2675 PH/58 Allow for the insertion of a newline as well as a space when a string
2676 is turned into more than one encoded-word during RFC 2047 encoding. The
2677 Sieve code now uses this.
2679 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2680 data_4xx, lost_connection, tls_required.
2682 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2683 message was being sent as an SMTP response.
2685 PH/61 Add -l and -k options to exicyclog.
2687 PH/62 When verifying, if an address was redirected to one new address, so that
2688 verification continued, and the new address failed or deferred after
2689 having set something in $address_data, the value of $address_data was not
2690 passed back to the ACL. This was different to the case when no
2691 redirection occurred. The value is now passed back in both cases.
2693 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2694 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2695 use login_cap.h, so on its own it isn't the distinguishing feature. The
2696 new name refers directly to the setclassresources() function.
2698 PH/65 Added configuration files for NetBSD3.
2700 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2702 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2703 is preferred over IPv4.
2705 PH/68 The bounce_return_message and bounce_return_body options were not being
2706 honoured for bounces generated during the reception of non-SMTP messages.
2707 In particular, this applied to messages rejected by the ACL. This bug has
2708 been fixed. However, if bounce_return_message is true and bounce_return_
2709 body is false, the headers that are returned for a non-SMTP message
2710 include only those that have been read before the error was detected.
2711 (In the case of an ACL rejection, they have all been read.)
2713 PH/69 The HTML version of the specification is now built in a directory called
2714 spec_html instead of spec.html, because the latter looks like a path with
2715 a MIME-type, and this confuses some software.
2717 PH/70 Catch two compiler warnings in sieve.c.
2719 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2720 function verify_get_ident() calls ip_connect() to connect a socket, but
2721 if the "connect()" function timed out, ip_connect() used to close the
2722 socket. However, verify_get_ident() also closes the socket later, and in
2723 between Exim writes to the log, which may get opened at this point. When
2724 the socket was closed in ip_connect(), the log could get the same file
2725 descriptor number as the socket. This naturally causes chaos. The fix is
2726 not to close the socket in ip_connect(); the socket should be closed by
2727 the function that creates it. There was only one place in the code where
2728 this was missing, in the iplookup router, which I don't think anybody now
2729 uses, but I've fixed it anyway.
2731 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2732 well as to direct DNS lookups. Otherwise the handling of names in host
2733 lists is inconsistent and therefore confusing.
2739 PH/01 Two changes to the default runtime configuration:
2741 (1) Move the checks for relay_from_hosts and authenticated clients from
2742 after to before the (commented out) DNS black list checks.
2744 (2) Add control=submission to the relay_from_hosts and authenticated
2745 clients checks, on the grounds that messages accepted by these
2746 statements are most likely to be submissions.
2748 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2750 (1) Generate an error if the third argument for the ${prvs expansion is
2753 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2756 (3) Reset the variables that are obtained from the first argument of
2757 ${prvscheck and used in the second argument before leaving the code,
2758 because their memory is reclaimed, so using them afterwards may do
2761 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2762 one (it's much easier than Tom thought :-).
2764 (5) Because of (4), we can now allow for the use of $prvscheck_result
2765 inside the third argument.
2767 PH/03 For some reason, the default setting of PATH when running a command from
2768 a pipe transport was just "/usr/bin". I have changed it to
2771 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2772 anything to be listed in the output from -bV.
2774 PH/05 When a filter generated an autoreply, the entire To: header line was
2775 quoted in the delivery log line, like this:
2777 => >A.N.Other <ano@some.domain> <original@ddress> ...
2779 This has been changed so that it extracts the operative address. There
2780 may be more than one such address. If so, they are comma-separated, like
2783 => >ano@some.domain,ona@other.domain <original@ddress> ...
2785 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2786 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2787 not being looked up in the rDNS to get a host name, Exim was showing the
2788 IP address twice in Received: lines, even though the IP addresses were
2789 identical. For example:
2791 Received: from [1.2.3.4] (helo=[1.2.3.4])
2793 However, if the real host name was known, it was omitting the HELO data
2794 if it matched the actual IP address. This has been tidied up so that it
2795 doesn't show the same IP address twice.
2797 PH/07 When both +timestamp and +memory debugging was on, the value given by
2798 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2799 called by the debug printing, thereby overwriting the timestamp buffer.
2800 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2802 PH/08 When the original message was included in an autoreply transport, it
2803 always said "this is a copy of the message, including all the headers",
2804 even if body_only or headers_only was set. It now gives an appropriate
2807 PH/09 Applied a patch from the Sieve maintainer which:
2809 o fixes some comments
2810 o adds the (disabled) notify extension core
2811 o adds some debug output for the result of if/elsif tests
2812 o points to the current vacation draft in the documentation
2813 and documents the missing references header update
2817 o fixes a bug in processing the envelope test (when testing
2818 multiple envelope elements, the last element determinted the
2821 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2822 Electronic Mail") by including:
2824 Auto-submitted: auto-generated
2826 in the messages that it generates (bounce messages and others, such as
2827 warnings). In the case of bounce messages for non-SMTP mesages, there was
2828 also a typo: it was using "Auto_submitted" (underscore instead of
2829 hyphen). Since every message generated by Exim is necessarily in response
2830 to another message, thes have all been changed to:
2832 Auto-Submitted: auto-replied
2834 in accordance with these statements in the RFC:
2836 The auto-replied keyword:
2838 - SHOULD be used on messages sent in direct response to another
2839 message by an automatic process,
2841 - MUST NOT be used on manually-generated messages,
2843 - MAY be used on Delivery Status Notifications (DSNs) and Message
2844 Disposition Notifications (MDNs),
2846 - MUST NOT be used on messages generated by automatic or periodic
2847 processes, except for messages which are automatic responses to
2850 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2851 to the default Received: header definition.
2853 PH/12 Added log selector acl_warn_skipped (default on).
2855 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2856 variables because (a) they are in the wrong storage pool and (b) even if
2857 they were copied, it wouldn't work properly because of the caching.
2859 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2860 checking when decoding. Apparently there are clients that generate
2861 overlong encoded strings. Why am I not surprised?
2863 PH/15 If the first argument of "${if match_address" was not empty, but did not
2864 contain an "@" character, Exim crashed. Now it writes a panic log message
2865 and treats the condition as false.
2867 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2869 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2870 extension "envelope-auth". The code is finished and in agreement with
2871 other implementations, but there is no documentation so far and in fact,
2872 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2873 not changing the active code.
2875 Print executed "if" and "elsif" statements when debugging is used. This
2876 helps a great deal to understand what a filter does.
2878 Document more things not specified clearly in RFC3028. I had all this
2879 sorted out, when out of a sudden new issues came to my mind. Oops."
2881 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2884 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2885 canonical form (as documented). However, after a host name lookup from
2886 the IP address, check_host() was doing a simple string comparison with
2887 addresses acquired from the DNS when checking that the found name did
2888 have the original IP as one of its addresses. Since any found IPv6
2889 addresses are likely to be in abbreviated form, the comparison could
2890 fail. Luckily, there already exists a function for doing the comparison
2891 by converting both addresses to binary, so now that is used instead of
2892 the text comparison.
2894 PH/20 There was another similar case to PH/19, when a complete host name was
2895 given in a host list; looking up its IP address could give an abbreviated
2896 form, whereas the current host's name might or might not be abbreviated.
2897 The same fix has been applied.
2903 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2904 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2907 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2908 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2910 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2912 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2913 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2914 Also, for "Auto-Submitted", treat anything other than "no" as
2915 non-personal, in accordance with RFC 3834. (Previously it treated
2916 anything starting "auto-" as non-personal.)
2918 TF/01 The control=submission/name=... option had a problem with syntax
2919 errors if the name included a slash character. The /name= option
2920 now slurps the rest of the string, so it can include any characters
2921 but it must come last in the list of options (after /sender_retain
2924 PH/05 Some modifications to the interface to the fake nameserver for the new
2932 TK/01 Added the "success_on_redirect" address verification option. See
2933 NewStuff for rationale and an example.
2935 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2937 PH/02 Patch to exigrep to allow it to work on syslog lines.
2939 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2940 fread() to skip over the body file's header line, because in Cygwin the
2941 header line is locked and is inaccessible.
2943 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2944 co-exist for some time) to make it clear that it is the Exim ID that is
2945 referenced, not the Message-ID: header line.
2947 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2948 string_format() function, because snprintf() does not exist on all
2951 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2952 ${stat: expansion item.
2954 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2955 protocol synchronization error", to keep the pedants happy.
2957 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2958 well as for IRIX systems, when gcc is being used. See the host.c source
2961 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2963 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2966 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2967 transport and to the smtp transport in LMTP mode.
2969 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2971 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2972 run a filter in a subprocess. This could lead to confusion in subsequent
2973 lookups in the parent process. There should also be a search_tidyup() at
2974 the end of the subprocess.
2976 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2977 only if the host matched helo_try_verify_hosts, which caused the
2978 verification to occur when the EHLO/HELO command was issued. The ACL just
2979 tested the remembered result. Now, if a previous verification attempt has
2980 not happened, "verify = helo" does it there and then.
2982 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2984 TK/03 Fix log output including CR from clamd.
2986 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2987 exist provoked a memory error which could cause a segfault.
2989 PH/15 Installed PCRE 6.2
2991 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2993 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2994 of the problem. Specifically, suggested +O2 rather than +O1 for the
2997 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2999 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3000 record, the port was not correctly logged when the outgoing_port log
3001 selector was set (it logged the transort's default port).
3003 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3004 fallback_hosts, and "hosts" in the smtp transport.
3006 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3007 host errors such as "Connection refused".
3009 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3010 authentication with radiusclient 0.4.9:
3012 - Error returned from rc_read_config was caught wrongly
3013 - Username/password not passed on to radius server due to wrong length.
3015 The presumption is that some radiusclient API changes for 4.51/PH/17
3016 were not taken care of correctly. The code is still untested by me (my
3017 Linux distribution still has 0.3.2 of radiusclient), but it was
3018 contributed by a Radius user.
3020 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3021 expanding the "port" option of the smtp transport.
3023 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3024 while reading a MIME header. Thanks to Tom Hughes for a patch.
3026 PH/24 Include config.h inside local_scan.h so that configuration settings are
3029 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3030 that in an ACL for RCPT (for example), you can examine exactly what was
3033 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3034 commands, but it was not correctly comparing the address with the actual
3035 client host address. Thus, it would show the EHLO address in Received:
3036 header lines when this was not necessary.
3038 PH/27 Added the % operator to ${eval:}.
3040 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3041 it should be ignoring failures (because with -C, for example, it has lost
3042 privilege). It wasn't ignoring creation failures other than "already
3045 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3048 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3049 because an input line was too long, either on its own, or by virtue of
3050 too many continuations, the temporary file was not being removed, and the
3051 return code was incorrect.
3053 PH/31 Missing "BOOL" in function definition in filtertest.c.
3055 PH/32 Applied Sieve patches from the maintainer.
3057 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3059 PH/33 Added "verify = not_blind".
3061 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3062 Local/Makefile (with some defaults set). These are used in built scripts
3063 such as exicyclog, but they have never been used in the exim_install
3064 script (though there are many overriding facilities there). I have
3065 arranged that the exim_install script now takes note of these two
3068 PH/35 Installed configuration files for Dragonfly.
3070 PH/36 When a locally submitted message by a trusted user did not contain a
3071 From: header, and the sender address was obtained from -f or from an SMTP
3072 MAIL command, and the trusted user did not use -F to supply a sender
3073 name, $originator_name was incorrectly used when constructing a From:
3074 header. Furthermore, $originator_name was used for submission mode
3075 messages from external hosts without From: headers in a similar way,
3076 which is clearly wrong.
3078 PH/37 Added control=suppress_local_fixups.
3080 PH/38 When log_selector = +received_sender was set, and the addition of the
3081 sender made the log line's construction buffer exactly full, or one byte
3082 less than full, an overflow happened when the terminating "\n" was
3085 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3086 when the result of a list match is failure because a DNS lookup failed.
3088 PH/40 RM_COMMAND is now used in the building process.
3090 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3091 the "build-* directories that it finds.
3093 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3094 address in a domain literal was a prefix of an interface address.
3096 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3097 when verifying a sender address, unless rewrite_headers is false.
3099 PH/44 Wrote a long comment about why errors_to addresses are verified as
3100 recipients, not senders.
3102 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3103 the ratelimit ACL was added.
3105 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3107 PH/46 Added extra information about PostgreSQL errors to the error string.
3109 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3110 avoiding the need to install special zones in a real server. This is
3111 backwards compatible; if it can't find the fake resolver, it drops back.
3112 Thus, both old and new test suites can be run.
3114 TF/02 Added util/ratelimit.pl
3116 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3117 clock is set back in time.
3119 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3120 Candler <B.Candler@pobox.com>.
3122 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3123 OK for addresses that are the result of redirections.
3125 PH/48 A number of further additions for the benefit of the new test suite,
3126 including a fake gethostbyname() that interfaces to the fake DNS resolver
3129 TF/06 The fix for widen_domains has also been applied to qualify_single and
3130 search_parents which are the other dnslookup options that can cause
3133 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3136 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3138 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3139 service" if it encountered an MX record with an empty target hostname.
3140 The message is now "an MX or SRV record indicated no SMTP service".
3142 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3143 if the DNS of the sending site is misconfigured. This is quite a
3144 common situation. This change restores the behaviour of treating a
3145 helo verification defer as a failure.
3147 PH/51 If self=fail was set on a router, the bounce message did not include the
3148 actual error message.
3154 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3156 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3157 command itself ended in error, the underlying message about the transport
3158 filter timeout was being overwritten with the pipe command error. Now the
3159 underlying error message should be appended to the second error message.
3161 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3163 PH/02 Reduce the amount of output that "make" produces by default. Full output
3164 can still be requested.
3166 PH/03 The warning log line about a condition test deferring for a "warn" verb
3167 was being output only once per connection, rather than after each
3168 occurrence (because it was using the same function as for successful
3169 "warn" verbs). This seems wrong, so I have changed it.
3171 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3172 it should not have, which might have caused a crash in the right
3173 circumstances, but probably never did.
3175 PH/04 Installed a modified version of Tony Finch's patch to make submission
3176 mode fix the return path as well as the Sender: header line, and to
3177 add a /name= option so that you can make the user's friendly name appear
3180 TF/03 Added the control = fakedefer ACL modifier.
3182 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3183 Mark Lowes for thorough testing.
3185 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3187 TK/03 Merged latest SRS patch from Miles Wilton.
3189 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3190 with the definition in sysexits.h (which is #included earlier).
3191 Fortunately, Exim does not actually use EX_OK. The code used to try to
3192 preserve the sysexits.h value, by assumimg that macro definitions were
3193 scanned for macro replacements. I have been disabused of this notion,
3194 so now the code just undefines EX_OK before #including unistd.h.
3196 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3197 in the smtp transport. When a block could not be written in a single
3198 write() function, the timeout was being re-applied to each part-write.
3199 This seems wrong - if the receiver was accepting one byte at a time it
3200 would take for ever. The timeout is now adjusted when this happens. It
3201 doesn't have to be particularly precise.
3203 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3204 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3206 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3207 without a domain if the check to <postmaster@domain> fails.
3209 SC/01 Eximstats: added -xls and the ability to specify output files
3210 (patch written by Frank Heydlauf).
3212 SC/02 Eximstats: use FileHandles for outputing results.
3214 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3216 SC/04 Eximstats: fixed display of large numbers with -nvr option
3218 SC/05 Eximstats: fixed merging of reports with empty tables.
3220 SC/06 Eximstats: added the -include_original_destination flag
3222 SC/07 Eximstats: removed tabs and trailing whitespace.
3224 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3226 TK/06 MBOX spool code: Add real "From " MBOX separator line
3227 so the .eml file is really in mbox format (even though
3228 most programs do not really care). Patch from Alex Miller.
3230 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3231 The latter is generated from $received_to and is only set if the
3232 message has one envelope recipient. SA can use these headers,
3233 obviously out-of-the-box. Patch from Alex Miller.
3235 PH/08 The ${def test on a variable was returning false if the variable's
3236 value was "0", contrary to what the specification has always said!
3237 The result should be true unless the variable is empty.
3239 PH/09 The syntax error of a character other than { following "${if
3240 def:variable_name" (after optional whitespace) was not being diagnosed.
3241 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3242 accidental colon was present, for example, could give incorrect results.
3244 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3245 result is used (not including appendfile, where other changes are about
3248 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3249 This involved changing a lot of size variables from int to off_t. It
3250 should work with maildirs and everything.
3252 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3253 spamd dying while we are connected to it.
3255 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3258 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3259 function for BDB 4.3.
3261 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3263 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3264 the sources for PCRE have been re-arranged and more files are now
3267 PH/15 The code I had for printing potentially long long variables in PH/11
3268 above was not the best (it lost precision). The length of off_t variables
3269 is now inspected at build time, and an appropriate printing format (%ld
3270 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3271 to be "long long int" or "long int". This is needed for the internal
3272 formatting function string_vformat().
3274 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3275 the configuration file to be ":syslog", then the script "guesses" where
3276 the logs files are, rather than using the compiled in default. In our
3277 case the guess is not the same as the compiled default, so the script
3278 suddenly stopped working when I started to use syslog. The patch checks
3279 to see if log_file_path is "". If so, it attempts to read it from exim
3280 with no configuration file to get the compiled in version, before it
3281 falls back to the previous guessing code."
3283 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3284 implementing BATV in an Exim configuration. See NewStuff for the gory
3287 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3288 Makefile that are specific to HP-UX.
3290 PH/18 If the "use_postmaster" option was set for a recipient callout together
3291 with the "random" option, the postmaster address was used as the MAIL
3292 FROM address for the random test, but not for the subsequent recipient
3293 test. It is now used for both.
3295 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3296 patch removes a few documentation additions to RFC 3028, because the
3297 latest draft now contains them. It adds the new en;ascii-case comparator
3298 and a new error check for 8bit text in MIME parts. Comparator and
3299 require names are now matched exactly. I enabled the subaddress
3300 extension, but it is not well tested yet (read: it works for me)."
3302 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3303 rework some of the code of TK/09 above to avoid the hardwired use of
3304 "%lld" and "long long". Replaced the call to snprintf() with a call to
3307 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3308 records point to non-existent hosts", "retry timeout exceeded", and
3309 "retry time not reached for any host after a long failure period".
3311 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3312 experimental DomainKeys support:
3314 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3315 (2) On an error such as an illegally used "control", the wrong name for
3316 the control was given.
3318 These problems did NOT occur unless DomainKeys support was compiled.
3320 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3322 PH/24 Added ${if match_ip condition.
3324 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3325 databases so that it will be absolutely obvious if a crash occurs in the
3326 DB library. This is a regular occurrence (often caused by mis-matched
3329 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3330 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3331 picked up on a user's system that detects such things. There doesn't seem
3332 to be a gcc warning option for this - only an attribute that has to be
3333 put on the function's prototype. It seems that in Fedora Core 4 they have
3334 set this on a number of new functions. No doubt there will be more in due
3337 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3338 specify a transport. However, if an address that was verified by such a
3339 router was the subject of a callout, Exim crashed because it tried to
3340 read the rcpt_include_affixes from the non-existent transport. Now it
3341 just assumes that the setting of that option is false. This bug was
3342 introduced by 4.51/PH/31.
3344 PH/28 Changed -d+all to exclude +memory, because that information is very
3345 rarely of interest, but it makes the output a lot bigger. People tend to
3346 do -d+all out of habit.
3348 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3349 code in os-type was giving problems when libc.so lives in lib64, like on
3352 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3353 aren't the modern standard, and it seems that some systems' include files
3354 don't always have them. Exim was already checking for some of the newer
3355 ones like T_AAAA, and defining it itself. I've added checks for all the
3356 record types that Exim uses.
3358 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3359 not automatically generating a new one, as it is supposed to. This
3360 prevented TLS from working. If the file did exist, but contained invalid
3361 data, a new version was generated, as expected. It was only the case of a
3362 non-existent file that was broken.
3364 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3365 with a change in libdomainkeys > 0.64.
3367 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3368 from DNS. If the selector record carries the flag, it now has
3369 precedence over the domain-wide flag.
3371 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3373 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3374 the use of a port name that isn't defined in /etc/services) occurred, the
3375 message was deferred as in a normal delivery, and thus remained on the
3376 spool, instead of being failed because of the mua_wrapper setting. This
3377 is now fixed, and I tidied up some of the mua_wrapper messages at the
3380 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3381 the messages in a hash of arrays rather than using individual hashes.
3382 This is a bit cleaner and results in dramatic memory savings, albeit
3383 at a slight CPU cost.
3385 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3386 as requested by Marc Sherman.
3388 SC/10 Eximstats: added histograms for user specified patterns as requested
3391 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3393 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3394 fopen() in the content-scanning modules that did not already have it.
3400 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3401 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3403 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3405 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3407 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3408 to test Sieve filters that use "vacation".
3410 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3411 that changes the way the GnuTLS parameters are stored in the cache file.
3412 The new format can be generated externally. For backward compatibility,
3413 if the data in the cache doesn't make sense, Exim assumes it has read an
3414 old-format file, and it generates new data and writes a new file. This
3415 means that you can't go back to an older release without removing the
3418 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3419 work if there are any delivery delays because "one_time" forces the
3420 parent to be marked "delivered", so its unseen clone is never tried
3421 again. For this reason, Exim now forbids the simultaneous setting of
3424 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3425 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3426 entry if you want to know the details. The fix, however, neglected to
3427 consider the case when local delivery batching is involved. The test for
3428 "previously delivered" was not happening when checking to see if an
3429 address could be batched with a previous (undelivered) one; under
3430 certain circumstances this could lead to multiple deliveries to the same
3433 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3434 in its include files, and this causes problems building Exim.
3436 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3437 header_syntax) but Exim was just ignoring anything given after a slash.
3438 In particular, this caused confusion with an attempt to use "verify =
3439 reverse_host_lookup/defer_ok". An error is now given when options are
3440 supplied for verify items that do not have them. (Maybe reverse_host_
3441 lookup should have a defer_ok option, but that's a different point.)
3443 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3444 defined by RFC 821) to 2048, because there were problems with some AUTH
3445 commands, and RFC 1869 says the size should be increased for extended
3446 SMTP commands that take arguments.
3448 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3451 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3452 "unknown" error; now it says that the functionality isn't in the binary.
3454 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3455 an address' error message when a string expansion fails (syntax or
3456 whatever). Otherwise the password may appear in the log. Following change
3457 PH/42 below, there is no longer a chance of it appearing in a bounce
3460 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3462 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3463 of its addresses was ever tried. (Bugzilla bug #2.)
3465 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3466 the result incorrectly in the debug output. (It correctly added a newline
3467 to what was transported.)
3469 TF/01 Added $received_time.
3471 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3472 commented out examples of how to interface to a virus scanner and to
3473 SpamAssassin. Also added commented examples of av_scanner and
3474 spamd_address settings.
3476 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3477 and controls are allowed in which ACLs. There were a couple of minor
3478 errors. Some of the entries in the conditions table (which is a table of
3479 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3480 negation of where the condition IS allowed.
3482 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3484 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3485 header file does not have a version number, so I've had to invent a new
3486 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3487 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3488 radiusclient), but it was contributed by a Radius user.
3490 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3491 files or directories, for OpenSSL.
3493 PH/19 When an Exim process that is running as root has to create an Exim log
3494 file, it does so in a subprocess that runs as exim:exim so as to get the
3495 ownership right at creation (otherwise, other Exim processes might see
3496 the file with the wrong ownership). There was no test for failure of this
3497 fork() call, which would lead to the process getting stuck as it waited
3498 for a non-existent subprocess. Forks do occasionally fail when resources
3499 run out. I reviewed all the other calls to fork(); they all seem to check
3502 PH/20 When checking for unexpected SMTP input at connect time (before writing
3503 the banner), Exim was not dealing correctly with a non-positive return
3504 from the read() function. If the client had disconnected by this time,
3505 the result was a log entry for a synchronization error with an empty
3506 string after "input=" when read() returned zero. If read() returned -1
3507 (an event I could not check), uninitialized data bytes were printed.
3508 There were reports of junk text (parts of files, etc) appearing after
3511 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3513 PH/22 Added support for macro redefinition, and (re)definition in between
3514 driver and ACL definitions.
3516 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3517 forgetting to use the resulting value; it was using the unexpanded value.
3519 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3520 hadn't been configured. The fix is from Juergen Kreileder, who
3521 understands it better than I do:
3523 "Here's what I see happening with three configured cyrus_sasl
3524 authenticators configured (plain, login, cram-md5):
3526 On startup auth_cyrus_sasl_init() gets called for each of these.
3527 This means three calls to sasl_listmech() without a specified mech_list.
3528 => SASL tests which mechs of all available mechs actually work
3529 => three warnings about OTP not working
3530 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3532 With the patch, sasl_listmech() also gets called three times. But now
3533 SASL's mech_list option is set to the server_mech specified in the the
3534 authenticator. Or in other words, the answer from sasl_listmech()
3535 gets limited to just the mech you're testing for (which is different
3537 => the return list contains just 'plain' or 'login', 'cram-md5' or
3538 nothing depending on the value of ob->server_mech.
3540 I've just tested the patch: Authentication still works fine,
3541 unavailable mechs specified in the exim configuration are still
3542 caught, and the auth.log warnings about OTP are gone."
3544 PH/25 When debugging is enabled, the contents of the command line are added
3545 to the debugging output, even when log_selector=+arguments is not
3548 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3549 answer is "GNU", and only if the return is "GNU/something" is the answer
3552 PH/27 $acl_verify_message is now set immediately after the failure of a
3553 verification in an ACL, and so is available in subsequent modifiers. In
3554 particular, the message can be preserved by coding like this:
3556 warn !verify = sender
3557 set acl_m0 = $acl_verify_message
3559 Previously, $acl_verify_message was set only while expanding "message"
3560 and "log_message" when a very denied access.
3562 PH/28 Modified OS/os.c-Linux with
3564 -#ifndef OS_LOAD_AVERAGE
3565 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3567 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3568 nomenclature these days.)
3570 PH/29 Installed patch from the Sieve maintainer that adds the options
3571 sieve_useraddress and sieve_subaddress to the redirect router.
3573 PH/30 In these circumstances:
3574 . Two addresses routed to the same list of hosts;
3575 . First host does not offer TLS;
3576 . First host accepts first address;
3577 . First host gives temporary error to second address;
3578 . Second host offers TLS and a TLS session is established;
3579 . Second host accepts second address.
3580 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3581 and peerdn, if requested) that were in fact used only for the second
3584 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3585 attention to any local part prefix or suffix that was matched by the
3586 router that accepted the address. It now behaves in the same way as it
3587 does for delivery: the affixes are removed from the local part unless
3588 rcpt_include_affixes is set on the transport.
3590 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3591 timeout during the DATA phase of an incoming message.
3593 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3594 applied a patch sanctioned by the Sieve maintainer.
3596 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3597 the uid or gid is negative. A case of a negative gid caused this to be
3598 noticed. The fix allows for either to be negative.
3600 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3601 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3604 PH/36 The change PH/12 above was broken. Fixed it.
3606 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3607 the grounds that routing the same address twice would always produce the
3608 same answer. This might have been true once, but it is certainly no
3609 longer true now. Routing a child address may depend on the previous
3610 routing that produced that child. Some complicated redirection strategies
3611 went wrong when messages had multiple recipients, and made Exim's
3612 behaviour dependent on the order in which the addresses were given.
3614 I have moved the duplicate checking until after the routing is complete.
3615 Exim scans the addresses that are assigned to local and remote
3616 transports, and removes any duplicates. This means that more work will be
3617 done, as duplicates will always all be routed, but duplicates are
3618 presumably rare, so I don't expect this is of any significance.
3620 For deliveries to pipes, files, and autoreplies, the duplicate checking
3621 still happens during the routing process, since they are not going to be
3624 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3625 It corrects a timeout issue with spamd. This is Ian's comment: "The
3626 background is that sometimes spamd either never reads data from a
3627 connection it has accepted, or it never writes response data. The exiscan
3628 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3629 blindly assumes that writes won't block so it may never time out."
3631 PH/39 Allow G after quota size as well as K and M.
3633 PH/40 The value set for $authenticated_id in an authenticator may not contain
3634 binary zeroes or newlines because the value is written to log lines and
3635 to spool files. There was no check on this. Now the value is run through
3636 the string_printing() function so that such characters are converted to
3637 printable escape sequences.
3639 PH/41 $message_linecount is a new variable that contains the total number of
3640 lines in the message. Compare $body_linecount, which is the count for the
3643 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3644 bounce and delay warning messages, except in certain special cases, which
3647 (a) An SMTP error message from a remote host;
3648 (b) A message specified in a :fail: redirection;
3649 (c) A message specified in a "fail" command in a system filter;
3650 (d) A message specified in a FAIL return from the queryprogram router;
3651 (e) A message specified by the cannot_route_message router option.
3653 In these cases only, Exim does include the error details in bounce and
3654 warning messages. There are also a few cases where bland messages such
3655 as "unrouteable address" or "local delivery error" are given.
3657 PH/43 $value is now also set for the "else" part of a ${run expansion.
3659 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3660 being worked on, but at least Exim now implements the latest version to
3663 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3664 process to complete was treated as a delivery failure, a timeout while
3665 writing the message to the pipe was logged, but erroneously treated as a
3666 successful delivery. Such timeouts include transport filter timeouts. For
3667 consistency with the overall process timeout, these timeouts are now
3668 treated as errors, giving rise to delivery failures by default. However,
3669 there is now a new Boolean option for the pipe transport called
3670 timeout_defer, which, if set TRUE, converts the failures into defers for
3671 both kinds of timeout. A transport filter timeout is now identified in
3674 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3675 systems where "make" and "gmake" are different, calling "gmake" at top
3676 level broke things. I've arranged for the value of $(MAKE) to be passed
3677 from the Makefile to this script so that it can call the same version of
3681 A note about Exim versions 4.44 and 4.50
3682 ----------------------------------------
3684 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3685 changes of various kinds. As a consequence, a big documentation update was
3686 needed. This delayed the release for rather longer than seemed good, especially
3687 in the light of a couple of (minor) security issues. Therefore, the changes
3688 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3689 release. So 4.44 and 4.50 are in effect two different branches that both start
3692 I have left the 4.50 change log unchanged; it contains all the changes since
3693 4.43. The change log for 4.44 is below; many of its items are identical to
3694 those for 4.50. This seems to be the most sensible way to preserve the
3695 historical information.
3701 1. Minor wording change to the doc/README.SIEVE file.
3703 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3704 computation of the current number of files was incorrect.
3706 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3707 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3710 4. Give more explanation in the error message when the command for a transport
3711 filter fails to execute.
3713 5. There are several places where Exim runs a non-Exim command in a
3714 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3715 was being done only for the command run by the queryprogram router. It is
3716 now done for all such subprocesses. The other cases are: ${run, transport
3717 filters, and the commands run by the lmtp and pipe transports.
3719 6. Added CONFIGURE_GROUP build-time option.
3721 7. Some older OS have a limit of 256 on the maximum number of file
3722 descriptors. Exim was using setrlimit() to set 1000 as a large value
3723 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3724 systems. I've change it so that if it can't get 1000, it tries for 256.
3726 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3727 was an oversight, and furthermore, ever since the addition of extra
3728 controls (e.g. 4.43/32), the checks on when to allow different forms of
3729 "control" were broken. There should now be diagnostics for all cases when a
3730 control that does not make sense is encountered.
3732 9. Added the /retain_sender option to "control=submission".
3734 10. $recipients is now available in the predata ACL (oversight).
3736 11. Tidy the search cache before the fork to do a delivery from a message
3737 received from the command line. Otherwise the child will trigger a lookup
3738 failure and thereby defer the delivery if it tries to use (for example) a
3739 cached ldap connection that the parent has called unbind on.
3741 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3742 of $address_data from the recipient verification was clobbered by the
3743 sender verification.
3745 13. The value of address_data from a sender verification is now available in
3746 $sender_address_data in subsequent conditions in the ACL statement.
3748 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3750 15. Added a new option "connect=<time>" to callout options, to set a different
3753 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3754 was its contents. (It was OK if the option was not defined at all.)
3756 17. A "Completed" log line is now written for messages that are removed from
3757 the spool by the -Mrm option.
3759 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3760 information about exactly what failed.
3762 19. Added -dd to debug only the daemon process.
3764 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3765 handles timeouts, both on the server side and network timeouts. Renamed the
3766 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3768 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3769 It is now set to "smtps".
3771 22. $host_address is now set to the target address during the checking of
3772 ignore_target_hosts.
3774 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3775 being passed; this would have caused $sender_host_name to have been used if
3776 matching the list had actually called for a host name (not very likely,
3777 since this list is usually IP addresses). A host name is now passed as
3780 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3781 code that specifies a non-restarting handler (typically sigaction() in
3782 modern systems) in an attempt to fix a rare and obscure crash bug.
3784 25. Narrowed the window for a race in the daemon that could cause it to ignore
3785 SIGCHLD signals. This is not a major problem, because they are used only to
3786 wake it up if nothing else does.
3788 26. A malformed maildirsize file could cause Exim to calculate negative values
3789 for the mailbox size or file count. Odd effects could occur as a result.
3790 The maildirsize information is now recalculated if the size or filecount
3793 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3794 support for a long time. Removed HAVE_SYS_VFS_H.
3796 28. Installed the latest version of exipick from John Jetmore.
3798 29. In an address list, if the pattern was not a regular expression, an empty
3799 subject address (from a bounce message) matched only if the pattern was an
3800 empty string. Non-empty patterns were not even tested. This was the wrong
3801 because it is perfectly reasonable to use an empty address as part of a
3802 database query. An empty address is now tested by patterns that are
3803 lookups. However, all the other forms of pattern expect the subject to
3804 contain a local part and a domain, and therefore, for them, an empty
3805 address still always fails if the pattern is not itself empty.
3807 30. Exim went into a mad DNS loop when attempting to do a callout where the
3808 host was specified on an smtp transport, and looking it up yielded more
3809 than one IP address.
3811 31. Re-factored the code for checking spool and log partition space into a
3812 function that finds that data and another that does the check. The former
3813 is then used to implement four new variables: $spool_space, $log_space,
3814 $spool_inodes, and $log_inodes.
3816 32. The RFC2047 encoding function was originally intended for short strings
3817 such as real names; it was not keeping to the 75-character limit for
3818 encoded words that the RFC imposes. It now respects the limit, and
3819 generates multiple encoded words if necessary. To be on the safe side, I
3820 have increased the buffer size for the ${rfc2047: expansion operator from
3823 33. It is now permitted to omit both strings after an "if" condition; if the
3824 condition is true, the result is "true". As before, when the second string
3825 is omitted, a false condition yields an empty string. This makes it less
3826 cumbersome to write custom ACL and router conditions.
3828 34. Failure to deliver a bounce message always caused it to be frozen, even if
3829 there was an errors_to setting on the router. The errors_to setting is now
3832 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3833 canonical form (fully expanded) before being placed in
3834 $sender_host_address.
3836 36. The table in the code that translates DNS record types into text (T_A to
3837 "A" for instance) was missing entries for NS and CNAME. It is just possible
3838 that this could have caused confusion if both these types were looked up
3839 for the same domain, because the text type is used as part of Exim's
3840 per-process caching. But the chance of anyone hitting this buglet seems
3843 37. The dnsdb lookup has been extended in a number of ways.
3845 (1) There is a new type, "zns", which walks up the domain tree until it
3846 finds some nameserver records. It should be used with care.
3848 (2) There is a new type, "mxh", which is like "mx" except that it returns
3849 just the host names, not the priorities.
3851 (3) It is now possible to give a list of domains (or IP addresses) to be
3852 looked up. The behaviour when one of the lookups defers can be
3853 controlled by a keyword.
3855 (4) It is now possible to specify the separator character for use when
3856 multiple records are returned.
3858 38. The dnslists ACL condition has been extended: it is now possible to supply
3859 a list of IP addresses and/or domains to be looked up in a particular DNS
3862 39. Added log_selector=+queue_time_overall.
3864 40. When running the queue in the test harness, wait just a tad after forking a
3865 delivery process, to get repeatability of debugging output.
3867 41. Include certificate and key file names in error message when GnuTLS fails
3868 to set them up, because the GnuTLS error message doesn't include the name
3869 of the failing file when there is a problem reading it.
3871 42. Allow both -bf and -bF in the same test run.
3873 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3875 44. The "Exiscan patch" is now merged into the mainline Exim source.
3877 45. Sometimes the final signoff response after QUIT could fail to get
3878 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3879 < 0 before doing a fflush(). This bug looks as though it goes back to the
3880 introduction of TLS in release 3.20, but "sometimes" must have been rare
3881 because the tests only now provoked it.
3883 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3884 (this can affect the format of dates).
3886 47. exim_tidydb, when checking for the continued existence of a message for
3887 which it has found a message-specific retry record, was not finding
3888 messages that were in split spool directories. Consequently, it was
3889 deleting retry records that should have stayed in existence.
3891 48. Steve fixed some bugs in eximstats.
3893 49. The SPA authentication driver was not abandoning authentication and moving
3894 on to the next authenticator when an expansion was forced to fail,
3895 contradicting the general specification for all authenticators. Instead it
3896 was generating a temporary error. It now behaves as specified.
3898 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3899 (the order specifies the preference for clients). The order is now AES256,
3900 AES128, 3DES, ARCFOUR128.
3902 51. Small patch to Sieve code - explicitly set From: when generating an
3905 52. Exim crashed if a remote delivery caused a very long error message to be
3906 recorded - for instance if somebody sent an entire SpamAssassin report back
3907 as a large number of 550 error lines. This bug was coincidentally fixed by
3908 increasing the size of one of Exim's internal buffers (big_buffer) that
3909 happened as part of the Exiscan merge. However, to be on the safe side, I
3910 have made the code more robust (and fixed the comments that describe what
3913 53. Now that there can be additional text after "Completed" in log lines (if
3914 the queue_time_overall log selector is set), a one-byte patch to exigrep
3915 was needed to allow it to recognize "Completed" as not the last thing in
3918 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3919 patch that reportedly fixes this has been added. I am not expert enough to
3920 create a test for it. This is what the patch creator wrote:
3922 "I found a little strange behaviour of ldap code when working with
3923 Windows 2003 AD Domain, where users was placed in more than one
3924 Organization Units. When I tried to give exim partial DN, the exit code
3925 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3926 But simultaneously result of request was absolutely normal ldap result,
3927 so I produce this patch..."
3929 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3930 REFERENCE, so I have modified the code to exclude the patch when that macro
3933 55. Some experimental protocols are using DNS PTR records for new purposes. The
3934 keys for these records are domain names, not reversed IP addresses. The
3935 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3936 leaves it alone. Component reversal etc. now happens only for IP addresses.
3939 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3941 57. Double the size of the debug message buffer (to 2048) so that more of very
3942 long debug lines gets shown.
3944 58. The exicyclog utility now does better if the number of log files to keep
3945 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3947 59. Two changes related to the smtp_active_hostname option:
3949 (1) $smtp_active_hostname is now available as a variable.
3950 (2) The default for smtp_banner uses $smtp_active_hostname instead
3951 of $primary_hostname.
3953 60. The host_aton() function is supposed to be passed a string that is known
3954 to be a valid IP address. However, in the case of IPv6 addresses, it was
3955 not checking this. This is a hostage to fortune. Exim now panics and dies
3956 if the condition is not met. A case was found where this could be provoked
3957 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3958 components; fortuitously, this particular loophole had already been fixed
3959 by change 4.50/55 above.
3961 If there are any other similar loopholes, the new check in host_aton()
3962 itself should stop them being exploited. The report I received stated that
3963 data on the command line could provoke the exploit when Exim was running as
3964 exim, but did not say which command line option was involved. All I could
3965 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3966 running as the user.
3969 61. There was a buffer overflow vulnerability in the SPA authentication code
3970 (which came originally from the Samba project). I have added a test to the
3971 spa_base64_to_bits() function which I hope fixes it.
3974 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3975 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3977 63. The daemon start-up calls getloadavg() while still root for those OS that
3978 need the first call to be done as root, but it missed one case: when
3979 deliver_queue_load_max is set with deliver_drop_privilege. This is
3980 necessary for the benefit of the queue runner, because there is no re-exec
3981 when deliver_drop_privilege is set.
3983 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3984 This has been fixed.
3986 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3987 was in use, was not putting the data itself into the right store pool;
3988 consequently, it could be overwritten for a subsequent message in the same
3989 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3992 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3994 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3995 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3996 the function were treating the return as a boolean value, which happened to
3997 work because 0=false and not-0=true, but is not correct code.
3999 68. The host_aton() function was not handling scoped IPv6 addresses (those
4000 with, for example, "%eth0" on the end) correctly.
4002 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4003 negated items (that is, ~something) in unsigned ints. Some compilers
4004 apparently mutter when there is no cast.
4006 70. If an address verification called from an ACL failed, and did not produce a
4007 user-specific message (i.e. there was only a "system" message), nothing was
4008 put in $acl_verify_message. In this situation, it now puts the system
4011 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4012 session; change 4.31/43 added the unwanted input to the log line - except
4013 that it did not do this in the start of session case. It now does.
4015 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4016 This is wrong and can cause the other end to generate a synchronization
4017 error if it is another Exim or anything else that does the synchronization
4018 check. A QUIT command is no longer sent after a timeout.
4020 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4021 during host lookups.
4023 74. The defer_ok option of callout verification was not working if it was used
4024 when verifying addresses in header lines, that is, for this case:
4026 verify = header_sender/callout=defer_ok
4028 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4029 those file descriptors could be used for SMTP connections. If anything
4030 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4031 could be sent to the SMTP client, causing chaos. The daemon now opens
4032 stdin, stdout, and stderr to /dev/null when it puts itself into the
4035 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4036 log by default. The user can override this with suitable Perl magic.
4038 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4039 the log message when discard triggers, was not working for the DATA ACL or
4040 for the non-SMTP ACL.
4042 78. Error message wording change in sieve.c.
4044 79. If smtp_accept_max_per_host was set, the number of connections could be
4045 restricted to fewer than expected, because the daemon was trying to set up
4046 a new connection before checking whether the processes handling previous
4047 connections had finished. The check for completed processes is now done
4048 earlier. On busy systems, this bug wouldn't be noticed because something
4049 else would have woken the daemon, and it would have reaped the completed
4052 80. If a message was submitted locally by a user whose login name contained one
4053 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4054 It caused a spool format error. I have fixed the spool reading code. A
4055 related problem was that the "from" clause in the Received: line became
4056 illegal because of the space(s). It is now covered by ${quote_local_part.
4058 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4061 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4063 83. Patch from Sieve maintainer to fix unterminated string problem in
4064 "vacation" handling.
4066 84. Some minor changes to the Linux configuration files to help with other
4067 OS variants using glibc.
4069 85. One more patch for Sieve to update vacation handling to latest spec.
4072 ----------------------------------------------------
4073 See the note above about the 4.44 and 4.50 releases.
4074 ----------------------------------------------------
4080 1. Change 4.43/35 introduced a bug that caused file counts to be
4081 incorrectly computed when quota_filecount was set in an appendfile
4084 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4085 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4088 3. Give more explanation in the error message when the command for a transport
4089 filter fails to execute.
4091 4. There are several places where Exim runs a non-Exim command in a
4092 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4093 was being done only for the command run by the queryprogram router. It is
4094 now done for all such subprocesses. The other cases are: ${run, transport
4095 filters, and the commands run by the lmtp and pipe transports.
4097 5. Some older OS have a limit of 256 on the maximum number of file
4098 descriptors. Exim was using setrlimit() to set 1000 as a large value
4099 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4100 systems. I've change it so that if it can't get 1000, it tries for 256.
4102 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4103 was an oversight, and furthermore, ever since the addition of extra
4104 controls (e.g. 4.43/32), the checks on when to allow different forms of
4105 "control" were broken. There should now be diagnostics for all cases when a
4106 control that does not make sense is encountered.
4108 7. $recipients is now available in the predata ACL (oversight).
4110 8. Tidy the search cache before the fork to do a delivery from a message
4111 received from the command line. Otherwise the child will trigger a lookup
4112 failure and thereby defer the delivery if it tries to use (for example) a
4113 cached ldap connection that the parent has called unbind on.
4115 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4116 of $address_data from the recipient verification was clobbered by the
4117 sender verification.
4119 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4120 was its contents. (It was OK if the option was not defined at all.)
4122 11. A "Completed" log line is now written for messages that are removed from
4123 the spool by the -Mrm option.
4125 12. $host_address is now set to the target address during the checking of
4126 ignore_target_hosts.
4128 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4129 being passed; this would have caused $sender_host_name to have been used if
4130 matching the list had actually called for a host name (not very likely,
4131 since this list is usually IP addresses). A host name is now passed as
4134 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4135 code that specifies a non-restarting handler (typically sigaction() in
4136 modern systems) in an attempt to fix a rare and obscure crash bug.
4138 15. Narrowed the window for a race in the daemon that could cause it to ignore
4139 SIGCHLD signals. This is not a major problem, because they are used only to
4140 wake it up if nothing else does.
4142 16. A malformed maildirsize file could cause Exim to calculate negative values
4143 for the mailbox size or file count. Odd effects could occur as a result.
4144 The maildirsize information is now recalculated if the size or filecount
4147 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4148 support for a long time. Removed HAVE_SYS_VFS_H.
4150 18. Updated exipick to current release from John Jetmore.
4152 19. Allow an empty sender to be matched against a lookup in an address list.
4153 Previously the only cases considered were a regular expression, or an
4156 20. Exim went into a mad DNS lookup loop when doing a callout where the
4157 host was specified on the transport, if the DNS lookup yielded more than
4160 21. The RFC2047 encoding function was originally intended for short strings
4161 such as real names; it was not keeping to the 75-character limit for
4162 encoded words that the RFC imposes. It now respects the limit, and
4163 generates multiple encoded words if necessary. To be on the safe side, I
4164 have increased the buffer size for the ${rfc2047: expansion operator from
4167 22. Failure to deliver a bounce message always caused it to be frozen, even if
4168 there was an errors_to setting on the router. The errors_to setting is now
4171 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4172 canonical form (fully expanded) before being placed in
4173 $sender_host_address.
4175 24. Updated eximstats to version 1.33
4177 25. Include certificate and key file names in error message when GnuTLS fails
4178 to set them up, because the GnuTLS error message doesn't include the name
4179 of the failing file when there is a problem reading it.
4181 26. Expand error message when OpenSSL has problems setting up cert/key files.
4184 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4185 (this can affect the format of dates).
4187 28. exim_tidydb, when checking for the continued existence of a message for
4188 which it has found a message-specific retry record, was not finding
4189 messages that were in split spool directories. Consequently, it was
4190 deleting retry records that should have stayed in existence.
4192 29. eximstats updated to version 1.35
4193 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4194 1.35 - bugfix such that pie charts by volume are generated correctly
4196 30. The SPA authentication driver was not abandoning authentication and moving
4197 on to the next authenticator when an expansion was forced to fail,
4198 contradicting the general specification for all authenticators. Instead it
4199 was generating a temporary error. It now behaves as specified.
4201 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4202 (the order specifies the preference for clients). The order is now AES256,
4203 AES128, 3DES, ARCFOUR128.
4205 31. Small patch to Sieve code - explicitly set From: when generating an
4208 32. Exim crashed if a remote delivery caused a very long error message to be
4209 recorded - for instance if somebody sent an entire SpamAssassin report back
4210 as a large number of 550 error lines. This bug was coincidentally fixed by
4211 increasing the size of one of Exim's internal buffers (big_buffer) that
4212 happened as part of the Exiscan merge. However, to be on the safe side, I
4213 have made the code more robust (and fixed the comments that describe what
4216 33. Some experimental protocols are using DNS PTR records for new purposes. The
4217 keys for these records are domain names, not reversed IP addresses. The
4218 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4219 leaves it alone. Component reversal etc. now happens only for IP addresses.
4222 34. The host_aton() function is supposed to be passed a string that is known
4223 to be a valid IP address. However, in the case of IPv6 addresses, it was
4224 not checking this. This is a hostage to fortune. Exim now panics and dies
4225 if the condition is not met. A case was found where this could be provoked
4226 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4227 components; fortuitously, this particular loophole had already been fixed
4228 by change 4.50/55 or 4.44/33 above.
4230 If there are any other similar loopholes, the new check in host_aton()
4231 itself should stop them being exploited. The report I received stated that
4232 data on the command line could provoke the exploit when Exim was running as
4233 exim, but did not say which command line option was involved. All I could
4234 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4235 running as the user.
4238 35. There was a buffer overflow vulnerability in the SPA authentication code
4239 (which came originally from the Samba project). I have added a test to the
4240 spa_base64_to_bits() function which I hope fixes it.
4243 36. The daemon start-up calls getloadavg() while still root for those OS that
4244 need the first call to be done as root, but it missed one case: when
4245 deliver_queue_load_max is set with deliver_drop_privilege. This is
4246 necessary for the benefit of the queue runner, because there is no re-exec
4247 when deliver_drop_privilege is set.
4249 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4250 was in use, was not putting the data itself into the right store pool;
4251 consequently, it could be overwritten for a subsequent message in the same
4252 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4255 38. Sometimes the final signoff response after QUIT could fail to get
4256 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4257 < 0 before doing a fflush(). This bug looks as though it goes back to the
4258 introduction of TLS in release 3.20, but "sometimes" must have been rare
4259 because the tests only now provoked it.
4265 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4266 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4267 Now it returns an integer. A number of calls were still expecting a T/F
4268 return. Fortuitously, in all cases, the tests worked in OK situations,
4269 which is the norm. However, things would have gone wrong on any write
4270 failures on the smtp file descriptor. This function is used when sending
4271 messages over SMTP and also when doing verify callouts.
4273 2. When Exim is called to do synchronous delivery of a locally submitted
4274 message (the -odf or -odi options), it no longer closes stderr before doing
4277 3. Implemented the mua_wrapper option.
4279 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4281 5. Implemented the functions header_remove(), header_testname(),
4282 header_add_at_position(), and receive_remove_recipient(), and exported them
4285 6. If an ACL "warn" statement specified the addition of headers, Exim already
4286 inserted X-ACL-Warn: at the start if there was no header name. However, it
4287 was not making this test for the second and subsequent header lines if
4288 there were newlines in the string. This meant that an invalid header could
4289 be inserted if Exim was badly configured.
4291 7. Allow an ACL "warn" statement to add header lines at the start or after all
4292 the Received: headers, as well as at the end.
4294 8. Added the rcpt_4xx retry error code.
4296 9. Added postmaster_mailfrom=xxx to callout verification option.
4298 10. Added mailfrom=xxxx to the callout verification option, for verify=
4301 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4302 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4304 12. Inserted some casts to stop certain compilers warning when using pointer
4305 differences as field lengths or precisions in printf-type calls (mostly
4306 affecting debugging statements).
4308 13. Added optional readline() support for -be (dynamically loaded).
4310 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4311 same clock tick as a message's arrival, so that its received time was the
4312 same as the "first fail" time on the retry record, and that message
4313 remained on the queue past the ultimate address timeout, every queue runner
4314 would try a delivery (because it was past the ultimate address timeout) but
4315 after another failure, the ultimate address timeout, which should have then
4316 bounced the address, did not kick in. This was a "< instead of <=" error;
4317 in most cases the first failure would have been in the next clock tick
4318 after the received time, and all would be well.
4320 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4321 being recognized when the domain list was tested by the match_domain
4322 condition in an expansion string.
4324 16. Added the ${str2b64: operator.
4326 17. Exim was always calling setrlimit() to set a large limit for the number of
4327 processes, without checking whether the existing limit was already
4328 adequate. (It did check for the limit on file descriptors.) Furthermore,
4329 errors from getrlimit() and setrlimit() were being ignored. Now they are
4330 logged to the main and panic logs, but Exim does carry on, to try to do its
4331 job under whatever limits there are.
4333 18. Imported PCRE 5.0.
4335 19. Trivial typo in log message " temporarily refused connection" (the leading
4338 20. If the log selector return_path_on_delivery was set and an address was
4339 redirected to /dev/null, the delivery process crashed because it assumed
4340 that a return path would always be set for a "successful" delivery. In this
4341 case, the whole delivery is bypassed as an optimization, and therefore no
4344 21. Internal re-arrangement: the function for sending a challenge and reading
4345 a response while authentication was assuming a zero-terminated challenge
4346 string. It's now changed to take a pointer and a length, to allow for
4347 binary data in such strings.
4349 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4351 23. Exim was not respecting finduser_retries when seeking the login of the
4352 uid under which it was called; it was always trying 10 times. (The default
4353 setting of finduser_retries is zero.) Also, it was sleeping after the final
4354 failure, which is pointless.
4356 24. Implemented tls_on_connect_ports.
4358 25. Implemented acl_smtp_predata.
4360 26. If the domain in control=submission is set empty, Exim assumes that the
4361 authenticated id is a complete email address when it generates From: or
4362 Sender: header lines.
4364 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4365 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4366 chown and chgrp in /bin and hostname in /usr/bin.
4368 28. Exim was keeping the "process log" file open after each use, just as it
4369 does for the main log. This opens the possibility of it remaining open for
4370 long periods when the USR1 signal hits a daemon. Occasional processlog
4371 errors were reported, that could have been caused by this. Anyway, it seems
4372 much more sensible not to leave this file open at all, so that is what now
4375 29. The long-running daemon process does not normally write to the log once it
4376 has entered its main loop, and it closes the log before doing so. This is
4377 so that log files can straightforwardly be renamed and moved. However,
4378 there are a couple of unusual error situations where the daemon does write
4379 log entries, and I had neglected to close the log afterwards.
4381 30. The text of an SMTP error response that was received during a remote
4382 delivery was being truncated at 512 bytes. This is too short for some of
4383 the long messages that one sometimes sees. I've increased the limit to
4386 31. It is now possible to make retry rules that apply only when a message has a
4387 specific sender, in particular, an empty sender.
4389 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4390 it possible to be selective about when SMTP synchronization is enforced.
4392 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4394 32. Implemented hosts_connection_nolog.
4396 33. Added an ACL for QUIT.
4398 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4401 35. Added mailbox_size and mailbox_filecount to appendfile.
4403 36. Added control = no_multiline_responses to ACLs.
4405 37. There was a bug in the logic of the code that waits for the clock to tick
4406 in the case where the clock went backwards by a substantial amount such
4407 that the microsecond fraction of "now" was more than the microsecond
4408 fraction of "then" (but the whole seconds number was less).
4410 38. Added support for the libradius Radius client library this is found on
4411 FreeBSD (previously only the radiusclient library was supported).
4417 1. When certain lookups returned multiple values in the form name=value, the
4418 quoting of the values was not always being done properly. Specifically:
4419 (a) If the value started with a double quote, but contained no whitespace,
4421 (b) If the value contained whitespace other than a space character (i.e.
4422 tabs or newlines or carriage returns) it was not quoted.
4423 This fix has been applied to the mysql and pgsql lookups by writing a
4424 separate quoting function and calling it from the lookup code. The fix
4425 should probably also be applied to nisplus, ibase and oracle lookups, but
4426 since I cannot test any of those, I have not disturbed their existing code.
4428 2. A hit in the callout cache for a specific address caused a log line with no
4429 reason for rejecting RCPT. Now it says "Previous (cached) callout
4430 verification failure".
4432 3. There was an off-by-one bug in the queryprogram router. An over-long
4433 return line was truncated at 256 instead of 255 characters, thereby
4434 overflowing its buffer with the terminating zero. As well as fixing this, I
4435 have increased the buffer size to 1024 (and made a note to document this).
4437 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4438 when Exim is waiting for an SMTP response from a remote server, Exim
4439 restarts its select() call on the socket, thereby resetting its timeout.
4440 This is not a problem when such interrupts are rare. Somebody set up a cron
4441 job to run exiwhat every 2 minutes, which is less than the normal select()
4442 timeout (5 or 10 minutes). This meant that the select() timeout never
4443 kicked in because it was always reset. I have fixed this by comparing the
4444 time when an interrupt arrives with the time at the start of the first call
4445 to select(). If more time than the timeout has elapsed, the interrupt is
4446 treated as a timeout.
4448 5. Some internal re-factoring in preparation for the addition of Sieve
4449 extensions (by MH). In particular, the "personal" test is moved to a
4450 separate function, and given an option for scanning Cc: and Bcc: (which is
4451 not set for Exim filters).
4453 6. When Exim created an email address using the login of the caller as the
4454 local part (e.g. when creating a From: or Sender: header line), it was not
4455 quoting the local part when it contained special characters such as @.
4457 7. Installed new OpenBSD configuration files.
4459 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4460 try to make them clearer.
4462 9. Callout options, other than the timeout value, were being ignored when
4463 verifying sender addresses in header lines. For example, when using
4465 verify = header_sender/callout=no_cache
4467 the cache was (incorrectly) being used.
4469 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4470 only the Cygwin environment.
4472 11. When return_path_on_delivery was set as a log selector, if different remote
4473 addresses in the same message used different return paths and parallel
4474 remote delivery occurred, the wrong values would sometimes be logged.
4475 (Whenever a remote delivery process finished, the return path value from
4476 the most recently started remote delivery process was logged.)
4478 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4479 lines when AUTH and/or TLS are in use. This is the "received protocol"
4480 field. Exim used to use "asmtp" for authenticated SMTP, without any
4481 indication (in the protocol name) for TLS use. Now it follows the RFC and
4482 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4483 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4484 names appear in log lines as well as in Received: header lines.
4486 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4487 extensions, and comparison tests, and to fix some bugs.
4489 14. Changes to the "personal" filter test:
4491 (1) The test was buggy in that it was just doing the equivalent of
4492 "contains" tests on header lines. For example, if a user's address was
4493 anne@some.where, the "personal" test would incorrectly be true for
4495 To: susanne@some.where
4497 This test is now done by extracting each address from the header in turn,
4498 and checking the entire address. Other tests that are part of "personal"
4499 are now done using regular expressions (for example, to check local parts
4500 of addresses in From: header lines).
4502 (2) The list of non-personal local parts in From: addresses has been
4503 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4504 taken from the Sieve specification recommendations.
4506 (3) If the message contains any header line starting with "List-" it is
4507 treated as non-personal.
4509 (4) The test for "circular" in the Subject: header line has been removed
4510 because it now seems ill-conceived.
4512 15. Minor typos in src/EDITME comments corrected.
4514 16. Installed latest exipick from John Jetmore.
4516 17. If headers_add on a router specified a text string that was too long for
4517 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4518 of string_sprintf() is now avoided.
4520 18. $message_body_size was not set (it was always zero) when running the DATA
4521 ACL and the local_scan() function.
4523 19. For the "mail" command in an Exim filter, no default was being set for
4524 the once_repeat time, causing a random time value to be used if "once" was
4525 specified. (If the value happened to be <= 0, no repeat happened.) The
4526 default is now 0s, meaning "never repeat". The "vacation" command was OK
4527 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4528 (I found it when inspecting the code).
4530 20. There is now an overall timeout for performing a callout verification. It
4531 defaults to 4 times the callout timeout, which applies to individual SMTP
4532 commands during the callout. The overall timeout applies when there is more
4533 than one host that can be tried. The timeout is checked before trying the
4534 next host. This prevents very long delays if there are a large number of
4535 hosts and all are timing out (e.g. when the network connections are timing
4536 out). The value of the overall timeout can be changed by specifying an
4537 additional sub-option for "callout", called "maxwait". For example:
4539 verify = sender/callout=5s,maxwait=20s
4541 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4542 to the end before writing, but this should make it even safer).
4544 22. Exim was forgetting that it had advertised PIPELINING for the second and
4545 subsequent messages on an SMTP connection. It was also not resetting its
4546 memory on STARTTLS and an internal HELO.
4548 23. When Exim logs an SMTP synchronization error within a session, it now
4549 records whether PIPELINING has been advertised or not.
4551 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4552 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4553 rather than long int.
4555 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4557 26. Added the never_mail option to autoreply.
4563 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4564 crash if the getsockname() call failed; this can happen if a connection is
4565 closed very soon after it is established. The problem was simply in the
4566 order in which certain operations were done, causing Exim to try to write
4567 to the SMTP stream before it had set up the file descriptor. The bug has
4568 been fixed by making things happen in the correct order.
4574 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4575 before the connection was closed, thus losing the rejection response.
4577 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4578 some early Solaris releases, but causes trouble in current releases where
4579 socklen_t is defined.
4581 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4584 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4587 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4588 that releasing the top of it at the end releases what was used for sub-
4589 expansions (unless the block got too big). However, discard this block if
4590 the first thing is a variable or header, so that we can use its block when
4591 it is dynamic (useful for very large $message_headers, for example).
4593 6. Lookups now cache *every* query, not just the most recent. A new, separate
4594 store pool is used for this. It can be recovered when all lookup caches are
4595 flushed. Lookups now release memory at the end of their result strings.
4596 This has involved some general refactoring of the lookup sources.
4598 7. Some code has been added to the store_xxx() functions to reduce the amount
4599 of flapping under certain conditions.
4601 8. log_incoming_interface used to affect only the <= reception log lines. Now
4602 it causes the local interface and port to be added to several more SMTP log
4603 lines, for example "SMTP connection from", and rejection lines.
4605 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4607 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4609 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4610 could be overwritten at the end of the current message (or the start of a
4611 new message if it was set in a HELO ACL). The value is now preserved for
4612 the duration of the SMTP connection.
4614 12. If a transport had a headers_rewrite setting, and a matching header line
4615 contained an unqualified address, that address was qualified, even if it
4616 did not match any rewriting rules. The underlying bug was that the values
4617 of the flags that permit the existence of unqualified sender and recipient
4618 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4619 non-local messages, and by -bnq for local messages) were not being
4620 preserved with the message after it was received.
4622 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4623 "next input=" as part of the text comprising the host identity instead of
4624 the correct text. The code was using the same buffer for two different
4625 strings. However, depending on which order the printing function evaluated
4626 its arguments, the bug did not always show up. Under Linux, for example, my
4627 test suite worked just fine.
4629 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4630 doesn't work with some older versions of Perl. It has been changed to "my",
4631 which in any case is probably the better facility to use.
4633 15. A really picky compiler found some instances of statements for creating
4634 error messages that either had too many or two few arguments for the format
4637 16. The size of the buffer for calls to the DNS resolver has been increased
4638 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4639 for addresses that have a lot of PTR records. This alleviates a problem; it
4640 does not fully solve it.
4642 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4643 buffer now truncates the list and logs the incident, which is the same
4644 action as happens when Exim is looking up a host name and its aliases.
4645 Previously in this situation something unpredictable would happen;
4646 sometimes it was "internal error: store_reset failed".
4648 18. If a server dropped the connection unexpectedly when an Exim client was
4649 using GnuTLS and trying to read a response, the client delivery process
4650 crashed while trying to generate an error log message.
4652 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4653 string, for example:
4655 warn message = H1: something\nH2: something
4657 the text was added as a single header line from Exim's point of view
4658 though it ended up OK in the delivered message. However, searching for the
4659 second and subsequent header lines using $h_h2: did not work. This has been
4660 fixed. Similarly, if a system filter added multiple headers in this way,
4661 the routers could not see them.
4663 20. Expanded the error message when iplsearch is called with an invalid key to
4664 suggest using net-iplsearch in a host list.
4666 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4667 ACLs are no longer actually imposed (and a message to that effect is
4670 22. If a "gecos" field in a passwd entry contained escaped characters, in
4671 particular, if it contained a \" sequence, Exim got it wrong when building
4672 a From: or a Sender: header from that name. A second bug also caused
4673 incorrect handling when an unquoted " was present following a character
4674 that needed quoting.
4676 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4677 was not being matched caselessly.
4679 24. Arranged for all hyphens in the exim.8 source to be escaped with
4682 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4683 properly. Recipient callout cache records were still being keyed to include
4684 the sender, even when use_sender was set false. This led to far more
4685 callouts that were necessary. The sender is no longer included in the key
4686 when use_sender is false.
4688 26. Added "control = submission" modifier to ACLs.
4690 27. Added the ${base62d: operator to decode base 62 numbers.
4692 28. dnsdb lookups can now access SRV records.
4694 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4695 the configuration file.
4697 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4698 (-v) mode. This makes the output for a verbose queue run more intelligible.
4700 31. Added a use_postmaster feature to recipient callouts.
4702 32. Added the $body_zerocount variable, containing the number of binary zero
4703 bytes in the message body.
4705 33. The time of last modification of the "new" subdirectory is now used as the
4706 "mailbox time last read" when there is a quota error for a maildir
4709 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4711 35. Added +ignore_unknown as a special item in host lists.
4713 36. Code for decoding IPv6 addresses in host lists is now included, even if
4714 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4715 address was recognized as an IP address, but was then not correctly decoded
4716 into binary, causing unexpected and incorrect effects when compared with
4723 1. Very minor rewording of debugging text in manualroute to say "list of
4724 hosts" instead of "hostlist".
4726 2. If verify=header_syntax was set, and a header line with an unqualified
4727 address (no domain) and a large number of spaces between the end of the
4728 name and the colon was received, the reception process suffered a buffer
4729 overflow, and (when I tested it) crashed. This was caused by some obsolete
4730 code that should have been removed. The fix is to remove it!
4732 3. When running in the test harness, delay a bit after writing a bounce
4733 message to get a bit more predictability in the log output.
4735 4. Added a call to search_tidyup() just before forking a reception process. In
4736 theory, someone could use a lookup in the expansion of smtp_accept_max_
4737 per_host which, without the tidyup, could leave open a database connection.
4739 5. Added the variables $recipient_data and $sender_data which get set from a
4740 lookup success in an ACL "recipients" or "senders" condition, or a router
4741 "senders" option, similar to $domain_data and $local_part_data.
4743 6. Moved the writing of debug_print from before to after the "senders" test
4746 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4747 problems for message scanning, either using a data ACL, or using
4748 local_scan() because the Received: header was not generated till after they
4749 were called (in order to set the time as the time of reception completion).
4750 I have revised the way this works. The header is now generated after the
4751 body is received, but before the ACL or local_scan() are called. After they
4752 are run, the timestamp in the header is updated.
4758 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4759 before starting a queue runner without re-exec. This happened only when
4760 deliver_drop_privilege was set or when the Exim user was set to root. The
4761 effect of the bug was that timeouts during subsequent deliveries caused
4762 crashes instead of being properly handled. The handler is now left at its
4763 default (and expected) setting.
4765 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4766 message, again when deliver_drop_privilege is set or Exim is run as root.
4767 The bug described in (1) was not present in this case, but the tidying up
4768 of the other signals was missing. I have made the two cases consistent.
4770 3. The ignore_target_hosts setting on a manualroute router was being ignored
4771 for hosts that were looked up using the /MX notation.
4773 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4776 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4777 operated on the sender address. After changing the $sender_address to <>
4778 for the sender address verify, Exim was re-instated it as the original
4779 (before rewriting) address, but remembering that it had rewritten it, so it
4780 wasn't rewriting it again. This bug also had the effect of breaking the
4781 sender address verification caching when the sender address was rewritten.
4783 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4784 This has been changed so that if the ip literal address matches
4785 ignore_target_hosts, the router declines.
4787 7. Added expansion conditions match_domain, match_address, and match_local_
4788 part (NOT match_host).
4790 8. The placeholder for the Received: header didn't have a length field set.
4792 9. Added code to Exim itself and to exim_lock to test for a specific race
4793 condition that could lead to file corruption when using MBX delivery. The
4794 issue is with the lockfile that is created in /tmp. If this file is removed
4795 after a process has opened it but before that process has acquired a lock,
4796 there is the potential for a second process to recreate the file and also
4797 acquire a lock. This could lead to two Exim processes writing to the file
4798 at the same time. The added code performs the same test as UW imapd; it
4799 checks after acquiring the lock that its file descriptor still refers to
4800 the same named file.
4802 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4803 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4806 11. Added the smtp_active_hostname option. If used, this will typically be made
4807 to depend on the incoming interface address. Because $interface_address is
4808 not set up until the daemon has forked a reception process, error responses
4809 that can happen earlier (such as "too many connections") no longer contain
4812 12. If an expansion in a condition on a "warn" statement fails because a lookup
4813 defers, the "warn" statement is abandoned, and the next ACL statement is
4814 processed. Previously this caused the whole ACL to be aborted.
4816 13. Added the iplsearch lookup type.
4818 14. Added ident_timeout as a log selector.
4820 15. Added tls_certificate_verified as a log selector.
4822 16. Added a global option tls_require_ciphers (compare the smtp transport
4823 option of the same name). This controls incoming TLS connections.
4825 17. I finally figured out how to make tls_require_ciphers do a similar thing
4826 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4827 before starting the TLS session.
4829 18. Tabs are now shown as \t in -bP output.
4831 19. If the log selector return_path_on_delivery was set, Exim crashed when
4832 bouncing a message because it had too many Received: header lines.
4834 20. If two routers both had headers_remove settings, and the first one included
4835 a superfluous trailing colon, the final name in the first list and the
4836 first name in the second list were incorrectly joined into one item (with a
4837 colon in the middle).
4843 1. Added -C and -D options to the exinext utility, mainly to make it easier
4844 to include in the automated testing, but these could be helpful when
4845 multiple configurations are in use.
4847 2. The exinext utility was not formatting the output nicely when there was
4848 an alternate port involved in the retry record key, nor when there was a
4849 message id as well (for retries that were specific to a specific message
4850 and a specific host). It was also confused by IPv6 addresses, because of
4851 the additional colons they contain. I have fixed the IPv4 problem, and
4852 patched it up to do a reasonable job for IPv6.
4854 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4855 delivery, the log line now contains "pipelined" if PIPELINING was used.
4857 4. An SMTP transport process used to panic and die if the bind() call to set
4858 an explicit outgoing interface failed. This has been changed; it is now
4859 treated in the same way as a connect() failure.
4861 5. A reference to $sender_host_name in the part of a conditional expansion
4862 that was being skipped was still causing a DNS lookup. This no longer
4865 6. The def: expansion condition was not recognizing references to header lines
4866 that used bh_ and bheader_.
4868 7. Added the _cache feature to named lists.
4870 8. The code for checking quota_filecount in the appendfile transport was
4871 allowing one more file than it should have been.
4873 9. For compatibility with Sendmail, the command line option
4881 and sets the incoming protocol and host name (for trusted callers). The
4882 host name and its colon can be omitted when only the protocol is to be set.
4883 Note the Exim already has two private options, -pd and -ps, that refer to
4884 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4885 "s", but I don't think that's a major issue.
4887 10. A number of refactoring changes to the code, none of which should affect
4890 (a) The number of logging options was getting close to filling up the
4891 32-bit word that was used as a bit map. I have split them into two classes:
4892 those that are passed in the argument to log_write(), and those that are
4893 only ever tested independently outside of that function. These are now in
4894 separate 32-bit words, so there is plenty of room for expansion again.
4895 There is no change in the user interface or the logging behaviour.
4897 (b) When building, for example, log lines, the code previously used a
4898 macro that called string_cat() twice, in order to add two strings. This is
4899 not really sufficiently general. Furthermore, there was one instance where
4900 it was actually wrong because one of the argument was used twice, and in
4901 one call a function was used. (As it happened, calling the function twice
4902 did not affect the overall behaviour.) The macro has been replaced by a
4903 function that can join an arbitrary number of extra strings onto a growing
4906 (c) The code for expansion conditions now uses a table and a binary chop
4907 instead of a serial search (which was left over from when there were very
4908 few conditions). Also, it now recognizes conditions like "pam" even when
4909 the relevant support is not compiled in: a suitably worded error message is
4910 given if an attempt is made to use such a condition.
4912 11. Added ${time_interval:xxxxx}.
4914 12. A bug was causing one of the ddress fields not to be passed back correctly
4915 from remote delivery subprocesses. The field in question was not being
4916 subsequently used, so this caused to problems in practice.
4918 13. Added new log selectors queue_time and deliver_time.
4920 14. Might have fixed a bug in maildirsizefile handling that threw up
4921 "unexpected character" debug warnings, and recalculated the data
4922 unnecessarily. In any case, I expanded the warning message to give more
4925 15. Added the message "Restricted characters in address" to the statements in
4926 the default ACL that block characters like @ and % in local parts.
4928 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4929 Three changes have been made:
4931 (a) There was a serious bug; a negative response to MAIL caused the whole
4932 recipient domain to be cached as invalid, thereby blocking all messages
4933 to all local parts at the same domain, from all senders. This bug has
4934 been fixed. The domain is no longer cached after a negative response to
4935 MAIL if the sender used is not empty.
4937 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4940 (c) A new callout option, "use_sender" has been added for people who want
4941 the modified behaviour.
4947 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4950 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4951 indeed breaks things for older releases.
4953 3. Added additional logging to the case where there is a problem reading data
4954 from a filter that is running in a subprocess using a pipe, in order to
4955 try to track down a specific problem.
4957 4. Testing facility fudge: when running in the test harness and attempting
4958 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4959 getting "No route to host". Convert this to a timeout.
4961 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4964 6. Some OS don't have socklen_t but use size_t instead. This affects the
4965 fifth argument of getsockopt() amongst other things. This is now
4966 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4967 can be set for individual OS. I have set it for SunOS5, OSF1, and
4968 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4969 some earlier ones do not.
4971 7. Change 4.30/15 was not doing the test caselessly.
4973 8. The standard form for an IPv6 address literal was being rejected by address
4974 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4975 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4976 this, as well as the form without the "IPv6" on the front (but only when
4977 address literals are enabled, of course).
4979 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4981 10. Exim crashed if a message with an empty sender address specified by -f
4982 encountered a router with an errors_to setting. This could be provoked only
4983 by a command such as
4987 where an empty string was supplied; "<>" did not hit this bug.
4989 11. Installed PCRE release 4.5.
4991 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4992 remained set. It is now erased.
4994 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4995 times from message ids (which are base 36 rather than the normal 62).
4997 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4998 were being counted as actual protocol errors, and logged if the log
4999 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5000 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5001 and DATA following a set of rejected RCPTs do not count as protocol errors.
5002 In other words, Exim assumes they were pipelined, though this may not
5003 actually be the case. Of course, in all cases the client gets an
5004 appropriate error code.
5006 15. If a lookup fails in an ACL condition, a message about the failure may
5007 be available; it is used if testing the ACL cannot continue, because most
5008 such messages specify what the cause of the deferral is. However, some
5009 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5010 that caused an old message to be retained and used if a later statement
5011 caused a defer, replacing the real cause of the deferral.
5013 16. If an IP address had so many PTR records that the DNS lookup buffer
5014 was not large enough to hold them, Exim could crash while trying to process
5015 the truncated data. It now detects and logs this case.
5017 17. Further to 4.21/58, another change has been made: if (and only if) the
5018 first line of a message (the first header line) ends with CRLF, a bare LF
5019 in a subsequent header line has a space inserted after it, so as not to
5020 terminate the header.
5022 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5023 unnecessarily, used atoi() instead of strtol(), and didn't check the
5024 termination when getting file sizes from file names by regex.
5026 19. Completely re-implemented the support for maildirsize files, in the light
5027 of a number of problems with the previous contributed implementation
5028 (4.30/29). In particular:
5030 . If the quota is zero, the maildirsize file is maintained, but no quota is
5033 . If the maildir directory does not exist, it is created before any attempt
5034 to write a maildirsize file.
5036 . The quota value in the file is just a cache; if the quota is changed in
5037 the transport, the new value overrides.
5039 . A regular expression is available for excluding directories from the
5042 20. The autoreply transport checks the characters in options that define the
5043 message's headers; it allows continued headers, but it was checking with
5044 isspace() after an embedded newline instead of explicitly looking for a
5047 21. If all the "regular" hosts to which an address was routed had passed their
5048 expiry times, and had not reached their retry times, the address was
5049 bounced, even if fallback hosts were defined. Now Exim should go on to try
5052 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5053 equivalent code in the SMTP transport. Some hosts send humungous responses
5054 to HELO/EHLO, more than 1024 it seems.
5056 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5057 gives compiler warnings in some environments. I've now done it "properly",
5060 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5061 (because of problems with the built-in one) was declared to return uschar *
5062 instead of char *, causing compiler failure.
5064 25. Fixed a file descriptor leak when processing alias/forward files.
5066 26. Fixed a minor format string issue in dbfn.c.
5068 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5070 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5071 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5074 29. When a sender address is verified, it is cached, to save repeating the test
5075 when there is more than one recipient in a message. However, when the
5076 verification involves a callout, it is possible for different callout
5077 options to be set for different recipients. It is too complicated to keep
5078 track of this in the cache, so now Exim always runs a verification when a
5079 callout is required, relying on the callout cache for the optimization.
5080 The overhead is duplication of the address routing, but this should not be
5083 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5084 to be verified with callout=postmaster, and the main callout worked but the
5085 postmaster check failed, the verification correctly failed. However, if a
5086 subsequent RCPT command asked for sender verification *without* the
5087 postmaster check, incorrect caching caused this verification also to fail,
5090 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5091 it was not caching the DNS options (qualify_single, search_parents) that
5092 were used when the lookup failed. A subsequent lookup with different
5093 options therefore always gave the same answer, though there were cases
5094 where it should not have. (Example: a "domains = !$mx_any" option on a
5095 dnslookup router: the "domains" option is always processed without any
5096 widening, but the router might have qualify_single set.) Now Exim uses the
5097 cached value only when the same options are set.
5099 32. Added John Jetmore's "exipick" utility to the distribution.
5101 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5102 than a timeout (e.g. a certificate is required, and is not provided), an
5103 Exim server now closes the connection immediately. Previously it waited for
5104 the client to close - but if the client is SSL, it seems that they each
5105 wait for each other, leading to a delay before one of them times out.
5107 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5108 maintained 0.8.x compatibility because I don't think many are using it, and
5109 it is clearly obsolete.
5111 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5114 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5115 client certificate was expired. A simple patch fixes this, though I don't
5116 understand the full logic of why the verify callback is called multiple
5119 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5120 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5121 which causes problems with some clients (such as the Certicom SSL Plus
5122 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5123 disables the coutermeasure allowing Eudora to connect."
5125 38. Exim was not checking that a write() to a log file succeeded. This could
5126 lead to Bad Things if a log got too big, in particular if it hit a file
5127 size limit. Exim now panics and dies if it cannot write to a log file, just
5128 as it does if it cannot open a log file.
5130 39. Modified OS/Makefile-Linux so that it now contains
5132 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5134 The two -D definitions ensure that Exim is compiled with large file
5135 support, which makes it possible to handle log files that are bigger than
5138 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5139 instance) a domain was checked against a named list that involved a lookup,
5140 causing $domain_data to be set, then another domain was checked against the
5141 same list, then the first domain was re-checked, the value of $domain_data
5142 after the final check could be wrong. In particular, if the second check
5143 failed, it could be set empty. This bug probably also applied to
5146 41. The strip_trailing_dot option was not being applied to the address given
5147 with the -f command-line option.
5149 42. The code for reading a message's header from the spool was incrementing
5150 $received_count, but never initializing it. This meant that the value was
5151 incorrect (doubled) while delivering a message in the same process in which
5152 it was received. In the most common configuration of Exim, this never
5153 happens - a fresh exec is done - but it can happen when
5154 deliver_drop_privilege is set.
5156 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5157 it now includes up to 150 characters of the unexpected data in the log
5160 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5161 and building data strings. The size of both of these buffers was 10 000
5162 bytes - far larger than anybody would *ever* want, thought I. Needless to
5163 say, somebody hit the limit. I have increased the maximum line length to
5164 20 000 and the maximum data length of concatenated lines to 100 000. I have
5165 also fixed two bugs, because there was no checking on these buffers. Tsk,
5166 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5167 buffer is too small.
5169 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5170 lsearch lookups. Now it does.
5172 46. When parsing a route_list item in a manualroute router, a fixed-length
5173 buffer was used for the list of hosts. I made this 1024 bytes long,
5174 thinking that nobody would ever have a list of hosts that long. Wrong.
5175 Somebody had a whole pile of complicated expansion conditions, and the
5176 string was silently truncated, leading to an expansion error. It turns out
5177 that it is easier to change to an unlimited length (owing to other changes
5178 that have happened since this code was originally written) than to build
5179 structure for giving a limitation error. The length of the item that
5180 expands into the list of hosts is now unlimited.
5182 47. The lsearch lookup could not handle data where the length of text line was
5183 more than 4095 characters. Such lines were truncated, leading to shortened
5184 data being returned. It should now handle lines of any length.
5186 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5187 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5190 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5191 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5192 no longer happens in comment lines. A list of replacements is now placed
5193 at the head of all of the source files, except those whose only change is
5194 to replace PERL_COMMAND in the very first #! line.
5196 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5197 messages on the queue, with a bottom-up merge sort, using code contributed
5198 by Michael Haardt. This should make operations like -bp somewhat faster on
5199 large queues. It won't affect queue runners, except when queue_run_in_order
5202 51. Installed eximstats 1.31 in the distribution.
5204 52. Added support for SRV lookups to the dnslookup router.
5206 53. If an ACL referred to $message_body or $message_body_end, the value was not
5207 reset for any messages that followed in the same SMTP session.
5209 54. The store-handling optimization for building very long strings was not
5210 differentiating between the different store pools. I don't think this
5211 actually made any difference in practice, but I've tidied it.
5213 55. While running the routers to verify a sender address, $sender_address
5214 was still set to the sender address. This is wrong, because when routing to
5215 send a bounce to the sender, it would be empty. Therefore, I have changed
5216 it so that, while verifying a sender address, $sender_address is set to <>.
5217 (There is no change to what happens when verifying a recipient address.)
5219 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5220 target A or AAAA records (if not already returned) without resetting the
5221 qualify_single or search_parents options of the DNS resolver. These are
5222 inappropriate in this case because the targets of MX and SRV records must
5223 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5224 target that, when qualified, matched something in the local domain. These
5225 two options are now turned off when doing these lookups.
5227 57. It seems that at least some releases of Reiserfs (which does not have the
5228 concept of a fixed number of inodes) returns zero and not -1 for the
5229 number of available inodes. This interacted badly with check_spool_inodes,
5230 which assumed that -1 was the "no such thing" setting. What I have done is
5231 to check that the total number of inodes is greater than zero before doing
5232 the test of how many are available.
5234 58. When a "warn" ACL statement has a log_message modifier, the message is
5235 remembered, and not repeated. This is to avoid a lot of repetition when a
5236 message has many recipients that cause the same warning to be written.
5237 Howewer, Exim was preserving the list of already written lines for an
5238 entire SMTP session, which doesn't seem right. The memory is now reset if a
5239 new message is started.
5241 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5242 debugging output unless log_rewrite was also set.
5244 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5245 of (int)(handle) when we know that handle contains (void *)(-1).
5247 61. The Exim daemon panic-logs an error return when it closes the incoming
5248 connection. However "connection reset by peer" seems to be common, and
5249 isn't really an error worthy of noting specially, so that particular error
5252 62. When Exim is trying to find all the local interfaces, it used to panic and
5253 die if the ioctl to get the interface flags failed. However, it seems that
5254 on at least one OS (Solaris 9) it is possible to have an interface that is
5255 included in the list of interfaces, but for which you get a failure error
5256 for this call. This happens when the interface is not "plumbed" into a
5257 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5258 failure of the "get flags" call assumes that the interface is down.
5260 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5261 makes life easier for people who are doing arithmetic on fields extracted
5262 from dates, where you often get leading zeros that should not be
5263 interpreted as octal.
5265 64. Added qualify_domain to the redirect router, to override the global
5268 65. If a pathologically long header line contained very many addresses (the
5269 report of this problem mentioned 10 000) and each of them was rewritten,
5270 Exim could use up a very large amount of memory. (It kept on making new
5271 copies of the header line as it rewrote, and never released the old ones.)
5272 At the expense of a bit more processing, the header rewriting function has
5273 been changed so that it no longer eats memory in this way.
5275 66. The generation of the Received: header has been moved from the time that a
5276 message starts to be received, to the time that it finishes. The timestamp
5277 in the Received: header should now be very close to that of the <= log
5278 line. There are two side-effects of this change:
5280 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5281 logged header lines no longer include the local Received: line, because
5282 it has not yet been created. The same applies to a copy of the message
5283 that is returned to a non-SMTP sender when a message is rejected.
5285 (b) When a filter file is tested using -bf, no additional Received: header
5286 is added to the test message. After some thought, I decided that this
5289 This change does not affect the value of $received_for. It is still set
5290 after address rewriting, but before local_scan() is called.
5292 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5294 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5295 gave an unhelpful panic error message, and a defer error. I have managed to
5296 change this behaviour so that it now rejects any supplied certificate,
5297 which seems right, as the list of acceptable certificates is empty.
5299 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5300 gave an unhelpful defer error. I have not managed to make this reject any
5301 supplied certificates, but the error message it gives is "no certificate
5302 supplied", which is not helpful.
5304 70. exigrep's output now also includes lines that are not associated with any
5305 message, but which match the given pattern. Implemented by a patch from
5306 Martin Sluka, which also tidied up the Perl a bit.
5308 71. Recipient callout verification, like sender verification, was using <> in
5309 the MAIL FROM command. This isn't really the right thing, since the actual
5310 sender may affect whether the remote host accepts the recipient or not. I
5311 have changed it to use the actual sender in the callout; this means that
5312 the cache record is now keyed on a recipient/sender pair, not just the
5313 recipient address. There doesn't seem to be a real danger of callout loops,
5314 since a callout by the remote host to check the sender would use <>.
5315 [SEE ABOVE: changed after hitting problems.]
5317 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5318 temporary errors. However, in the case of such a code being given after
5319 the end of a data transmission (i.e. after ".") Exim was failing to write
5320 a retry record for the message. (Yes, there was some broken host that was
5321 actually sending 8xx at this point.)
5323 73. An unknown lookup type in a host list could cause Exim to panic-die when
5324 the list was checked. (An example that provoked this was putting <; in the
5325 middle of a list instead of at the start.) If this happened during a DATA
5326 ACL check, a -D file could be left lying around. This kind of configuration
5327 error no longer causes Exim to die; instead it causes a defer errror. The
5328 incident is still logged to the main and panic logs.
5330 74. Buglet left over from Exim 3 conversion. The message "too many messages
5331 in one connection" was written to the rejectlog but not the mainlog, except
5332 when address rewriting (yes!) was being logged.
5334 75. Added write_rejectlog option.
5336 76. When a system filter was run not as root (that is, when system_filter_user
5337 was set), the values of the $n variables were not being returned to the
5338 main process; thus, they were not subsequently available in the $sn
5341 77. Added +return_path_on_delivery log selector.
5343 78. A connection timeout was being treated differently from recipients deferred
5344 when testing hosts_max_try with a message that was older than the host's
5345 retry timeout. (The host should not be counted, thus allowing all hosts to
5346 be tried at least once before bouncing.) This may have been the cause of an
5347 occasionally reported bug whereby a message would remain on the queue
5348 longer than the retry timeout, but would be bounced if a delivery was
5349 forced. I say "may" because I never totally pinned down the problem;
5350 setting up timeout/retry tests is difficult. See also the next item.
5352 79. The ultimate address timeout was not being applied to errors that involved
5353 a combination of host plus message (for example, a timeout on a MAIL
5354 command). When an address resolved to a number of possible hosts, and they
5355 were not all tried for each delivery (e.g. because of hosts_max_try), a
5356 message could remain on the queue longer than the retry timeout.
5358 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5361 81. Fixed an obscure SMTP outgoing bug which required at least the following
5362 conditions: (a) there was another message waiting for the same server;
5363 (b) the server returned 5xx to all RCPT commands in the first message so
5364 that the message was not completed; (c) the server dropped the connection
5365 or gave a negative response to the RSET that Exim sends to abort the
5366 transaction. The observed case was a dropped connection after DATA that had
5367 been sent in pipelining mode. That is, the server had advertised PIPELINING
5368 but was not implementing it correctly. The effect of the bug was incorrect
5369 behaviour, such as trying another host, and this could lead to a crash.
5375 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5376 and daemon.c were passed as pointers to ints; they should have been
5377 pointers to socklen_t variables (which are typically unsigned ints).
5379 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5382 3. Fixed a really odd bug that affected only the testing scheme; patching a
5383 certain fixed string in the binary changed the value of another string that
5384 happened to be identical to the end of the original first string.
5386 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5387 name", it returns that address as the IP address. On some operating
5388 systems (e.g. Solaris), it also passes back the IP address string as the
5389 "host name". However, on others (e.g. Linux), it passes back an empty
5390 string. Exim wasn't checking for this, and was changing the host name to an
5391 empty string, assuming it had been canonicized.
5393 5. Although rare, it is permitted to have more than one PTR record for a given
5394 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5395 all the names associated with an address, because they do in Solaris.
5396 However, it seems that they do not in Linux for data that comes from the
5397 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5398 I found this out when I moved to a new Linux workstation and tried to run
5399 the Exim test suite.
5401 To get round this problem I have changed the code so that it now does its
5402 own call to the DNS to look up PTR records when searching for a host name.
5403 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5404 addresses that are only in /etc/hosts are still found.
5406 This behaviour is, however, controlled by an option called host_lookup_
5407 order, which defaults to "bydns:byaddr". If people want to use the other
5408 order, or indeed, just use one or the other means of lookup, they can
5409 specify it in this variable.
5411 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5412 some operating systems, this comes back from gethostbyaddr() as an empty
5413 string, and this is what Exim used to test for. However, it seems that in
5414 other systems, "." is yielded. Exim now tests for this case too.
5416 7. The values of check_spool_space and check_log_space are now held internally
5417 as a number of kilobytes instead of an absolute number of bytes. If a
5418 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5419 kilobyte. This means that much larger values can be stored.
5421 8. Exim monitor: an attempt to get the action menu when not actually pointing
5422 at a message produces an empty menu entitled "No message selected". This
5423 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5424 no entries in it ("Shell widget menu has zero width and/or height"). So I
5425 have added a single, blank menu entry in this case.
5427 9. Added ${quote_local_part.
5429 10. MIME decoding is now applied to the contents of Subject: header lines when
5432 11. Now that a reference to $sender_host_address automatically causes a reverse
5433 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5434 host lookup before query-style lookups in lists that might use this
5435 variable. This has therefore been abolished, and the "net-" prefix is no
5436 longer necessary for query-style lookups.
5438 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5439 have been a typo for LFLAGS, so it has been changed.
5441 13. The install script calls Exim with "-C /dev/null" in order to find the
5442 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5443 to be output. Howeve, since Exim outputs its version number before the
5444 error, it didn't break the script. It just looked ugly. I fixed this by
5445 always allowing "-C /dev/null" if the caller is root.
5447 14. Ignore overlarge ACL variable number when reading spool file - insurance
5448 against a later release with more variables having written the file.
5450 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5451 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5452 this, as well as the form without the "IPv6" on the front.
5454 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5455 OS/Makefile-Darwin file.
5457 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5458 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5460 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5461 variable. However, some OS use ints for inodes. I've added cast to long int
5462 to get rid of the compiler warning.
5464 19. I had forgotten to lock out "/../" in configuration file names when
5465 ALT_CONFIG_PREFIX was set.
5467 20. Routers used for verification do not need to specify transports. However,
5468 if such a router generated a host list, and callout was configured, Exim
5469 crashed, because it could not find a port number from the (non-existent)
5470 transport. It now assumes port 25 in this circumstance.
5472 21. Added the -t option to exigrep.
5474 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5475 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5476 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5477 them is set, LOOKUP_LSEARCH is forced.
5479 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5481 24. Added sender and host information to the "rejected by local_scan()" log
5482 line; previously there was no indication of these.
5484 25. Added .include_if_exists.
5486 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5487 Linux. It turns out that not all file systems support this. Apparently some
5488 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5489 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5490 not supported on the file descriptor, is now ignored when Exim is trying to
5491 sync a directory. This applies only to Linux.
5493 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5495 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5496 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5499 29. Added support for maildirsize files from supplied patch (modified a bit).
5501 30. The use of :fail: followed by an empty string could lead Exim to respond to
5502 sender verification failures with (e.g.):
5504 550 Verification failed for <xxx>
5505 550 Sender verify failed
5507 where the first response line was missing the '-' that indicates it is not
5508 the final line of the response.
5510 31. The loop for finding the name of the user that called Exim had a hardwired
5511 limit of 10; it now uses the value of finduser_retries, which is used for
5512 all other user lookups.
5514 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5517 33. Exim was neglecting to zero errno before one call of strtol() when
5518 expanding a string and expecting an integer value. On some systems this
5519 resulted in spurious "integer overflow" errors. Also, it was casting the
5520 result into an int without checking.
5522 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5523 did not work. The code looks as if it has *never* worked, though it appears
5524 to have been documented since at least releast 1.62. I have made it work.
5526 35. The "timeout_DNS" error in retry rules, also documented since at least
5527 1.62, also never worked. As it isn't clear exactly what this means, and
5528 clearly it isn't a major issue, I have abolished the feature by treating it
5529 as "timeout", and writing a warning to the main and panic logs.
5531 36. The display of retry rules for -brt wasn't always showing the error code
5534 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5535 timeout_connect_A, timeout_connect_MX.
5537 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5538 to the empty sender.
5540 39. The daemon was not analysing the content of -oX till after it had closed
5541 stderr and disconnected from the controlling terminal. This meant that any
5542 syntax errors were only noted on the panic log, and the return code from
5543 the command was 0. By re-arranging the code a little, I've made the
5544 decoding happen first, so such errors now appear on stderr, and the return
5545 code is 1. However, the actual setting up of the sockets still happens in
5546 the disconnected process, so errors there are still only recorded on the
5549 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5550 connections (as happens on some IP stacks) was logged at start up time as
5551 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5552 it from "IPv6 and IPv4", which means that two separate sockets are being
5555 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5556 says whether AF_INET or AF_INET6 was passed as an argument.
5558 42. Exiwhat output was messed up when time zones were included in log
5561 43. Exiwhat now gives more information about the daemon's listening ports,
5562 and whether -tls-on-connect was used.
5564 44. The "port" option of the smtp transport is now expanded.
5566 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5567 silently ignored. Now an error message is written to the main and panic
5570 46. There's a new ACL modifier called "logwrite" which writes to a log file
5571 as soon as it is encountered.
5573 47. Added $local_user_uid and $local_user_gid at routing time.
5575 48. Exim crashed when trying to verify a sender address that was being
5578 49. Exim was recognizing only a space character after ".include". It now also
5579 recognizes a tab character.
5581 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5582 extracting the relevant information from the specification. The man page no
5583 longer contains scrambled data for the -d option, and I've added a section
5584 at the front about calling Exim under different names.
5586 51. Added "extra_headers" argument to the "mail" command in filter files.
5588 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5591 53. Installed eximstats 1.29.
5593 54. Added transport_filter_timeout as a generic transport option.
5595 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5596 Cc: header lines. This was required by RFC 822, but it not required by RFC
5599 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5600 incoming messages that did not have them. Now it does so only if the
5601 message originates locally, that is, if there is no associated remote host
5602 address. When Resent- header lines are present, this applies to the Resent-
5603 lines rather than the non-Resent- lines.
5605 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5606 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5608 58. Messages for configuration errors now include the name of the main
5609 configuration file - useful now that there may be more than one file in a
5610 list (.included file names were always shown).
5612 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5613 for those rare installations that do not start the daemon as root or run it
5614 setuid root. I've cut out the call to initgroups() if the daemon is not
5617 60. The Exim user and group can now be bound into the binary as text strings
5618 that are looked up at the start of Exim's processing.
5620 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5622 62. Added $mailstore_basename variable.
5624 63. Installed patch to sieve.c from Michael Haardt.
5626 64. When Exim failed to open the panic log after failing to open the main log,
5627 the original message it was trying to log was written to stderr and debug
5628 output, but if they were not available (the usual case in production), it
5629 was lost. Now it is written to syslog before the two lines that record the
5630 failures to open the logs.
5632 65. Users' Exim filters run in subprocesses under the user's uid. It is
5633 possible for a "deliver" command or an alias in a "personal" command to
5634 provoke an address rewrite. If logging of address rewriting is configured,
5635 this fails because the process is not running as root or exim. There may be
5636 a better way of dealing with this, but for the moment (because 4.30 needs
5637 to be released), I have disabled address rewrite logging when running a
5638 filter in a non-root, non-exim process.
5644 1. The buildconfig auxiliary program wasn't quoting the value set for
5645 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5646 not defined. This bug was masked in 4.22 by the effect that was fixed in
5649 2. Some messages that were rejected after a message id was allocated were
5650 shown as "incomplete" by exigrep. It no longer does this for messages that
5651 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5653 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5654 have allow_domain_literals set, the ID did not get logged in the <= line.
5655 Domain literals are now always recognized in Message-ID: header lines.
5657 4. The first argument for a ${extract expansion item is the key name or field
5658 number. Leading and trailing spaces in this item were not being ignored,
5659 causing some misleading effects.
5661 5. When deliver_drop_privilege was set, single queue runner processes started
5662 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5663 same command in the process it spins off) were not dropping privilege.
5665 6. When the daemon running as "exim" started a queue runner, it always
5666 re-executed Exim in the spun-off process. This is a waste of effort when
5667 deliver_drop_privilege is set. The new process now just calls the
5668 queue-runner function directly.
5674 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5677 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5678 ignored. Though the use of -oP was forcing the writing of a pid file, it
5679 was always written to the default place.
5681 3. If the message "no IP address found for host xxxx" is generated during
5682 incoming verification, it is now followed by identification of the incoming
5683 connection (so you can more easily find what provoked it).
5685 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5687 5. Added some features to "harden" Exim a bit more against certain attacks:
5689 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5690 be put in Local/Makefile. This is like the never_users runtime option,
5691 but it cannot be overridden. The default setting is "root".
5693 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5694 prefix string with which any file named in a -C command line option
5697 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5698 is retained for -C and -D only if the caller of Exim is root. Without
5699 it, the exim user may also use -C and -D and retain privilege.
5701 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5702 command line option is disabled.
5704 6. Macro names set by the -D option must start with an upper case letter, just
5705 like macro names defined in the configuration file.
5707 7. Added "dereference=" facility to LDAP.
5709 8. Two instances of the typo "uknown" in the source files are fixed.
5711 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5712 the Configure-Makefile script screwed up while processing it.
5714 10. Incorporated PCRE 4.4.
5716 11. The SMTP synchronization check was not operating right at the start of an
5717 SMTP session. For example, it could not catch a HELO sent before the client
5718 waited for the greeting. There is now a check for outstanding input at the
5719 point when the greeting is written. Because of the duplex, asynchronous
5720 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5721 way, but not yet received, when the check is performed.
5723 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5724 on TCP/IP sockets, because this apparently causes some broken clients to
5727 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5728 unchanged) from the Cygwin maintainer.
5730 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5731 was supported instead of testing for mbx. Effectively a typo.
5733 15. The spa authenticator server code was not checking that the input it
5734 received was valid base64.
5736 16. The debug output line for the "set" modifier in ACLs was not showing the
5737 name of the variable that was being set.
5739 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5741 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5742 lookup on its own. Something else was needed to trigger the lookup. For
5743 example, a match in host_lookup or the need for a host name in a host list.
5744 Now, if $sender_host_name is referenced and the host name has not yet been
5745 looked up, a lookup is performed. If the lookup fails, the variable remains
5746 empty, and $host_lookup_failed is set to "1".
5748 19. Added "eqi" as a case-independent comparison operator.
5750 20. The saslauthd authentication condition could segfault if neither service
5751 nor realm was specified.
5753 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5754 error message that was logged was misleading, and incoming SMTP
5755 connections were dropped. The message is now more accurate, and temporary
5756 errors are given to SMTP connections.
5758 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5759 (except RSET) with a 421 error, until QUIT is received. However, it was
5760 failing to send a response to QUIT.
5762 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5763 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5764 ACL is now run after the helo_try_verify_hosts code.
5766 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5767 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5768 case-independent, so other case variants are also recognized). Apparently
5769 some systems use these upper case variants.
5771 25. If more than two messages were waiting for the same host, and a transport
5772 filter was specified for the transport, Exim sent two messages over the
5773 same TCP/IP connection, and then failed with "socket operation on non-
5774 socket" when it tried to send the third.
5776 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5778 27. The extern definition of crypt16() in expand.c was not being excluded when
5779 the OS had its own crypt16() function.
5781 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5782 as a preferred synonym for return_size_limit, both as an option and as an
5785 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5787 30. Changed the default configuration ACL to relax the local part checking rule
5788 for addresses that are not in any local domains. For these addresses,
5789 slashes and pipe symbols are allowed within local parts, but the sequence
5790 /../ is explicitly forbidden.
5792 31. SPA server authentication was not clearing the challenge buffer before
5795 32. log_message in a "warn" ACL statement was writing to the reject log as
5796 well as to the main log, which contradicts the documentation and doesn't
5797 seem right (because no rejection is happening). So I have stopped it.
5799 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5800 However, I am unable to do any testing of this.
5802 34. Fixed an infelicity in the appendfile transport. When checking directories
5803 for a mailbox, to see if any needed to be created, it was accidentally
5804 using path names with one or more superfluous leading slashes; tracing
5805 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5807 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5808 contents are added to the log line that is written for every discarded
5809 recipient. (Previously a log_message setting was ignored.)
5811 36. The ${quote: operator now quotes the string if it is empty.
5813 37. The install script runs exim in order to find its version number. If for
5814 some reason other than non-existence or emptiness, which it checks, it
5815 could not run './exim', it was installing it with an empty version number,
5816 i.e. as "exim-". This error state is now caught, and the installation is
5819 38. An argument was missing from the function that creates an error message
5820 when Exim fails to connect to the socket for saslauthd authentication.
5821 This could cause Exim to crash, or give a corrupted message.
5823 39. Added isip, isip4, and isip6 to ${if conditions.
5825 40. The ACL variables $acl_xx are now saved with the message, and can be
5826 accessed later in routers, transports, and filters.
5828 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5829 strings in the file are not string-expanded.
5831 42. If a MAIL command specified a SIZE value that was too large to fit into an
5832 int variable, the check against message_size_limit failed. Such values are
5833 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5834 day this will have to be increased, but I don't think I want to be around
5835 when emails are that large.
5842 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5843 iconv() is not standard in FreeBSD.
5845 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5846 IPv6 enabled. The observed symptom was a segmentation fault on return from
5847 the function os_common_find_running_interfaces() in src/os.c.
5849 3. In the check_special_case() function in daemon.c I had used "errno" as an
5850 argument name, which causes warnings on some systems. This was basically a
5851 typo, since it was named "eno" in the comments!
5853 4. The code that waits for the clock to tick (at a resolution of some fraction
5854 of a second) so as to ensure message-id uniqueness was always waiting for
5855 at least one whole tick, when it could have waited for less. [This is
5856 almost certainly not relevant at current processor speeds, where it is
5857 unlikely to ever wait at all. But we try to future-proof.]
5859 5. The function that sleeps for a time interval that includes fractions of a
5860 second contained a race. It did not block SIGALRM between setting the
5861 timer, and suspending (a couple of lines later). If the interval was short
5862 and the sigsuspend() was delayed until after it had expired, the suspension
5863 never ended. On busy systems this could lead to processes getting stuck for
5866 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5867 process, before it forks any delivery processes. The open lookup caching
5868 mechanism meant that the open file or database connection was passed into
5869 the delivery process. The problem was that delivery processes always tidy
5870 up cached lookup data. This could cause a problem for the next delivery
5871 process started by the queue runner, because the external queue runner
5872 process does not know about the closure. So the next delivery process
5873 still has data in the lookup cache. In the case of a file lookup, there was
5874 no problem because closing a file descriptor in a subprocess doesn't affect
5875 the parent. However, if the lookup was caching a connection to a database,
5876 the connection was closed, and the second delivery process was likely to
5877 see errors such as "PGSQL: query failed: server closed the connection
5878 unexpectedly". The problem has been fixed by closing all cached lookups
5879 in a queue runner before running a delivery process.
5881 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5882 seem to have the "const" qualifier which it has on other OS. I've
5885 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5886 *of the same type* (client or server) with the same public name that an
5887 error should be diagnosed.
5889 9. When Exim looked up a host name for an IP address, but failed to find the
5890 original IP address when looking up the host name (a safety check), it
5891 output the message "<ip address> does not match any IP for NULL", which was
5892 confusing, to say the least. The bug was that the host name should have
5893 appeared instead of "NULL".
5895 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5896 user that is built into the binary, and the -C or -D options is used, root
5897 privilege is dropped before the configuration file is read. In addition,
5898 logging is switched to stderr instead of the normal log files. If the
5899 configuration then re-defines the Exim user, the unprivileged environment
5900 is probably not what is expected, so Exim logs a panic warning message (but
5903 However, if deliver_drop_privilege is set, the unprivileged state may well
5904 be exactly what is intended, so the warning has been cut out in that case,
5905 and Exim is allowed to try to write to its normal log files.
5911 1. smtp_return_error_details was not giving details for temporary sender
5912 or receiver verification errors.
5914 2. Diagnose a configuration error if two authenticators have the same public
5917 3. Exim used not to create the message log file for a message until the first
5918 delivery attempt. This could be confusing when incoming messages were held
5919 for policy or load reasons. The message log file is now created at the time
5920 the message is received, and an initial "Received" line is written to it.
5922 4. The automatically generated man page for command line options had a minor
5923 bug that caused no ill effects; however, a more serious problem was that
5924 the procedure for building the man page automatically didn't always
5925 operate. Consequently, release 4.20 contains an out-of-date version. This
5926 shouldn't happen again.
5928 5. When building Exim with embedded Perl support, the script that builds the
5929 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5930 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5932 6. The freeze_tell option was not being used for messages that were frozen on
5933 arrival, either by an ACL or by local_scan().
5935 7. Added the smtp_incomplete_transaction log selector.
5937 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5938 was accepting AUTH without a new EHLO.
5940 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5941 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5944 10. Logging of TCP/IP connections (when configured) now happens in the main
5945 daemon process instead of the child process, so that the TCP/IP connection
5946 count is more accurate (but it can never be perfect).
5948 11. The use of "drop" in a nested ACL was not being handled correctly in the
5949 outer ACL. Now, if condition failure induced by the nested "drop" causes
5950 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5951 or "require"), the connection is dropped.
5953 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5954 that yield "discard" can now be used with an "accept" or a "discard" verb,
5955 but an error is generated for any others (because I can't see a useful way
5956 to define what should happen).
5958 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5959 are now processed in the same way as lines in the Exim configuration file.
5960 In particular, continuation lines are supported.
5962 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5964 15. Added -ti meaning -t -i.
5966 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5967 domains, and warn by logging if others are found.
5969 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5970 returned from GIFCONF when Exim is trying to find the list of interfaces on
5971 a host. The code in os.c has been modified to copy each ifreq to an aligned
5972 structure in all cases.
5974 Also, in some cases, the returned ifreq's were being copied to a 'struct
5975 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5976 means the last couple of bytes of an IPv6 address could be chopped if the
5977 ifreq contained only a normal sockaddr (14 bytes storage).
5979 18. Named domain lists were not supported in the hosts_treat_as_local option.
5980 An entry such as +xxxx was not recognized, and was treated as a literal
5983 19. Ensure that header lines added by a DATA ACL are included in the reject log
5984 if the ACL subsequently rejects the message.
5986 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5987 MD5 (which is deprecated).
5989 21. When testing a filter file using -bf, Exim was writing a message when it
5990 took the sender from a "From " line in the message, but it was not doing so
5991 when it took $return_path from a Return-Path: header line. It now does.
5993 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5994 with a valid header line field name (a series of printing characters
5995 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5997 23. Changed "disc" in the source to "disk" to conform to the documentation and
5998 the book and for uniformity.
6000 24. Ignore Sendmail's -Ooption=value command line item.
6002 25. When execve() failed while trying to run a command in a pipe transport,
6003 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6004 could be confused with a return value of 69 from the command itself. This
6005 has been changed to 127, the value the shell returns if it is asked to run
6006 a non-existent command. The wording for the related log line suggests a
6007 non-existent command as the problem.
6009 26. If received_header_text expands to an empty string, do not add a Received:
6010 header line to the message. (Well, it adds a token one on the spool, but
6011 marks it "old" so that it doesn't get used or transmitted.)
6013 27. Installed eximstats 1.28 (addition of -nt option).
6015 28. There was no check for failure on the call to getsockname() in the daemon
6016 code. This can fail if there is a shortage of resources on the system, with
6017 ENOMEM, for example. A temporary error is now given on failure.
6019 29. Contrary to the C standard, it seems that in some environments, the
6020 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6021 program. Exim now does this explicitly; it affects the formatting of
6022 timestamps using strftime().
6024 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6025 complaints. I've now initialized all the variables, to avoid this.
6027 32. Header lines added by a system filter were not being "seen" during
6028 transport-time rewrites.
6030 33. The info_callback() function passed to OpenSSL is set up with type void
6031 (*)(SSL *, int, int), as described somewhere. However, when calling the
6032 function (actually a macro) that sets it up, the type void(*)() is
6033 expected. I've put in a cast to prevent warnings from picky compilers.
6035 34. If a DNS black list lookup found a CNAME record, but there were no A
6036 records associated with the domain it pointed at, Exim crashed.
6038 35. If a DNS black list lookup returned more than one A record, Exim ignored
6039 all but the first. It now scans all returned addresses if a particular IP
6040 value is being sought. In this situation, the contents of the
6041 $dnslist_value variable are a list of all the addresses, separated by a
6044 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6045 to accept a host name and all its aliases if the forward lookup for any of
6046 them yielded the IP address of the incoming connection. Now it accepts only
6047 those names whose forward lookup yields the correct IP address. Any other
6048 names are discarded. This closes a loophole whereby a rogue DNS
6049 administrator could create reverse DNS records to break through a
6050 wildcarded host restriction in an ACL.
6052 37. If a user filter or a system filter that ran in a subprocess used any of
6053 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6054 the wrong values were passed to the pipe command ($thisaddress had the
6055 value of $0, $0 had the value of $1, etc). This bug was introduced by
6056 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6058 38. Improved the line breaking for long SMTP error messages from ACLs.
6059 Previously, if there was no break point between 40 and 75 characters, Exim
6060 left the rest of the message alone. Two changes have been made: (a) I've
6061 reduced the minimum length to 35 characters; (b) if it can't find a break
6062 point between 35 and 75 characters, it looks ahead and uses the first one
6063 that it finds. This may give the occasional overlong line, but at least the
6064 remaining text gets split now.
6066 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6067 file descriptors might be low, and that setting 1000 would always raise it.
6068 It turns out that in some environments, the limit is already over 1000 and
6069 that lowering it causes trouble. So now Exim takes care not to decrease it.
6071 40. When delivering a message, the value of $return_path is set to $sender_
6072 address at the start of routing (routers may change the value). By an
6073 oversight, this default was not being set up when an address was tested by
6074 -bt or -bv, which affected the outcome if any router or filter referred to
6077 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6078 log only when "message" or "log_message" are set. However, if one of the
6079 conditions was an address verification, or a call to a nested ACL, the
6080 messages generated by the underlying test were being passed through. This
6081 no longer happens. The underlying message is available in $acl_verify_
6082 message for both "message" and "log_message" expansions, so it can be
6083 passed through if needed.
6085 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6086 new expansion $bh_ to give the encoded byte string without charset
6087 translation. Translation happens only if iconv() is available; HAVE_ICONV
6088 indicates this at build time. HEADERS_CHARSET gives the charset to
6089 translate to; headers_charset can change it in the configuration, and
6090 "headers charset" can change it in an individual filter file.
6092 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6093 that creates RFC 2047 encoded "words" labels them as that charset instead
6094 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6095 expansion operator; (ii) when Exim creates a From: line for a local
6096 message; (iii) when a header line is rewritten to include a "phrase" part.
6098 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6099 buggy, causing it to skip the first lines of messages whose message ID
6100 ended in 'D'. This would not have bitten before Exim release 4.14, because
6101 message IDs were unlikely to end in 'D' before then. The effect was to have
6102 incorrect size information for certain domains.
6104 45. #include "config.h" was missing at the start of the crypt16.c module. This
6105 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6108 46. If there was a timeout during a "random" callout check, Exim treated it as
6109 a failure of the random address, and carried on sending RSET and the real
6110 address. If the delay was just some slowness somewhere, the response to the
6111 original RCPT would be taken as a response to RSET and so on, causing
6112 mayhem of various kinds.
6114 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6115 when I implemented it. It didn't allow for the fact that some option values
6116 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6117 the right test for positive values.
6119 48. Domain names in DNS records are case-independent. Exim always looks them up
6120 in lower case. Some resolvers return domain names in exactly the case they
6121 appear in the zone file, that is, they may contain uppercase letters. Not
6122 all resolvers do this - some return always lower case. Exim was treating a
6123 change of case by a resolver as a change of domain, similar to a widening
6124 of a domain abbreviation. This triggered its re-routing code and so it was
6125 trying to route what was effectively the same domain again. This normally
6126 caused routing to fail (because the router wouldn't handle the domain
6127 twice). Now Exim checks for this case specially, and just changes the
6128 casing of the domain that it ultimately uses when it transmits the message
6131 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6134 50. If a filter generated a file delivery with a non-absolute name (possible if
6135 no home directory exists for the router), the forbid_file option was not
6138 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6139 the existing equality matching.
6141 52. Exim was using ints instead of ino_t variables in some places where it was
6142 dealing with inode numbers.
6144 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6145 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6146 called TMPDIR, and if it finds it is different, it changes its value.
6148 54. The smtp_printf() function is now made available to local_scan() so
6149 additional output lines can be written before returning. There is also an
6150 smtp_fflush() function to enable the detection of a dropped connection.
6151 The variables smtp_input and smtp_batched_input are exported to
6154 55. Changed the default runtime configuration: the message "Unknown user"
6155 has been removed from the ACL, and instead placed on the localuser router,
6156 using the cannot_route_message feature. This means that any verification
6157 failures that generate their own messages won't get overridden. Similarly,
6158 the "Unrouteable address" message that was in the ACL for unverifiable
6159 relay addresses has also been removed.
6161 56. Added hosts_avoid_esmtp to the smtp transport.
6163 57. The exicyclog script was not checking for the esoteric option
6164 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6165 will work only if exicyclog is run under the appropriate euid.
6167 58. Following a discussion on the list, the rules by which Exim recognises line
6168 endings on incoming messages have been changed. The -dropcr and drop_cr
6169 options are now no-ops, retained only for backwards compatibility. The
6170 following line terminators are recognized: LF CRLF CR. However, special
6171 processing applies to CR:
6173 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6174 nor a local message in the state where . is a terminator.
6176 (ii) If a bare CR is encountered in a header line, an extra space is added
6177 after the line terminator so as not to end the header. The reasoning
6178 behind this is that bare CRs in header lines are most likely either
6179 to be mistakes, or people trying to play silly games.
6181 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6182 was being incorrectly given as 18 bytes larger than it should have been.
6183 This is a VOB (very old bug).
6185 60. This may never have affected anything current, but just in case it has:
6186 When the local host is found other than at the start of a list of hosts,
6187 the local host, those with the same MX, and any that follow, are discarded.
6188 When the list in question was part of a longer list of hosts, the following
6189 hosts (not currently being processed) were also being discarded. This no
6190 longer happens. I'm not sure if this situation could ever has previously
6193 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6196 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6197 header. This is something that is recommended in a new Internet Draft, and
6198 is something that is documented as being done by Sendmail. There are two
6199 possible values. For messages generated by the autoreply transport, Exim
6202 Auto-Submitted: auto-replied
6204 whereas for all other generated messages (e.g. bounces) it adds
6206 Auto-Submitted: auto-generated
6208 63. The "personal" condition in filters now includes a test for the
6209 Auto-Submitted: header. If it contains the string "auto-" the message it
6210 not considered personal.
6212 64. Added rcpt_include_affixes as a generic transport option.
6214 65. Added queue_only_override (default true).
6216 66. Added the syslog_duplication option.
6218 67. If what should have been the first header line of a message consisted of
6219 a space followed by a colon, Exim was mis-interpreting it as a header line.
6220 It isn't of course - it is syntactically invalid and should therefore be
6221 treated as the start of the message body. The misbehaviour could have
6222 caused a number of strange effects, including loss of data in subsequent
6223 header lines, and spool format errors.
6225 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6226 client host had authenticated. This control can now be exercised by an ACL
6227 for more flexibility.
6229 69. By default, callouts do not happen when testing with -bh. There is now a
6230 variant, -bhc, which does actually run the callout code, including
6231 consulting and updating the callout cache.
6233 70. Added support for saslauthd authentication, courtesy of Alexander
6236 71. If statvfs() failed on the spool or log directories while checking their
6237 size for availability, Exim confusingly gave the error "space shortage".
6238 Furthermore, in debugging mode it crashed with a floating point exception.
6239 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6240 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6241 serious problem, Exim now writes to the main and panic logs when this
6242 happens, with details of the failure. It then refuses to accept the
6243 incoming message, giving the message "spool directory problem" or "log
6244 directory problem" with a 421 code for SMTP messages.
6246 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6247 0, 1, and 2 exist, because some OS complain for execs without them (see
6248 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6249 descriptors. However, the code omitted to check that the open succeeded,
6250 causing mysterious errors if for some reason the permissions on /dev/null
6251 got screwed. Now Exim writes a message to the main and panic logs, and
6252 bombs out if it can't open /dev/null.
6254 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6255 interact so that it is all more flexible. It is supposed to remain
6256 backwards compatible. Also added extra_local_interfaces.
6258 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6259 to bomb out with an assertion failure - to the client this appears as a
6260 connection drop. This problem occurs in the part of the code that was taken
6261 from the Samba project. Fortunately, the assertion is in a very simple
6262 function, so I have fixed this by reproducing the function inline in the
6263 one place where it is called, and arranging for authentication to fail
6264 instead of killing the process with assert().
6266 75. The SPA client code was not working when the server requested OEM rather
6267 than Unicode encoding.
6269 76. Added code to make require_files with a specific uid setting more usable in
6270 the case where statting the file as root fails - usually a non-root-mounted
6271 NFS file system. When this happens and the failure is EACCES, Exim now
6272 forks a subprocess and does the per-uid checking as the relevant uid.
6274 77. Added process_log_path.
6276 78. If log_file_path was not explicitly set, a setting of check_log_space or
6277 check_log_inodes was ignored.
6279 79. If a space check for the spool or log partitions fails, the incident is now
6280 logged. Of course, in the latter case the data may get lost...
6282 80. Added the %p formatting code to string_format() so that it can be used to
6283 print addresses in debug_print(). Adjusted all the address printing in the
6284 debugging in store.c to use %p rather than %d.
6286 81. There was a concern that a line of code in smtp_in.c could overflow a
6287 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6288 initially expressed, the concern was not well-founded, because trailing
6289 spaces are removed early. However, if the trailing spaces were followed by
6290 a NULL, they did not get removed, so the overflow was possible. Two fixes
6293 (a) I re-wrote the offending code in a cleaner fashion.
6294 (b) If an incoming SMTP command contains a NULL character, it is rejected
6297 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6298 runs initgroups(), so that if the Exim user is in any additional groups,
6299 they will be used during message reception.
6305 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob