1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way. Bug 68.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and priority
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
61 and remote servers. Remote servers can be IPv6 addresses, and
64 JH/17 Bug 68: The spamd_address main option now supports an optional
65 timeout value per server.
67 JH/18 Bug 1581: Router and transport options headers_add/remove can
68 now have the list separator specified.
70 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
73 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
76 JH/21 Support for the A6 type of dns record is withdrawn.
78 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
79 rather than the verbs used.
81 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
82 from 255 to 1024 chars.
84 JH/24 Verification callouts now attempt to use TLS by default.
86 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
87 are generic router options now. The defaults didn't change.
89 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
90 Original patch from Alexander Shikoff, worked over by JH.
92 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
93 files if ZCAT_COMMAND is not executable.
95 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
97 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
99 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
100 Normally benign, it bites when the pair was led to by a CNAME;
101 modern usage is to not canoicalize the domain to a CNAME target
102 (and we were inconsistent anyway for A-only vs AAAA+A).
104 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
106 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
107 when evaluating $sender_host_dnssec.
109 JH/31 Check the HELO verification lookup for DNSSEC, adding new
110 $sender_helo_dnssec variable.
115 TL/01 When running the test suite, the README says that variables such as
116 no_msglog_check are global and can be placed anywhere in a specific
117 test's script, however it was observed that placement needed to be near
118 the beginning for it to behave that way. Changed the runtest perl
119 script to read through the entire script once to detect and set these
120 variables, reset to the beginning of the script, and then run through
121 the script parsing/test process like normal.
123 TL/02 The BSD's have an arc4random API. One of the functions to induce
124 adding randomness was arc4random_stir(), but it has been removed in
125 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
126 function when detected.
128 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
129 cause callback expansion.
131 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
132 syntax errors in an expansion can be treated as a string instead of
133 logging or causing an error, due to the internal use of bool_lax
134 instead of bool when processing it.
136 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
137 server certificates when making smtp deliveries.
139 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
141 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
143 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
145 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
146 Merged patch from Sebastian Wiedenroth.
148 JH/05 Fix results-pipe from transport process. Several recipients, combined
149 with certificate use, exposed issues where response data items split
150 over buffer boundaries were not parsed properly. This eventually
151 resulted in duplicates being sent. This issue only became common enough
152 to notice due to the introduction of conection certificate information,
153 the item size being so much larger. Found and fixed by Wolfgang Breyha.
155 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
156 size buffer was used, resulting in syntax errors when an expansion
159 JH/07 Add support for directories of certificates when compiled with a GnuTLS
160 version 3.3.6 or later.
162 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
163 is EXPERIMENTAL_EVENT, the main-configuration and transport options
164 both become "event_action", the variables become $event_name, $event_data
165 and $event_defer_errno. There is a new variable $verify_mode, usable in
166 routers, transports and related events. The tls:cert event is now also
167 raised for inbound connections, if the main configuration event_action
170 TL/06 In test suite, disable OCSP for old versions of openssl which contained
171 early OCSP support, but no stapling (appears to be less than 1.0.0).
173 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
174 server certificate names available under the smtp transport option
175 "tls_verify_cert_hostname" now do not permit multi-component wildcard
178 JH/10 Time-related extraction expansions from certificates now use the main
179 option "timezone" setting for output formatting, and are consistent
180 between OpenSSL and GnuTLS compilations. Bug 1541.
182 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
183 encoded parameter in the incoming message. Bug 1558.
185 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
186 include certificate info, eximon was claiming there were spoolfile
189 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
191 JH/14 Log delivery-related information more consistently, using the sequence
192 "H=<name> [<ip>]" wherever possible.
194 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
195 are problematic for Debian distribution, omit them from the release
198 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
200 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
203 JH/17 Fix a null-indirection in certextract expansions when a nondefault
204 output list separator was used.
209 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
210 checkers that were complaining about end of non-void function with no
213 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
214 This was a regression intruduced in 4.83 by another bugfix.
216 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
218 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
219 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
225 TF/01 Correctly close the server side of TLS when forking for delivery.
227 When a message was received over SMTP with TLS, Exim failed to clear up
228 the incoming connection properly after forking off the child process to
229 deliver the message. In some situations the subsequent outgoing
230 delivery connection happened to have the same fd number as the incoming
231 connection previously had. Exim would try to use TLS and fail, logging
232 a "Bad file descriptor" error.
234 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
235 utilities have not been installed.
237 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
238 temporary space as the ACL may create new global variables.
240 TL/01 LDAP support uses per connection or global context settings, depending
241 upon the detected version of the libraries at build time.
243 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
244 to extract and use the src ip:port in logging and expansions as if it
245 were a direct connection from the outside internet. PPv2 support was
246 updated based on HAProxy spec change in May 2014.
248 JH/02 Add ${listextract {number}{list}{success}{fail}}.
250 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
251 Properly escape header and check for NULL return.
253 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
256 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
258 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
259 characters in header names, implemented as a verify condition.
260 Contributed by Michael Fischer v. Mollard.
262 TL/05 Rename SPF condition results err_perm and err_temp to standardized
263 results permerror and temperror. Previous values are deprecated but
264 still accepted. In a future release, err_perm and err_temp will be
265 completely removed, which will be a backward incompatibility if the
266 ACL tests for either of these two old results. Patch contributed by
267 user bes-internal on the mailing list.
269 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
271 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
272 selectors, in both main and reject logs.
274 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
277 JH/07 Add malware type "sock" for talking to simple daemon.
279 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
281 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
282 routers/transports under cutthrough routing.
284 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
285 numbers. Touch up "bool" conditional to keep the same definition.
287 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
289 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
291 JH/12 Expand items in router/transport headers_add or headers_remove lists
292 individually rather than the list as a whole. Bug 1452.
294 Required for reasonable handling of multiple headers_ options when
295 they may be empty; requires that headers_remove items with embedded
296 colons must have them doubled (or the list-separator changed).
298 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
299 view the policy declared in the DMARC record. Currently, $dmarc_status
300 is a combined value of both the record presence and the result of the
303 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
305 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
306 dnslookup router and the smtp transport (applying to the forward
309 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
310 of ldap servers used for a specific lookup. Patch provided by Heiko
313 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
314 New variable $lookup_dnssec_authenticated for observability.
316 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
317 Patch submitted by Lars Timman.
319 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
321 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
322 Requires trusted mode and valid format message id, aborts otherwise.
323 Patch contributed by Heiko Schlichting.
325 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
326 certextract with support for various fields. Bug 1358.
328 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
329 is requested by default, modifiable by smtp transport option
332 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
333 operate on certificate variables to give certificate fingerprints
334 Also new ${sha256:cert_variable}.
336 JH/23 The PRDR feature is moved from being Experimental into the mainline.
338 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
339 Christian Aistleitner.
341 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
343 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
344 file. Patch from Wolfgang Breyha.
346 JH/25 Expand the coverage of the delivery $host and $host_address to
347 client authenticators run in verify callout. Bug 1476.
349 JH/26 Port service names are now accepted for tls_on_connect_ports, to
350 align with daemon_smtp_ports. Bug 72.
352 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
353 support and error reporting did not work properly.
355 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
356 and is readable. Patch from Andrew Colin Kissa.
358 TL/14 Enhance documentation of ${run expansion and how it parses the
359 commandline after expansion, particularly in the case when an
360 unquoted variable expansion results in an empty value.
362 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
364 PP/02 Fix internal collision of T_APL on systems which support RFC3123
365 by renaming away from it. Addresses GH issue 15, reported by
368 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
370 TL/15 SECURITY: prevent double expansion in math comparison functions
371 (can expand unsanitized data). Not remotely exploitable.
378 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
380 PP/02 Make -n do something, by making it not do something.
381 When combined with -bP, the name of an option is not output.
383 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
386 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
387 $sender_host_name and config options to manage this, and basic check
390 PP/05 DSCP support for outbound connections and control modifier for inbound.
392 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
393 (Only plugin which currently uses this is kerberos4, which nobody should
394 be using, but we should make it available and other future plugins might
395 conceivably use it, even though it would break NAT; stuff *should* be
396 using channel bindings instead).
398 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
399 name; added for Sendmail compatibility; requires admin caller.
400 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
401 just ignore it); requires trusted caller.
402 Also parse but ignore: -Ac -Am -X<logfile>
405 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
407 TL/02 Add +smtp_confirmation as a default logging option.
409 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
410 Patch by Magnus Holmgren from 2007-02-20.
412 TL/04 Bugzilla 1281 - Spec typo.
413 Bugzilla 1283 - Spec typo.
414 Bugzilla 1290 - Spec grammar fixes.
416 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
418 TL/06 Add Experimental DMARC support using libopendmarc libraries.
420 TL/07 Fix an out of order global option causing a segfault. Reported to dev
421 mailing list by by Dmitry Isaikin.
423 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
425 JH/02 Support "G" suffix to numbers in ${if comparisons.
427 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
429 NM/01 Bugzilla 1197 - Spec typo
430 Bugzilla 1196 - Spec examples corrections
432 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
434 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
435 gnutls_enable_pkcs11, but renamed to more accurately indicate its
438 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
439 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
441 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
442 "acl {{name}{arg}...}", and optional args on acl condition
445 JH/05 Permit multiple router/transport headers_add/remove lines.
447 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
449 JH/07 Avoid using a waiting database for a single-message-only transport.
450 Performance patch from Paul Fisher. Bugzilla 1262.
452 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
455 JH/09 Add $headers_added variable, with content from use of ACL modifier
456 add_header (but not yet added to the message). Bugzilla 199.
458 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
459 Pulled from Bugzilla 817 by Wolfgang Breyha.
461 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
463 (nb: this is the same fix as in Exim 4.80.1)
465 JH/11 Add A= logging on delivery lines, and a client_set_id option on
468 JH/12 Add optional authenticated_sender logging to A= and a log_selector
471 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
473 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
474 advertise SMTP AUTH mechanism to us, instead of a generic
475 protocol violation error. Also, make Exim more robust to bad
476 data from the Dovecot auth socket.
478 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
480 When a queue runner is handling a message, Exim first routes the
481 recipient addresses, during which it prunes them based on the retry
482 hints database. After that it attempts to deliver the message to
483 any remaining recipients. It then updates the hints database using
486 So if a recipient address works intermittently, it can get repeatedly
487 deferred at routing time. The retry hints record remains fresh so the
488 address never reaches the final cutoff time.
490 This is a fairly common occurrence when a user is bumping up against
491 their storage quota. Exim had some logic in its local delivery code
492 to deal with this. However it did not apply to per-recipient defers
493 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
495 This change adds a proper retry rule check during routing so that the
496 final cutoff time is checked against the message's age. We only do
497 this check if there is an address retry record and there is not a
498 domain retry record; this implies that previous attempts to handle
499 the address had the retry_use_local_parts option turned on. We use
500 this as an approximation for the destination being like a local
501 delivery, as in LMTP.
503 I suspect this new check makes the old local delivery cutoff check
504 redundant, but I have not verified this so I left the code in place.
506 TF/02 Correct gecos expansion when From: is a prefix of the username.
508 Test 0254 submits a message to Exim with the header
512 When I ran the test suite under the user fanf2, Exim expanded
513 the header to contain my full name, whereas it should have added
514 a Resent-Sender: header. It erroneously treats any prefix of the
515 username as equal to the username.
517 This change corrects that bug.
519 GF/01 DCC debug and logging tidyup
520 Error conditions log to paniclog rather than rejectlog.
521 Debug lines prefixed by "DCC: " to remove any ambiguity.
523 TF/03 Avoid unnecessary rebuilds of lookup-related code.
525 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
526 Bug spotted by Jeremy Harris; was flawed since initial commit.
527 Would have resulted in OCSP responses post-SNI triggering an Exim
528 NULL dereference and crash.
530 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
532 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
533 Bug detection, analysis and fix by Samuel Thibault.
534 Bugzilla 1331, Debian bug #698092.
536 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
538 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
539 Server implementation by Todd Lyons, client by JH.
540 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
541 config variable "prdr_enable" controls whether the server
542 advertises the facility. If the client requests PRDR a new
543 acl_data_smtp_prdr ACL is called once for each recipient, after
544 the body content is received and before the acl_smtp_data ACL.
545 The client is controlled by bolth of: a hosts_try_prdr option
546 on the smtp transport, and the server advertisement.
547 Default client logging of deliveries and rejections involving
548 PRDR are flagged with the string "PRDR".
550 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
551 fclose(). Diagnosis by Todd Lyons.
553 PP/17 Update configure.default to handle IPv6 localhost better.
554 Patch by Alain Williams (plus minor tweaks).
557 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
558 This is now consistent with GnuTLS, and is now documented: the
559 previous undocumented portable approach to treating the option as
560 unset was to force an expansion failure. That still works, and
561 an empty string is now equivalent.
563 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
564 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
565 not performing validation itself.
567 PP/20 Added force_command boolean option to pipe transport.
568 Patch from Nick Koston, of cPanel Inc.
570 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
573 TF/04 Added udpsend ACL modifer and hexquote expansion operator
575 PP/21 Fix eximon continuous updating with timestamped log-files.
576 Broken in a format-string cleanup in 4.80, missed when I repaired the
577 other false fix of the same issue.
578 Report and fix from Heiko Schlichting.
581 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
582 Report from Prashanth Katuri.
584 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
585 It's SecureTransport, so affects any MacOS clients which use the
586 system-integrated TLS libraries, including email clients.
588 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
589 using a MIME ACL for non-SMTP local injection.
590 Report and assistance in diagnosis by Warren Baker.
592 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
594 JH/16 Fix comparisons for 64b. Bugzilla 1385.
596 TL/09 Add expansion variable $authenticated_fail_id to keep track of
597 last id that failed so it may be referenced in subsequent ACL's.
599 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
602 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
603 ldap library initialization, allowing self-signed CA's to be
604 used. Also properly sets require_cert option later in code by
605 using NULL (global ldap config) instead of ldap handle (per
606 session). Bug diagnosis and testing by alxgomz.
608 TL/12 Enhanced documentation in the ratelimit.pl script provided in
609 the src/util/ subdirectory.
611 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
612 renamed to Transport Post Delivery Action by Jeremy Harris, as
615 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
616 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
617 redis_servers = needs to be configured which will be used by the redis
618 lookup. Patch from Warren Baker, of The Packet Hub.
620 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
622 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
623 hostname or reverse DNS when processing a host list. Used suggestions
624 from multiple comments on this bug.
626 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
628 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
629 Missed a few lines, added it to make the runtest require no keyboard
632 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
633 contains upper case chars. Make router use caseful_local_part.
635 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
636 support when GnuTLS has been built with p11-kit.
642 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
644 This, or similar/improved, will also be change PP/11 of 4.82.
650 PP/01 Handle short writes when writing local log-files.
651 In practice, only affects FreeBSD (8 onwards).
652 Bugzilla 1053, with thanks to Dmitry Isaikin.
654 NM/01 Bugzilla 949 - Documentation tweak
656 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
659 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
661 PP/02 Implemented gsasl authenticator.
663 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
665 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
666 `pkg-config foo` for cflags/libs.
668 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
669 with rest of GSASL and with heimdal_gssapi.
671 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
672 `pkg-config foo` for cflags/libs for the TLS implementation.
674 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
675 properties get this fed in as external SSF. A number of robustness
676 and debugging improvements to the cyrus_sasl authenticator.
678 PP/08 cyrus_sasl server now expands the server_realm option.
680 PP/09 Bugzilla 1214 - Log authentication information in reject log.
681 Patch by Jeremy Harris.
683 PP/10 Added dbmjz lookup type.
685 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
687 PP/12 MAIL args handles TAB as well as SP, for better interop with
688 non-compliant senders.
689 Analysis and variant patch by Todd Lyons.
691 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
692 Bug report from Lars Müller <lars@samba.org> (via SUSE),
693 Patch from Dirk Mueller <dmueller@suse.com>
695 PP/13 tls_peerdn now print-escaped for spool files.
696 Observed some $tls_peerdn in wild which contained \n, which resulted
697 in spool file corruption.
699 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
700 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
701 or write after TLS renegotiation, which otherwise led to messages
704 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
705 as a tracking header (ie: a signed header comes before the signature).
706 Patch from Wolfgang Breyha.
708 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
709 comma-sep list; embedded commas doubled.
711 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
713 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
715 Report and patch from Dmitry Banschikov.
717 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
718 Removed SSL_clear() after SSL_new() which led to protocol negotiation
719 failures. We appear to now support TLS1.1+ with Exim.
721 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
722 lets Exim select keys and certificates based upon TLS SNI from client.
723 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
724 before an outbound SMTP session. New log_selector, +tls_sni.
726 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
727 NULL dereference. Report and patch from Alun Jones.
729 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
730 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
731 Not seeing resolver debug output on NetBSD, but suspect this is a
732 resolver implementation change.
734 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
735 Left warnings. Added "eximon gdb" invocation mode.
737 PP/21 Defaulting "accept_8bitmime" to true, not false.
739 PP/22 Added -bw for inetd wait mode support.
741 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
742 locate the relevant includes and libraries. Made this the default.
744 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
745 Bugzilla 1246, report and most of solution from Tomasz Kusy.
747 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
748 This may cause build issues on older platforms.
750 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
751 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
752 gnutls_require_mac & gnutls_require_protocols (no longer supported).
753 Added SNI support via GnuTLS too.
754 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
756 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
758 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
759 Applied second patch from Janne, implementing suggestion to default
760 multiple-strings-in-record handling to match SPF spec.
762 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
764 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
765 read-only, out of scope).
766 Patch from Wolfgang Breyha, report from Stuart Northfield.
768 PP/29 Fix three issues highlighted by clang analyser static analysis.
769 Only crash-plausible issue would require the Cambridge-specific
770 iplookup router and a misconfiguration.
771 Report from Marcin Mirosław.
773 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
775 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
776 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
777 As part of this, removing so much warning spew let me fix some minor
778 real issues in debug logging.
780 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
781 assignment on my part. Fixed.
783 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
784 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
785 Janne Snabb (who went above and beyond: thank you).
787 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
788 string otherwise requires a connection and a bunch more work and it's
789 relatively easy to get wrong. Should also expose TLS library linkage
792 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
793 64-bit ${eval} (JH/03).
795 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
796 GNU libc to support some of the 64-bit stuff, should not lead to
797 conflicts. Defined before os.h is pulled in, so if a given platform
798 needs to override this, it can.
800 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
801 protection layer was required, which is not implemented.
802 Bugzilla 1254, patch from Wolfgang Breyha.
804 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
805 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
806 tls_dhparam take prime identifiers. Also unbreak combination of
807 OpenSSL+DH_params+TLSSNI.
809 PP/39 Disable SSLv2 by default in OpenSSL support.
815 PP/01 Solaris build fix for Oracle's LDAP libraries.
816 Bugzilla 1109, patch from Stephen Usher.
818 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
820 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
823 TF/02 Fix a couple more cases where we did not log the error message
824 when unlink() failed. See also change 4.74-TF/03.
826 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
827 lock up or crash if it happened to be inside a call to libc when it
828 got a SIGUSR1 from exiwhat.
830 The SIGUSR1 handler appends the current process status to the process
831 log which is later printed by exiwhat. It used to use the general
832 purpose logging code to do this, but several functions it calls are
833 not safe for signals.
835 The new output code in the SIGUSR1 handler is specific to the process
836 log, and simple enough that it's easy to inspect for signal safety.
837 Removing some special cases also simplifies the general logging code.
838 Removing the spurious timestamps from the process log simplifies
841 TF/04 Improved ratelimit ACL condition.
843 The /noupdate option has been deprecated in favour of /readonly which
844 has clearer semantics. The /leaky, /strict, and /readonly update modes
845 are mutually exclusive. The update mode is no longer included in the
846 database key; it just determines when the database is updated. (This
847 means that when you upgrde Exim will forget old rate measurements.)
849 Exim now checks that the per_* options are used with an update mode that
850 makes sense for the current ACL. For example, when Exim is processing a
851 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
852 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
853 must specify per_mail/readonly. If you omit the update mode it defaults to
854 /leaky where that makes sense (as before) or /readonly where required.
856 The /noupdate option is now undocumented but still supported for
857 backwards compatibility. It is equivalent to /readonly except that in
858 ACLs where /readonly is required you may specify /leaky/noupdate or
859 /strict/noupdate which are treated the same as /readonly.
861 A useful new feature is the /count= option. This is a generalization
862 of the per_byte option, so that you can measure the throughput of other
863 aggregate values. For example, the per_byte option is now equivalent
864 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
866 The per_rcpt option has been generalized using the /count= mechanism
867 (though it's more complicated than the per_byte equivalence). When it is
868 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
869 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
870 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
871 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
872 Note that using per_rcpt with a non-readonly update mode in more than
873 one ACL will cause the recipients to be double-counted. (The per_mail
874 and per_byte options don't have this problem.)
876 The handling of very low rates has changed slightly. If the computed rate
877 is less than the event's count (usually one) then this event is the first
878 after a long gap. In this case the rate is set to the same as this event's
879 count, so that the first message of a spam run is counted properly.
881 The major new feature is a mechanism for counting the rate of unique
882 events. The new per_addr option counts the number of different
883 recipients that someone has sent messages to in the last time period. It
884 behaves like per_rcpt if all the recipient addresses are different, but
885 duplicate recipient addresses do not increase the measured rate. Like
886 the /count= option this is a general mechanism, so the per_addr option
887 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
888 example, measure the rate that a client uses different sender addresses
889 with the options per_mail/unique=$sender_address. There are further
890 details in the main documentation.
892 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
894 TF/06 Removed a few PCRE remnants.
896 TF/07 Automatically extract Exim's version number from tags in the git
897 repository when doing development or release builds.
899 PP/02 Raise smtp_cmd_buffer_size to 16kB.
900 Bugzilla 879. Patch from Paul Fisher.
902 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
903 Heavily based on revision 40f9a89a from Simon Arlott's tree.
906 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
908 PP/05 Variable $av_failed, true if the AV scanner deferred.
909 Bugzilla 1078. Patch from John Horne.
911 PP/06 Stop make process more reliably on build failure.
912 Bugzilla 1087. Patch from Heiko Schlittermann.
914 PP/07 Make maildir_use_size_file an _expandable_ boolean.
915 Bugzilla 1089. Patch from Heiko Schlittermann.
917 PP/08 Handle ${run} returning more data than OS pipe buffer size.
918 Bugzilla 1131. Patch from Holger Weiß.
920 PP/09 Handle IPv6 addresses with SPF.
921 Bugzilla 860. Patch from Wolfgang Breyha.
923 PP/10 GnuTLS: support TLS 1.2 & 1.1.
925 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
928 PP/11 match_* no longer expand right-hand-side by default.
929 New compile-time build option, EXPAND_LISTMATCH_RHS.
930 New expansion conditions, "inlist", "inlisti".
932 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
934 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
936 PP/14 fix log_write() format string regression from TF/03.
937 Bugzilla 1152. Patch from Dmitry Isaikin.
943 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
945 PP/02 Harmonised TLS library version reporting; only show if debugging.
946 Layout now matches that introduced for other libraries in 4.74 PP/03.
948 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
950 PP/04 New "dns_use_edns0" global option.
952 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
955 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
956 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
958 TK/01 Updated PolarSSL code to 0.14.2.
959 Bugzilla 1097. Patch from Andreas Metzler.
961 PP/07 Catch divide-by-zero in ${eval:...}.
964 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
967 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
968 format-string attack -- SECURITY: remote arbitrary code execution.
970 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
971 time unintentionally subject to list matching rules, letting the header
972 cause arbitrary Exim lookups (of items which can occur in lists, *not*
973 arbitrary string expansion). This allowed for information disclosure.
975 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
976 INT_MIN/-1 -- value coerced to INT_MAX.
982 NM/01 Workround for PCRE version dependency in version reporting
985 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
986 This fixes portability to compilers other than gcc, notably
987 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
989 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
990 makefiles for portability to HP-UX and POSIX correctness.
992 PP/01 Permit LOOKUP_foo enabling on the make command-line.
993 Also via indented variable definition in the Makefile.
994 (Debugging by Oliver Heesakkers).
996 PP/02 Restore caching of spamd results with expanded spamd_address.
997 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
999 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1000 Improves build reliability. Fix from: Frank Elsner
1002 NM/02 Fix wide character breakage in the rfc2047 coding
1003 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1005 NM/03 Allow underscore in dnslist lookups
1006 Fixes bug 1026. Patch from Graeme Fowler
1008 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1009 Code patches from Adam Ciarcinski of NetBSD.
1011 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1014 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1015 is logged, to avoid truncation. Patch from John Horne.
1017 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1018 Patch from Jakob Hirsch.
1020 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1021 SQL string expansion failure details.
1022 Patch from Andrey Oktyabrski.
1024 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1025 Patch from Simon Arlott.
1027 PP/09 New lookups functionality failed to compile on old gcc which rejects
1028 extern declarations in function scope.
1029 Patch from Oliver Fleischmann
1031 PP/10 Use sig_atomic_t for flags set from signal handlers.
1032 Check getgroups() return and improve debugging.
1033 Fixed developed for diagnosis in bug 927 (which turned out to be
1036 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1037 Patch from Mark Zealey.
1039 PP/12 Bugzilla 1056: Improved spamd server selection.
1040 Patch from Mark Zealey.
1042 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1043 Based on patch from Heiko Schlittermann.
1045 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1046 Patch from Uwe Doering, sign-off by Michael Haardt.
1048 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1049 variable declaration deep within a block. Bug and patch from
1052 PP/15 lookups-Makefile IRIX compatibilty coercion.
1054 PP/16 Make DISABLE_DKIM build knob functional.
1056 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1057 Patch by Simon Arlott
1059 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1060 variable argument macros. Our copy now differs from upstream.
1066 TF/01 Failure to get a lock on a hints database can have serious
1067 consequences so log it to the panic log.
1069 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1070 controlled using the smtp_confirmation log selector.
1072 TF/03 Include the error message when we fail to unlink a spool file.
1074 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1075 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1076 for maintaining out-of-tree patches for some time.
1078 PP/01 Bugzilla 139: Documentation and portability issues.
1079 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1080 Handle per-OS dynamic-module compilation flags.
1082 PP/02 Let /dev/null have normal permissions.
1083 The 4.73 fixes were a little too stringent and complained about the
1084 permissions on /dev/null. Exempt it from some checks.
1085 Reported by Andreas M. Kirchwitz.
1087 PP/03 Report version information for many libraries, including
1088 Exim version information for dynamically loaded libraries. Created
1089 version.h, now support a version extension string for distributors
1090 who patch heavily. Dynamic module ABI change.
1092 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1093 privilege escalation vulnerability whereby the Exim run-time user
1094 can cause root to append content of the attacker's choosing to
1097 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1100 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1101 If dropping privileges for untrusted macros, we disabled normal logging
1102 on the basis that it would fail; for the Exim run-time user, this is not
1103 the case, and it resulted in successful deliveries going unlogged.
1104 Fixed. Reported by Andreas Metzler.
1110 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1111 only prepend for the Resent-* case. Fixes regression introduced in
1112 Exim 4.70 by NM/22 for Bugzilla 607.
1114 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1115 increasing numbers of administrators be bitten by this.
1117 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1119 PP/03 Bugzilla 994: added openssl_options main configuration option.
1121 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1123 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1125 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1127 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1128 without a peer certificate, leading to a segfault because of an
1129 assumption that peers always have certificates. Be a little more
1130 paranoid. Problem reported by Martin Tscholak.
1132 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1133 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1134 NB: ClamAV planning to remove STREAM in "middle of 2010".
1135 CL also introduces -bmalware, various -d+acl logging additions and
1136 more caution in buffer sizes.
1138 PP/09 Implemented reverse_ip expansion operator.
1140 PP/10 Bugzilla 937: provide a "debug" ACL control.
1142 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1144 PP/12 Bugzilla 973: Implement --version.
1146 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1148 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1150 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1152 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1153 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1154 ignore trailing whitespace.
1156 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1158 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1161 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1162 Notification from Dr Andrew Aitchison.
1164 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1165 ExtendedDetectionInfo response format.
1166 Notification from John Horne.
1168 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1171 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1172 XSL and documented dependency on system catalogs, with examples of how
1175 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1178 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1179 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1180 configuration file which is writeable by the Exim user or group.
1182 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1183 of configuration files to cover files specified with the -C option if
1184 they are going to be used with root privileges, not just the default
1187 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1188 option (effectively making it always true).
1190 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1191 files to be used while preserving root privileges.
1193 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1194 that rogue child processes cannot use them.
1196 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1197 run-time user, instead of root.
1199 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1200 Exim run-time user without dropping privileges.
1202 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1203 result string, instead of calling string_vformat() twice with the same
1206 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1207 for other users. Others should always drop root privileges if they use
1208 -C on the command line, even for a whitelisted configure file.
1210 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1212 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1218 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1219 $data_path, and $header_path variables; fixed documentation bugs and
1222 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1223 exipick to access non-standard spools, including the "frozen" queue
1226 NM/01 Bugzilla 965: Support mysql stored procedures.
1227 Patch from Alain Williams
1229 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1231 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1232 Patch from Andreas Metzler
1234 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1235 Patch from Kirill Miazine
1237 NM/05 Bugzilla 671: Added umask to procmail example.
1239 JJ/03 installed exipick 20100323.0, fixing doc bug
1241 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1242 directory. Notification and patch from Dan Rosenberg.
1244 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1246 TK/02 Improve log output when DKIM signing operation fails.
1248 MH/01 Treat the transport option dkim_domain as a colon separated
1249 list, not as a single string, and sign the message with each element,
1250 omitting multiple occurences of the same signer.
1252 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1253 Bugzilla 985, 986. Patch by Simon Arlott
1255 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1256 Patch by Simon Arlott
1258 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1259 MBX locking. Notification from Dan Rosenberg.
1265 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1267 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1269 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1271 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1273 NM/04 Bugzilla 915: Flag broken perl installation during build.
1279 TK/01 Added patch by Johannes Berg that expands the main option
1280 "spamd_address" if it starts with a dollar sign.
1282 TK/02 Write list of recipients to X-Envelope-Sender header when building
1283 the mbox-format spool file for content scanning (suggested by Jakob
1286 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1287 (http://www.dcc-servers.net/) support via dccifd. Activated by
1288 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1290 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1291 by Mark Daniel Reidel <mr@df.eu>.
1293 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1294 When building exim an external PCRE library is now needed -
1295 PCRE is a system library on the majority of modern systems.
1296 See entry on PCRE_LIBS in EDITME file.
1298 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1299 conversation. Added nologin parameter to request.
1300 Patch contributed by Kirill Miazine.
1302 TF/01 Do not log submission mode rewrites if they do not change the address.
1304 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1306 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1307 log files in place. Contributed by Roberto Lima.
1309 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1311 TF/03 Bugzilla 615: When checking the local_parts router precondition
1312 after a local_part_suffix or local_part_prefix option, Exim now
1313 does not use the address's named list lookup cache, since this
1314 contains cached lookups for the whole local part.
1316 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1317 Robert Millan. Documentation is in experimental-spec.txt.
1319 TF/04 Bugzilla 668: Fix parallel build (make -j).
1321 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1323 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1324 Patch provided by Jan Srzednicki.
1326 TF/05 Leading white space used to be stripped from $spam_report which
1327 wrecked the formatting. Now it is preserved.
1329 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1330 that they are available at delivery time.
1332 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1334 TF/08 TLS error reporting now respects the incoming_interface and
1335 incoming_port log selectors.
1337 TF/09 Produce a more useful error message if an SMTP transport's hosts
1338 setting expands to an empty string.
1340 NM/06 Bugzilla 744: EXPN did not work under TLS.
1341 Patch provided by Phil Pennock.
1343 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1344 Patch provided by Richard Godbee.
1346 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1347 acl_smtp_notquit, added index entry.
1349 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1350 Patch provided by Eugene Bujak.
1352 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1353 accept(). Patch provided by Maxim Dounin.
1355 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1356 Patch provided by Phil Pennock.
1358 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1360 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1361 Patch provided by Brad "anomie" Jorsch.
1363 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1364 Patch provided by Dean Brooks.
1366 TK/05 Add native DKIM support (does not depend on external libraries).
1368 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1369 Patch provided by Graeme Fowler.
1371 NM/16 Bugzilla 851: Documentation example syntax fix.
1373 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1375 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1378 NM/19 Bugzilla 745: TLS version reporting.
1379 Patch provided by Phil Pennock.
1381 NM/20 Bugzilla 167: bool: condition support.
1382 Patch provided by Phil Pennock.
1384 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1385 clients. Patch provided by Phil Pennock.
1387 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1388 Patch provided by Brad "anomie" Jorsch.
1390 NM/23 Bugzilla 687: Fix misparses in eximstats.
1391 Patch provided by Heiko Schlittermann.
1393 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1394 Patch provided by Heiko Schlittermann.
1396 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1397 plus update to original patch.
1399 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1401 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1402 Patch provided by David Brownlee.
1404 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1406 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1408 NM/30 Bugzilla 888: TLS documentation bugfixes.
1410 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1412 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1413 Unlike the original bugzilla I have changed all shell scripts in src tree.
1415 NM/33 Bugzilla 898: Transport filter timeout fix.
1416 Patch by Todd Rinaldo.
1418 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1419 Patch by Serge Demonchaux.
1421 NM/35 Bugzilla 39: Base64 decode bug fixes.
1422 Patch by Jakob Hirsch.
1424 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1426 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1428 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1430 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1436 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1437 ALT-N's libdkim that I have put here:
1438 http://duncanthrax.net/exim-experimental/
1440 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1441 were called 'true' and it seems that C99 defines that as a reserved
1442 keyword to be used with 'bool' variable types. That means you could
1443 not include C99-style headers which use bools without triggering
1444 build errors in sieve.c.
1446 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1447 as mailq or other aliases. Changed the --help handling significantly
1448 to do whats expected. exim_usage() emits usage/help information.
1450 SC/01 Added the -bylocaldomain option to eximstats.
1452 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1454 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1456 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1462 PH/01 Another patch from the Sieve maintainer.
1464 PH/02 When an IPv6 address is converted to a string for single-key lookup
1465 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1466 dots are used instead of colons so that keys in lsearch files need not
1467 contain colons. This was done some time before quoting was made available
1468 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1469 (notated using the quote facility) so as to distinguish them from IPv4
1470 keys. This meant that lookups for IP addresses in host lists did not work
1471 for iplsearch lookups.
1473 This has been fixed by arranging for IPv6 addresses to be expressed with
1474 colons if the lookup type is iplsearch. This is not incompatible, because
1475 previously such lookups could never work.
1477 The situation is now rather anomolous, since one *can* have colons in
1478 ordinary lsearch keys. However, making the change in all cases is
1479 incompatible and would probably break a number of configurations.
1481 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1484 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1485 conversion specification without a maximum field width, thereby enabling
1486 a rogue spamd server to cause a buffer overflow. While nobody in their
1487 right mind would setup Exim to query an untrusted spamd server, an
1488 attacker that gains access to a server running spamd could potentially
1489 exploit this vulnerability to run arbitrary code as the Exim user.
1491 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1492 $primary_hostname instead of what libspf2 thinks the hosts name is.
1494 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1495 a directory entry by the name of the lookup key. Previously, if a
1496 symlink pointed to a non-existing file or a file in a directory that
1497 Exim lacked permissions to read, a lookup for a key matching that
1498 symlink would fail. Now it is enough that a matching directory entry
1499 exists, symlink or not. (Bugzilla 503.)
1501 PH/03 The body_linecount and body_zerocount variables are now exported in the
1504 PH/04 Added the $dnslist_matched variable.
1506 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1507 This means they are set thereafter only if the connection becomes
1510 PH/06 Added the client_condition to authenticators so that some can be skipped
1511 by clients under certain conditions.
1513 PH/07 The error message for a badly-placed control=no_multiline_responses left
1514 "_responses" off the end of the name.
1516 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1518 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1519 (without spaces) instead of just copying the configuration text.
1521 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1523 PH/11 Added $max_received_linelength.
1525 PH/12 Added +ignore_defer and +include_defer to host lists.
1527 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1528 way in which PCRE > 7.0 is built.
1530 PH/14 Implemented queue_only_load_latch.
1532 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1533 MAIL command. The effect was to mangle the value on 64-bit systems.
1535 PH/16 Another patch from the Sieve maintainer.
1537 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1539 PH/18 If a system quota error occurred while trying to create the file for
1540 a maildir delivery, the message "Mailbox is full" was not appended to the
1541 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1542 only to a quota excession during the actual writing of the file.
1544 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1545 characters?) which causes problems in log lines. The DN values are now
1546 passed through string_printing() before being added to log lines.
1548 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1549 and InterBase are left for another time.)
1551 PH/21 Added message_body_newlines option.
1553 PH/22 Guard against possible overflow in moan_check_errorcopy().
1555 PH/23 POSIX allows open() to be a macro; guard against that.
1557 PH/24 If the recipient of an error message contained an @ in the local part
1558 (suitably quoted, of course), incorrect values were put in $domain and
1559 $local_part during the evaluation of errors_copy.
1565 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1566 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1569 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1570 issue a MAIL command.
1572 PH/02 In an ACL statement such as
1574 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1576 if a client was not listed at all, or was listed with a value other than
1577 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1578 the condition was not true (as it should be), so access was not denied.
1579 The bug was that the ! inversion was incorrectly passed on to the second
1580 item. This has been fixed.
1582 PH/03 Added additional dnslists conditions == and =& which are different from
1583 = and & when the dns lookup returns more than one IP address.
1585 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1586 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1588 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1589 FSYNC, which compiles an option called disable_fsync that allows for
1590 bypassing fsync(). The documentation is heavily laced with warnings.
1592 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1594 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1595 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1596 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1597 including adding "make clean"; (3) Added -fPIC when compiling the test
1598 dynamically loaded module, to get rid of a warning.
1600 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1601 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1602 = 0s. The bug is otherwise harmless.
1604 PH/07 There was a bug in the dovecot authenticator such that the value of
1605 $auth1 could be overwritten, and so not correctly preserved, after a
1606 successful authentication. This usually meant that the value preserved by
1607 the server_setid option was incorrect.
1609 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1611 PH/09 Installed PCRE release 7.0.
1613 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1614 run for batched SMTP input. It is now run at the start of every message
1615 in the batch. While fixing this I discovered that the process information
1616 (output by running exiwhat) was not always getting set for -bs and -bS
1617 input. This is fixed, and it now also says "batched" for BSMTP.
1619 PH/11 Added control=no_pipelining.
1621 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1622 patch, slightly modified), and move the expansion of helo_data till after
1623 the connection is made in the smtp transport (so it can use these
1626 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1628 PH/14 Added log_selector = +pid.
1630 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1632 PH/16 Add ${if forany and ${if forall.
1634 PH/17 Added dsn_from option to vary the From: line in DSNs.
1636 PH/18 Flush SMTP output before performing a callout, unless control =
1637 no_callout_flush is set.
1639 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1640 was true (the default) a successful delivery failed to delete the retry
1641 item, thus causing premature timeout of the address. The bug is now
1644 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1646 PH/21 Long custom messages for fakedefer and fakereject are now split up
1647 into multiline reponses in the same way that messages for "deny" and
1648 other ACL rejections are.
1650 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1651 with slight modification.
1653 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1654 draft, changing the syntax and factoring some duplicate code".
1656 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1657 for deliveries of the second and subsequent messages over the same SMTP
1660 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1661 ${reduce, with only minor "tidies".
1663 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1665 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1666 expansion side effects.
1668 PH/27 When a message times out after an over-quota error from an Exim-imposed
1669 quota, the bounce message says "mailbox is full". This message was not
1670 being given when it was a system quota that was exceeded. It now should
1673 MH/03 Made $recipients available in local_scan(). local_scan() already has
1674 better access to the recipient list through recipients_list[], but
1675 $recipients can be useful in postmaster-provided expansion strings.
1677 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1678 in the case of a MAIL command with additional options following the
1679 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1680 were accidentally chopped off.
1682 PH/29 SMTP synchronization checks are implemented when a command is read -
1683 there is a check that no more input is waiting when there shouldn't be
1684 any. However, for some commands, a delay in an ACL can mean that it is
1685 some time before the response is written. In this time, more input might
1686 arrive, invalidly. So now there are extra checks after an ACL has run for
1687 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1688 pipelining has not been advertised.
1690 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1692 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1693 returned in the SMTP response when smtp_return_error_details was false.
1694 This has been fixed.
1696 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1697 instead of the C I/O that was originally supplied, because problems were
1698 reported on Solaris.
1700 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1701 Exim which did not show up earlier: it was assuming that a call to
1702 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1703 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1704 was a macro that became an assignment, so it seemed to work. This has
1705 changed to a proper function call with a void return, hence the compile
1706 error. Exim's code has been fixed.
1708 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1711 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1713 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1715 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1716 because it turns out to be too big or there is a timeout) there is no
1717 "Completed" line in the log. When some messages of this type were
1718 selected by exigrep, they were listed as "not completed". Others were
1719 picked up by some special patterns. I have improved the selection
1720 criteria to be more general.
1722 PH/38 The host_find_failed option in the manualroute router can now be set
1723 to "ignore", to completely ignore a host whose IP address cannot be
1724 found. If all hosts are ignored, the behaviour is controlled by the new
1725 host_all_ignored option.
1727 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1728 homing or because of multiple MX records with /mx) generated more than
1729 one IP address, and the following item turned out to be the local host,
1730 all the secondary addresses of the first item were incorrectly removed
1731 from the list, along with the local host and any following hosts (which
1732 is what is supposed to happen).
1734 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1735 whoever called Exim into the -H file. In the case of the daemon it was
1736 behaving confusingly. When first started, it used values for whoever
1737 started the daemon, but after a SIGHUP it used the Exim user (because it
1738 calls itself on a restart). I have changed the code so that it now always
1741 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1742 message are rejected with the same error (e.g. no authentication or bad
1743 sender address), and a DATA command is nevertheless sent (as can happen
1744 with PIPELINING or a stupid MUA), the error message that was given to the
1745 RCPT commands is included in the rejection of the DATA command. This is
1746 intended to be helpful for MUAs that show only the final error to their
1749 PH/42 Another patch from the Sieve maintainer.
1751 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1752 Eximstats - Fixed some broken HTML links and added missing column headers
1754 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1755 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1757 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1759 PH/43 Yet another patch from the Sieve maintainer.
1761 PH/44 I found a way to check for a TCP/IP connection going away before sending
1762 the response to the final '.' that terminates a message, but only in the
1763 case where the client has not sent further data following the '.'
1764 (unfortunately, this is allowed). However, in many cases there won't be
1765 any further data because there won't be any more messages to send. A call
1766 to select() can be used: if it shows that the input is "ready", there is
1767 either input waiting, or the socket has been closed. An attempt to read
1768 the next input character can distinguish the two cases. Previously, Exim
1769 would have sent an OK response which the client would never have see.
1770 This could lead to message repetition. This fix should cure that, at
1771 least in a lot of common cases.
1773 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1774 advertised in response to EHLO.
1780 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1781 fixed by 4.65/MH/01 (is this a record?) are fixed:
1783 (i) An empty string was always treated as zero by the numeric comparison
1784 operators. This behaviour has been restored.
1786 (ii) It is documented that the numeric comparison operators always treat
1787 their arguments as decimal numbers. This was broken in that numbers
1788 starting with 0 were being interpreted as octal.
1790 While fixing these problems I realized that there was another issue that
1791 hadn't been noticed. Values of message_size_limit (both the global option
1792 and the transport option) were treated as octal if they started with 0.
1793 The documentation was vague. These values are now always treated as
1794 decimal, and I will make that clear in the documentation.
1800 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1801 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1804 MH/01 Don't check that the operands of numeric comparison operators are
1805 integers when their expansion is in "skipping" mode (fixes bug
1806 introduced by 4.64-PH/07).
1808 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1809 child addresses, Exim now panics and dies. Previously, because the count
1810 is held in a short int, deliveries were likely to be lost. As such a
1811 large number of recipients for a single message is ridiculous
1812 (performance will be very, very poor), I have chosen to impose a limit
1813 rather than extend the field.
1819 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1820 leftover -K file (the existence of which was triggered by #402).
1821 While we were at it, introduced process PID as part of the -K
1822 filename. This should rule out race conditions when creating
1825 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1826 processing considerably. Previous code took too long for large mails,
1827 triggering a timeout which in turn triggers #401.
1829 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1830 in the DK code in transports.c. sendfile() is not really portable,
1831 hence the _LINUX specificness.
1833 TF/01 In the add_headers option to the mail command in an Exim filter,
1834 there was a bug that Exim would claim a syntax error in any
1835 header after the first one which had an odd number of characters
1838 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1839 callout verification, Exim cached a "reject" for the entire domain. This
1840 is correct for most verifications, but it is not correct for a recipient
1841 verification with use_sender or use_postmaster set, because in that case
1842 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1843 case of MAIL FROM:<> rejection from other early rejections (e.g.
1844 rejection of HELO). When verifying a recipient using a non-null MAIL
1845 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1846 Whatever the result of the callout, the value of the domain cache is
1847 left unchanged (for any other kind of callout, getting as far as trying
1848 RCPT means that the domain itself is ok).
1850 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1853 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1854 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1855 session, the daemon ignores ECONNECT errors and logs others; it now
1856 ignores EPIPE as well.
1858 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1859 (quoted-printable decoding).
1861 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1862 later the small subsequent patch to fix an introduced bug.
1864 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1866 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1868 PH/08 An error is now given if message_size_limit is specified negative.
1870 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1871 to be given (somewhat) arbitrary names.
1873 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1876 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1877 miscellaneous code fixes
1879 PH/10 Added the log_reject_target ACL modifier to specify where to log
1882 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1883 hostname. This is wrong, because it relates to the incoming message (and
1884 probably the interface on which it is arriving) and not to the outgoing
1885 callout (which could be using a different interface). This has been
1886 changed to use the value of the helo_data option from the smtp transport
1887 instead - this is what is used when a message is actually being sent. If
1888 there is no remote transport (possible with a router that sets up host
1889 addresses), $smtp_active_hostname is used.
1891 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1892 tweaks were necessary in order to get it to work (see also 21 below):
1893 (a) The code assumed that strncpy() returns a negative number on buffer
1894 overflow, which isn't the case. Replaced with Exim's string_format()
1896 (b) There were several signed/unsigned issues. I just did the minimum
1897 hacking in of casts. There is scope for a larger refactoring.
1898 (c) The code used strcasecmp() which is not a standard C function.
1899 Replaced with Exim's strcmpic() function.
1900 (d) The code set only $1; it now sets $auth1 as well.
1901 (e) A simple test gave the error "authentication client didn't specify
1902 service in request". It would seem that Dovecot has changed its
1903 interface. Fortunately there's a specification; I followed it and
1904 changed what the client sends and it appears to be working now.
1906 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1909 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1910 address A is aliased to B and C, where B exists and C does not. Without
1911 -v the output is "A verified" because verification stops after a
1912 successful redirection if more than one address is generated. However,
1913 with -v the child addresses are also verified. Exim was outputting "A
1914 failed to verify" and then showing the successful verification for C,
1915 with its parentage. It now outputs "B failed to verify", showing B's
1916 parentage before showing the successful verification of C.
1918 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1919 look up a TXT record in a specific list after matching in a combined
1922 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1923 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1924 they consult the DNS. I had assumed they would set it the way they
1925 wanted; and indeed my experiments on Linux seem to show that in some
1926 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1927 To be on the safe side, however, I have now made the interface to
1928 host_find_byname() similar to host_find_bydns(), with an argument
1929 containing the DNS resolver options. The host_find_byname() function now
1930 sets these options at its start, just as host_find_bydns() does. The smtp
1931 transport options dns_qualify_single and dns_search_parents are passed to
1932 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1933 of host_find_byname() use the default settings of RES_DEFNAMES
1934 (qualify_single) but not RES_DNSRCH (search_parents).
1936 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1937 spool_read_header() do less string testing, by means of a preliminary
1938 switch on the second character of optional "-foo" lines. (This is
1939 overdue, caused by the large number of possibilities that now exist.
1940 Originally there were few.) While I was there, I also converted the
1941 str(n)cmp tests so they don't re-test the leading "-" and the first
1942 character, in the hope this might squeeze out yet more improvement.
1944 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1945 flag allowing group syntax was set by the header_syntax check but not
1946 turned off, possible causing trouble later; (2) The flag was not being
1947 set at all for the header_verify test, causing "group"-style headers to
1948 be rejected. I have now set it in this case, and also caused header_
1949 verify to ignore an empty address taken from a group. While doing this, I
1950 came across some other cases where the code for allowing group syntax
1951 while scanning a header line wasn't quite right (mostly, not resetting
1952 the flag correctly in the right place). These bugs could have caused
1953 trouble for malformed header lines. I hope it is now all correct.
1955 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1956 with the "reply" argument non-NULL. The code, however (which originally
1957 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1958 but it didn't always do it. This confused somebody who was copying the
1959 code for some other use. I have removed all the tests.
1961 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1962 feature that was used to support insecure browsers during the U.S. crypto
1963 embargo. It requires special client support, and Exim is probably the
1964 only MTA that supported it -- and would never use it because real RSA is
1965 always available. This code has been removed, because it had the bad
1966 effect of slowing Exim down by computing (never used) parameters for the
1967 RSA_EXPORT functionality.
1969 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1970 authenticator to fail if there's a tab character in the incoming data
1971 (there should never be unless someone is messing about, as it's supposed
1972 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1973 if the connection is using TLS or if the remote IP is the same as the
1974 local IP, and the "valid-client-cert option" if a client certificate has
1977 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1978 authenticators. This can be used for authorization after authentication
1979 succeeds. (In the case of plaintext, it servers for both authentication
1982 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1983 if any retry times were supplied.
1985 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1986 connection, where there is no client IP address to check. In this
1987 situation, the verify now always succeeds.
1989 PH/25 Applied John Jetmore's -Mset patch.
1991 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1993 PH/27 In a string expansion for a processed (not raw) header when multiple
1994 headers of the same name were present, leading whitespace was being
1995 removed from all of them, but trailing whitespace was being removed only
1996 from the last one. Now trailing whitespace is removed from each header
1997 before concatenation. Completely empty headers in a concatenation (as
1998 before) are ignored.
2000 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2001 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2003 PH/29 [Removed. This was a change that I later backed out, and forgot to
2004 correct the ChangeLog entry (that I had efficiently created) before
2005 committing the later change.]
2007 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2008 address errors (4xx response to RCPT) over the same connection as other
2009 messages routed to the same hosts. Such deliveries are always "forced",
2010 so retry times are not inspected. This resulted in far too many retries
2011 for the affected addresses. The effect occurred only when there were more
2012 hosts than the hosts_max_try setting in the smtp transport when it had
2013 the 4xx errors. Those hosts that it had tried were not added to the list
2014 of hosts for which the message was waiting, so if all were tried, there
2015 was no problem. Two fixes have been applied:
2017 (i) If there are any address or message errors in an SMTP delivery, none
2018 of the hosts (tried or untried) are now added to the list of hosts
2019 for which the message is waiting, so the message should not be a
2020 candidate for sending over the same connection that was used for a
2021 successful delivery of some other message. This seems entirely
2022 reasonable: after all the message is NOT "waiting for some host".
2023 This is so "obvious" that I'm not sure why it wasn't done
2024 previously. Hope I haven't missed anything, but it can't do any
2025 harm, as the worst effect is to miss an optimization.
2027 (ii) If, despite (i), such a delivery is accidentally attempted, the
2028 routing retry time is respected, so at least it doesn't keep
2029 hammering the server.
2031 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2032 in ${readsocket because some servers need this prod.
2034 PH/32 Added some extra debug output when updating a wait-xxx database.
2036 PH/33 The hint "could be header name not terminated by colon", which has been
2037 given for certain expansion errors for a long time, was not being given
2038 for the ${if def:h_colon_omitted{... case.
2040 PH/34 The spec says: "With one important exception, whenever a domain list is
2041 being scanned, $domain contains the subject domain." There was at least
2042 one case where this was not true.
2044 PH/35 The error "getsockname() failed: connection reset by peer" was being
2045 written to the panic log as well as the main log, but it isn't really
2046 panic-worthy as it just means the connection died rather early on. I have
2047 removed the panic log writing for the ECONNRESET error when getsockname()
2050 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2051 runs only) independently of the message's sender address. This meant
2052 that, if the 4xx error was in fact related to the sender, a different
2053 message to the same recipient with a different sender could confuse
2054 things. In particualar, this can happen when sending to a greylisting
2055 server, but other circumstances could also provoke similar problems.
2056 I have changed the default so that the retry time for these errors is now
2057 based a combination of the sender and recipient addresses. This change
2058 can be overridden by setting address_retry_include_sender=false in the
2061 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2062 remote server are returned as part of bounce messages. This was not
2063 happening for LMTP over a pipe (the lmtp transport), but now it is the
2064 same for both kinds of LMTP.
2066 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2067 in header lines that were in fact CNAMEs. This is no longer the case.
2069 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2070 and queue runs started by the daemon processed all messages. This has
2071 been fixed so that -R and -S can now usefully be given with -q<time>.
2073 PH/40 Import PCRE release 6.7 (fixes some bugs).
2075 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2077 PH/42 Give an error if -q is specified more than once.
2079 PH/43 Renamed the variables $interface_address and $interface_port as
2080 $received_ip_address and $received_port, to make it clear that these
2081 values apply to message reception, and not to the outgoing interface when
2082 a message is delivered. (The old names remain recognized, of course.)
2084 PH/44 There was no timeout on the connect() call when using a Unix domain
2085 socket in the ${readsocket expansion. There now is.
2087 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2088 be meaningful with "accept".
2090 SC/01 Eximstats V1.43
2091 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2093 SC/02 Eximstats V1.44
2094 Use a glob alias rather than an array ref in the generated
2095 parser. This improves both readability and performance.
2097 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2098 Collect SpamAssassin and rejection statistics.
2099 Don't display local sender or destination tables unless
2100 there is data to show.
2101 Added average volumes into the top table text output.
2103 SC/04 Eximstats V1.46
2104 Collect data on the number of addresses (recipients)
2105 as well as the number of messages.
2107 SC/05 Eximstats V1.47
2108 Added 'Message too big' to the list of mail rejection
2109 reasons (thanks to Marco Gaiarin).
2111 SC/06 Eximstats V1.48
2112 Mainlog lines which have GMT offsets and are too short to
2113 have a flag are now skipped.
2115 SC/07 Eximstats V1.49 (Alain Williams)
2116 Added the -emptyok flag.
2118 SC/08 Eximstats V1.50
2119 Fixes for obtaining the IP address from reject messages.
2121 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2122 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2123 whitesspace changes from 4.64-PH/27
2125 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2128 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2129 are found, allow negative numbers in numeric criteria)
2131 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2133 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2134 to match changes made in 4.64-PH/43
2136 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2138 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2139 "domain" is actually an IP address, and force a failure. This locks out
2140 those revolvers/nameservers that support "A-for-A" lookups, in
2141 contravention of the specifications.
2143 PH/48 When a host name was looked up from an IP address, and the subsequent
2144 forward lookup of the name timed out, the host name was left in
2145 $sender_host_name, contrary to the specification.
2147 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2148 restricted to single-key lookups, Exim was not diagnosing an error if
2149 * or *@ was used with a query-style lookup.
2151 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2153 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2154 long ago, but noone interested enough thought of it. Let's just say that
2155 the "1.1" means that there are some new functions that weren't there at
2156 some point in the past.
2158 PH/51 Error processing for expansion failure of helo_data from an smtp
2159 transport during callout processing was broken.
2161 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2162 tested/used via the -bh/-bhc/-bs options.
2164 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2165 bug, fixed in subsequent PCRE releases).
2167 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2168 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2170 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2176 SC/01 Use a glob alias rather than an array ref in eximstats generated
2177 parser. This improves both readability and performance.
2179 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2180 Don't display local sender or destination tables in eximstats unless
2181 there is data to show.
2182 Added average volumes into the eximstats top table text output.
2184 SC/03 Collect data on the number of addresses (recipients) as well
2185 as the number of messages in eximstats.
2187 TF/01 Correct an error in the documentation for the redirect router. Exim
2188 does (usually) call initgroups() when daemonizing.
2190 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2191 with consistent privilege compared to when running as a daemon.
2193 TF/03 Note in the spec that $authenticated_id is not set for local
2194 submissions from trusted users.
2196 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2197 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2199 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2200 by adding some example configuration directives to the default
2201 configuration file. A little bit of work is required to uncomment the
2202 directives and define how usernames and passwords are checked, but
2203 there is now a framework to start from.
2205 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2206 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2207 without this. I don't know how relevant this is to other LDAP libraries.
2209 PH/02 Add the verb name to the "unknown ACL verb" error.
2211 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2213 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2215 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2216 directory not expanded when it should be if an expanded home directory
2217 was set for the address (which is overridden by the transport).
2219 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2222 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2223 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2224 because it is too late at that time, and has no effect.
2226 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2227 security issue with \' (bugzilla #107). I could not use the
2228 PQescapeStringConn() function, because it needs a PGconn value as one of
2231 PH/08 When testing addresses using -bt, indicate those final addresses that
2232 are duplicates that would not cause an additional delivery. At least one
2233 person was confused, thinking that -bt output corresponded to deliveries.
2234 (Suppressing duplicates isn't a good idea as you lose the information
2235 about possibly different redirections that led to the duplicates.)
2237 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2238 systems where poll() doesn't work, in particular OS X.
2240 PH/10 Added more information to debugging output for retry time not reached.
2242 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2243 operations in malware.c.
2245 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2248 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2249 syslog_duplication set false, log lines that would normally be written
2250 both the the main log and to the reject log were not written to syslog at
2253 PH/14 In the default configuration, change the use of "message" in ACL warn
2254 statements to "add_header".
2256 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2257 not followed by a command (e.g. "seen endif").
2259 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2260 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2263 PH/17 Added extra conditions to the default value of delay_warning_condition
2267 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2268 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2269 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2272 The Auto-Submitted: and various List- headers are standardised, whereas I
2273 don't think Precedence: ever was.
2275 PH/18 Refactored debugging code in route_finduser() to show more information,
2276 in particular, the error code if getpwnam() issues one.
2278 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2279 This is apparently needed in addition to the PH/07 change above to avoid
2280 any possible encoding problems.
2282 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2283 but not after initializing Perl.
2285 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2286 output them only if debugging. By default they are written stderr,
2287 apparently, which is not desirable.
2289 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2292 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2295 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2297 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2298 authenticated or an ident call has been made. Suppress the default
2299 values for $authenticated_id and $authenticated_sender (but permit -oMai
2300 and -oMas) when testing with -bh.
2302 PH/24 Re-jigged the order of the tests in the default configuration so that the
2303 tests for valid domains and recipients precede the DNS black list and CSA
2304 tests, on the grounds that those ones are more expensive.
2306 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2307 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2308 This bug exists in every version of Exim that I still have, right back to
2311 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2312 However, an attempt to turn on case-sensitivity in a regex key by
2313 including (?-i) didn't work because the subject string was already
2314 lowercased, and the effects were non-intuitive. It turns out that a
2315 one-line patch can be used to allow (?-i) to work as expected.
2321 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2322 other effects) broke the use of negated acl sub-conditions.
2324 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2327 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2328 "Deny" causes Exim to reject the incoming connection with a 554 error.
2329 Unfortunately, if there is a major crisis, such as a disk failure,
2330 tcp-wrappers gives "deny", whereas what one would like would be some
2331 kind of temporary error. A kludge has been added to help with this.
2332 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2333 554 error is used if errno is still zero or contains ENOENT (which occurs
2334 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2337 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2339 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2340 errors. Otherwise a message that provokes a temporary error (when other
2341 messages do not) can cause a whole host to time out.
2343 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2344 addresses were routed directly to files or pipes from a redirect router.
2345 File deliveries just didn't batch; pipe deliveries might have suffered
2348 PH/06 A failure to get a lock for a hints database would erroneously always say
2349 "Failed to get write lock", even when it was really a read lock.
2351 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2352 of 0600. This has been changed to use the value of the lockfile_mode
2353 option (which defaults to 0600).
2355 PH/08 Applied small patch from the Sieve maintainer.
2357 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2358 folder from quota calculations, a direct delivery into this folder messed
2359 up the contents of the maildirsize file. This was because the regex was
2360 used only to exclude .Trash (or whatever) when the size of the mailbox
2361 was calculated. There was no check that a delivery was happening into an
2362 excluded directory. This bug has been fixed by ignoring all quota
2363 processing for deliveries into excluded directories.
2365 PH/10 Added the maildirfolder_create_regex option to appendfile.
2371 PH/01 The code for finding all the local interface addresses on a FreeBSD
2372 system running IPv6 was broken. This may well have applied to all BSD
2373 systems, as well as to others that have similar system calls. The broken
2374 code found IPv4 interfaces correctly, but gave incorrect values for the
2375 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2376 that it would not match correctly against @[] and not recognize the IPv6
2379 PH/02 The ipliteral router was not recognizing addresses of the form user@
2380 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2382 PH/03 Added disable_ipv6.
2384 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2385 decoded form, because it is most often used to construct To: headers
2386 lines in autoreplies, and the decoded form may well be syntactically
2387 invalid. However, $reply_address has leading white space removed, and all
2388 newlines turned into spaces so that the autoreply transport does not
2391 PH/05 If group was specified without a user on a router, and no group or user
2392 was specified on a transport, the group from the router was ignored.
2394 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2395 for visible compile-time settings that can be used to change these
2396 numbers, for those that want even more. Backwards compatibility with old
2397 spool files has been maintained. However, going back to a previous Exim
2398 release will lost any variables that are in spool files.
2400 PH/07 Two small changes when running in the test harness: increase delay when
2401 passing a TCP/IP connection to a new process, in case the original
2402 process has to generate a bounce, and remove special handling of
2403 127.0.0.2 (sic), which is no longer necessary.
2405 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2406 be the same on different OS.
2408 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2411 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2412 whether --show-vars was specified or not
2414 JJ/02 exipick: Added support for new ACL variable spool format introduced
2417 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2418 syntactically invalid From: or Reply-to: line, and a filter used this to
2419 generate an autoreply, and therefore failed to obtain an address for the
2420 autoreply, Exim could try to deliver to a non-existent relative file
2421 name, causing unrelated and misleading errors. What now happens is that
2422 it logs this as a hard delivery error, but does not attempt to create a
2425 PH/11 The exinext utility has a -C option for testing purposes, but although
2426 the given file was scanned by exinext itself; it wasn't being passed on
2427 when Exim was called.
2429 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2430 an end-of-file indication when reading a command response.
2432 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2433 compiled. In many other places in Exim, IPv6 addresses are always
2434 recognized, so I have changed this. It also means that IPv4 domain
2435 literals of the form [IPV4:n.n.n.n] are now always recognized.
2437 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2438 used if the router is not running as root, for example, when verifying at
2439 ACL time, or when using -bh. The debugging output from this situation was
2440 non-existent - all you got was a failure to exec. I have made two
2443 (a) Failures to set uid/gid, the current directory, or a process leader
2444 in a subprocess such as that created by queryprogram now generate
2445 suitable debugging ouput when -d is set.
2447 (b) The queryprogram router detects when it is not running as root,
2448 outputs suitable debugging information if -d is set, and then runs
2449 the subprocess without attempting to change uid/gid.
2451 PH/15 Minor change to Makefile for building test_host (undocumented testing
2454 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2455 additional section of a DNS packet that returns MX or SRV records.
2456 Instead, it always explicitly searches for A/AAAA records. This avoids
2457 major problems that occur when a DNS server includes only records of one
2458 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2459 fixed another bug: if SRV records were looked up and the corresponding
2460 address records were *not* found in the additional section, the port
2461 values from the SRV records were lost.
2463 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2464 using the correct key (the original address) when searching the retry
2465 rules in order to find which one to use for generating the retry hint.
2467 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2468 adding the default one. Similarly, if it contains a Reply-To: header, the
2469 errors_reply_to option, if set, is not used.
2471 PH/19 When calculating a retry time, Exim used to measure the "time since
2472 failure" by looking at the "first failed" field in the retry record. Now
2473 it does not use this if it is later than than the arrival time of the
2474 message. Instead it uses the arrival time. This makes for better
2475 behaviour in cases where some deliveries succeed, thus re-setting the
2476 "first failed" field. An example is a quota failure for a huge message
2477 when small messages continue to be delivered. Without this change, the
2478 "time since failure" will always be short, possible causing more frequent
2479 delivery attempts for the huge message than are intended.
2480 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2482 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2483 $1, $2, $3) because the numerical variables can be reset during some
2484 expansion items (e.g. "match"), thereby losing the authentication data.
2486 PH/21 Make -bV show the size of off_t variables so that the test suite can
2487 decide whether to run tests for quotas > 2G.
2489 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2490 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2491 filecount value is greater than 2G or if a quota value is greater than 2G
2492 on a system where the size of off_t is not greater than 4, a panic error
2495 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2496 never match. The debug and -bh output now contains an explicit error
2497 message indicating a malformed IPv4 address or mask.
2499 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2500 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2501 PH/23 above applies.
2503 PH/25 Do not write to syslog when running in the test harness. The only
2504 occasion when this arises is a failure to open the main or panic logs
2505 (for which there is an explicit test).
2507 PH/26 Added the /no_tell option to "control=freeze".
2509 PH/27 If a host name lookup failed very early in a connection, for example, if
2510 the IP address matched host_lookup and the reverse lookup yielded a name
2511 that did not have a forward lookup, an error message of the form "no IP
2512 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2513 could be logged. Now it outputs the IP address instead of "NULL".
2515 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2516 allows the sender and the authenticated sender to be set when
2517 submitting a message from within Exim. Since child_open_exim() is
2518 documented for local_scan(), the new function should be too.
2520 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2521 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2522 results in an empty string is now treated as unset.
2524 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2526 PH/31 Added sender_verify_fail logging option.
2528 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2529 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2530 tidied the source and removed it altogether.
2532 PH/33 When a queue run was abandoned because the load average was too high, a
2533 log line was always written; now it is written only if the queue_run log
2534 selector is set. In addition, the log line for abandonment now contains
2535 information about the queue run such as the pid. This is always present
2536 in "start" and "stop" lines but was omitted from the "abandon" line.
2538 PH/34 Omit spaces between a header name and the colon in the error message that
2539 is given when verify = headers_syntax fails (if there are lots of them,
2540 the message gets confusing).
2542 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2543 names, as there are now some PTR records that contain slashes. This check
2544 is only to protect against broken name servers that fall over on strange
2545 characters, so the fact that it applies to all lookups doesn't matter.
2547 PH/36 Now that the new test suite is complete, we can remove some of the
2548 special code in Exim that was needed for the old test suite. For example,
2549 sorting DNS records because real resolvers return them in an arbitrary
2550 order. The new test suite's fake resolver always returns records in the
2553 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2554 bounces) except when queue_only is set, to avoid logging races between
2555 the different processes.
2557 PH/38 Panic-die if .include specifies a non-absolute path.
2559 PH/39 A tweak to the "H" retry rule from its user.
2561 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2562 a label. They prevented compilation on older perls.
2564 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2565 a warning to be raised on newish perls.
2567 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2568 on queue. Changes to match documented behaviour of showing count of
2569 messages matching specified criteria.
2571 PH/40 Changed the default ident timeout from 30s to 5s.
2573 PH/41 Added support for the use of login_cap features, on those BSD systems
2574 that have them, for controlling the resources used by pipe deliveries.
2576 PH/42 The content-scanning code uses fopen() to create files in which to put
2577 message data. Previously it was not paying any attention to the mode of
2578 the files. Exim runs with umask(0) because the rest of the code creates
2579 files with open(), and sets the required mode explicitly. Thus, these
2580 files were ending up world-writeable. This was not a big issue, because,
2581 being within the spool directory, they were not world-accessible. I have
2582 created a function called modefopen, which takes an additional mode
2583 argument. It sets umask(777), creates the file, chmods it to the required
2584 mode, then resets the umask. All the relevant calls to fopen() in the
2585 content scanning code have been changed to use this function.
2587 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2588 to 24 hours. This avoids potential overflow problems when processing G
2589 and H retry rules. I suspect nobody ever tinkers with this value.
2591 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2593 PH/45 When the plaintext authenticator is running as a client, the server's
2594 challenges are checked to ensure they are valid base64 strings. By
2595 default, the authentication attempt is cancelled if an invalid string is
2596 received. Setting client_ignore_invalid_base64 true ignores these errors.
2597 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2598 they are received. Thus, the responses can be made to depend on the
2599 challenges. If an invalid string is ignored, an empty string is placed in
2602 PH/46 Messages that are created by the autoreply transport now contains a
2603 References: header, in accordance with RFCs 2822 and 3834.
2605 PH/47 Added authenticated_sender_force to the smtp transport.
2607 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2609 PH/49 Installed latest patch from the Sieve maintainer.
2611 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2612 was also set, the appendfile transport was unnecessarily scanning a
2613 directory of message files (e.g. for maildir delivery) to find the count
2614 of files (along with the size), even though it did not need this
2615 information. It now does the scan only if it needs to find either the
2616 size of the count of files.
2618 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2620 PH/52 Two bugs concerned with error handling when the smtp transport is
2623 (i) Exim was not creating retry information for temporary errors given
2624 for individual recipients after the DATA command when the smtp transport
2625 was used in LMTP mode. This meant that they could be retried too
2626 frequently, and not timed out correctly.
2628 (ii) Exim was setting the flag that allows error details to be returned
2629 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2630 recipients that were returned after the DATA command.
2632 PH/53 This is related to PH/52, but is more general: for any failing address,
2633 when detailed error information was permitted to be returned to the
2634 sender, but the error was temporary, then after the final timeout, only
2635 "retry timeout exceeded" was returned. Now it returns the full error as
2636 well as "retry timeout exceeded".
2638 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2639 do this, and (what is worse) MTAs that accept it.
2641 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2642 will now be deprecated.
2644 PH/56 New os.c-cygwin from the Cygwin maintainer.
2646 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2647 formats (previously only available in exim formats via -bpr, -bpru,
2648 and -bpra. Now also available in native and exiqgrep formats)
2650 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2651 with very large, slow to parse queues
2653 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2655 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2657 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2658 responses to authentication challenges, though it was showing the
2659 challenges; (ii) I've removed the CR characters from the debug output for
2662 PH/58 Allow for the insertion of a newline as well as a space when a string
2663 is turned into more than one encoded-word during RFC 2047 encoding. The
2664 Sieve code now uses this.
2666 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2667 data_4xx, lost_connection, tls_required.
2669 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2670 message was being sent as an SMTP response.
2672 PH/61 Add -l and -k options to exicyclog.
2674 PH/62 When verifying, if an address was redirected to one new address, so that
2675 verification continued, and the new address failed or deferred after
2676 having set something in $address_data, the value of $address_data was not
2677 passed back to the ACL. This was different to the case when no
2678 redirection occurred. The value is now passed back in both cases.
2680 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2681 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2682 use login_cap.h, so on its own it isn't the distinguishing feature. The
2683 new name refers directly to the setclassresources() function.
2685 PH/65 Added configuration files for NetBSD3.
2687 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2689 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2690 is preferred over IPv4.
2692 PH/68 The bounce_return_message and bounce_return_body options were not being
2693 honoured for bounces generated during the reception of non-SMTP messages.
2694 In particular, this applied to messages rejected by the ACL. This bug has
2695 been fixed. However, if bounce_return_message is true and bounce_return_
2696 body is false, the headers that are returned for a non-SMTP message
2697 include only those that have been read before the error was detected.
2698 (In the case of an ACL rejection, they have all been read.)
2700 PH/69 The HTML version of the specification is now built in a directory called
2701 spec_html instead of spec.html, because the latter looks like a path with
2702 a MIME-type, and this confuses some software.
2704 PH/70 Catch two compiler warnings in sieve.c.
2706 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2707 function verify_get_ident() calls ip_connect() to connect a socket, but
2708 if the "connect()" function timed out, ip_connect() used to close the
2709 socket. However, verify_get_ident() also closes the socket later, and in
2710 between Exim writes to the log, which may get opened at this point. When
2711 the socket was closed in ip_connect(), the log could get the same file
2712 descriptor number as the socket. This naturally causes chaos. The fix is
2713 not to close the socket in ip_connect(); the socket should be closed by
2714 the function that creates it. There was only one place in the code where
2715 this was missing, in the iplookup router, which I don't think anybody now
2716 uses, but I've fixed it anyway.
2718 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2719 well as to direct DNS lookups. Otherwise the handling of names in host
2720 lists is inconsistent and therefore confusing.
2726 PH/01 Two changes to the default runtime configuration:
2728 (1) Move the checks for relay_from_hosts and authenticated clients from
2729 after to before the (commented out) DNS black list checks.
2731 (2) Add control=submission to the relay_from_hosts and authenticated
2732 clients checks, on the grounds that messages accepted by these
2733 statements are most likely to be submissions.
2735 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2737 (1) Generate an error if the third argument for the ${prvs expansion is
2740 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2743 (3) Reset the variables that are obtained from the first argument of
2744 ${prvscheck and used in the second argument before leaving the code,
2745 because their memory is reclaimed, so using them afterwards may do
2748 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2749 one (it's much easier than Tom thought :-).
2751 (5) Because of (4), we can now allow for the use of $prvscheck_result
2752 inside the third argument.
2754 PH/03 For some reason, the default setting of PATH when running a command from
2755 a pipe transport was just "/usr/bin". I have changed it to
2758 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2759 anything to be listed in the output from -bV.
2761 PH/05 When a filter generated an autoreply, the entire To: header line was
2762 quoted in the delivery log line, like this:
2764 => >A.N.Other <ano@some.domain> <original@ddress> ...
2766 This has been changed so that it extracts the operative address. There
2767 may be more than one such address. If so, they are comma-separated, like
2770 => >ano@some.domain,ona@other.domain <original@ddress> ...
2772 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2773 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2774 not being looked up in the rDNS to get a host name, Exim was showing the
2775 IP address twice in Received: lines, even though the IP addresses were
2776 identical. For example:
2778 Received: from [1.2.3.4] (helo=[1.2.3.4])
2780 However, if the real host name was known, it was omitting the HELO data
2781 if it matched the actual IP address. This has been tidied up so that it
2782 doesn't show the same IP address twice.
2784 PH/07 When both +timestamp and +memory debugging was on, the value given by
2785 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2786 called by the debug printing, thereby overwriting the timestamp buffer.
2787 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2789 PH/08 When the original message was included in an autoreply transport, it
2790 always said "this is a copy of the message, including all the headers",
2791 even if body_only or headers_only was set. It now gives an appropriate
2794 PH/09 Applied a patch from the Sieve maintainer which:
2796 o fixes some comments
2797 o adds the (disabled) notify extension core
2798 o adds some debug output for the result of if/elsif tests
2799 o points to the current vacation draft in the documentation
2800 and documents the missing references header update
2804 o fixes a bug in processing the envelope test (when testing
2805 multiple envelope elements, the last element determinted the
2808 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2809 Electronic Mail") by including:
2811 Auto-submitted: auto-generated
2813 in the messages that it generates (bounce messages and others, such as
2814 warnings). In the case of bounce messages for non-SMTP mesages, there was
2815 also a typo: it was using "Auto_submitted" (underscore instead of
2816 hyphen). Since every message generated by Exim is necessarily in response
2817 to another message, thes have all been changed to:
2819 Auto-Submitted: auto-replied
2821 in accordance with these statements in the RFC:
2823 The auto-replied keyword:
2825 - SHOULD be used on messages sent in direct response to another
2826 message by an automatic process,
2828 - MUST NOT be used on manually-generated messages,
2830 - MAY be used on Delivery Status Notifications (DSNs) and Message
2831 Disposition Notifications (MDNs),
2833 - MUST NOT be used on messages generated by automatic or periodic
2834 processes, except for messages which are automatic responses to
2837 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2838 to the default Received: header definition.
2840 PH/12 Added log selector acl_warn_skipped (default on).
2842 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2843 variables because (a) they are in the wrong storage pool and (b) even if
2844 they were copied, it wouldn't work properly because of the caching.
2846 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2847 checking when decoding. Apparently there are clients that generate
2848 overlong encoded strings. Why am I not surprised?
2850 PH/15 If the first argument of "${if match_address" was not empty, but did not
2851 contain an "@" character, Exim crashed. Now it writes a panic log message
2852 and treats the condition as false.
2854 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2856 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2857 extension "envelope-auth". The code is finished and in agreement with
2858 other implementations, but there is no documentation so far and in fact,
2859 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2860 not changing the active code.
2862 Print executed "if" and "elsif" statements when debugging is used. This
2863 helps a great deal to understand what a filter does.
2865 Document more things not specified clearly in RFC3028. I had all this
2866 sorted out, when out of a sudden new issues came to my mind. Oops."
2868 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2871 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2872 canonical form (as documented). However, after a host name lookup from
2873 the IP address, check_host() was doing a simple string comparison with
2874 addresses acquired from the DNS when checking that the found name did
2875 have the original IP as one of its addresses. Since any found IPv6
2876 addresses are likely to be in abbreviated form, the comparison could
2877 fail. Luckily, there already exists a function for doing the comparison
2878 by converting both addresses to binary, so now that is used instead of
2879 the text comparison.
2881 PH/20 There was another similar case to PH/19, when a complete host name was
2882 given in a host list; looking up its IP address could give an abbreviated
2883 form, whereas the current host's name might or might not be abbreviated.
2884 The same fix has been applied.
2890 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2891 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2894 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2895 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2897 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2899 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2900 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2901 Also, for "Auto-Submitted", treat anything other than "no" as
2902 non-personal, in accordance with RFC 3834. (Previously it treated
2903 anything starting "auto-" as non-personal.)
2905 TF/01 The control=submission/name=... option had a problem with syntax
2906 errors if the name included a slash character. The /name= option
2907 now slurps the rest of the string, so it can include any characters
2908 but it must come last in the list of options (after /sender_retain
2911 PH/05 Some modifications to the interface to the fake nameserver for the new
2919 TK/01 Added the "success_on_redirect" address verification option. See
2920 NewStuff for rationale and an example.
2922 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2924 PH/02 Patch to exigrep to allow it to work on syslog lines.
2926 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2927 fread() to skip over the body file's header line, because in Cygwin the
2928 header line is locked and is inaccessible.
2930 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2931 co-exist for some time) to make it clear that it is the Exim ID that is
2932 referenced, not the Message-ID: header line.
2934 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2935 string_format() function, because snprintf() does not exist on all
2938 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2939 ${stat: expansion item.
2941 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2942 protocol synchronization error", to keep the pedants happy.
2944 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2945 well as for IRIX systems, when gcc is being used. See the host.c source
2948 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2950 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2953 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2954 transport and to the smtp transport in LMTP mode.
2956 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2958 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2959 run a filter in a subprocess. This could lead to confusion in subsequent
2960 lookups in the parent process. There should also be a search_tidyup() at
2961 the end of the subprocess.
2963 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2964 only if the host matched helo_try_verify_hosts, which caused the
2965 verification to occur when the EHLO/HELO command was issued. The ACL just
2966 tested the remembered result. Now, if a previous verification attempt has
2967 not happened, "verify = helo" does it there and then.
2969 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2971 TK/03 Fix log output including CR from clamd.
2973 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2974 exist provoked a memory error which could cause a segfault.
2976 PH/15 Installed PCRE 6.2
2978 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2980 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2981 of the problem. Specifically, suggested +O2 rather than +O1 for the
2984 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2986 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2987 record, the port was not correctly logged when the outgoing_port log
2988 selector was set (it logged the transort's default port).
2990 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2991 fallback_hosts, and "hosts" in the smtp transport.
2993 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2994 host errors such as "Connection refused".
2996 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2997 authentication with radiusclient 0.4.9:
2999 - Error returned from rc_read_config was caught wrongly
3000 - Username/password not passed on to radius server due to wrong length.
3002 The presumption is that some radiusclient API changes for 4.51/PH/17
3003 were not taken care of correctly. The code is still untested by me (my
3004 Linux distribution still has 0.3.2 of radiusclient), but it was
3005 contributed by a Radius user.
3007 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3008 expanding the "port" option of the smtp transport.
3010 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3011 while reading a MIME header. Thanks to Tom Hughes for a patch.
3013 PH/24 Include config.h inside local_scan.h so that configuration settings are
3016 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3017 that in an ACL for RCPT (for example), you can examine exactly what was
3020 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3021 commands, but it was not correctly comparing the address with the actual
3022 client host address. Thus, it would show the EHLO address in Received:
3023 header lines when this was not necessary.
3025 PH/27 Added the % operator to ${eval:}.
3027 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3028 it should be ignoring failures (because with -C, for example, it has lost
3029 privilege). It wasn't ignoring creation failures other than "already
3032 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3035 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3036 because an input line was too long, either on its own, or by virtue of
3037 too many continuations, the temporary file was not being removed, and the
3038 return code was incorrect.
3040 PH/31 Missing "BOOL" in function definition in filtertest.c.
3042 PH/32 Applied Sieve patches from the maintainer.
3044 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3046 PH/33 Added "verify = not_blind".
3048 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3049 Local/Makefile (with some defaults set). These are used in built scripts
3050 such as exicyclog, but they have never been used in the exim_install
3051 script (though there are many overriding facilities there). I have
3052 arranged that the exim_install script now takes note of these two
3055 PH/35 Installed configuration files for Dragonfly.
3057 PH/36 When a locally submitted message by a trusted user did not contain a
3058 From: header, and the sender address was obtained from -f or from an SMTP
3059 MAIL command, and the trusted user did not use -F to supply a sender
3060 name, $originator_name was incorrectly used when constructing a From:
3061 header. Furthermore, $originator_name was used for submission mode
3062 messages from external hosts without From: headers in a similar way,
3063 which is clearly wrong.
3065 PH/37 Added control=suppress_local_fixups.
3067 PH/38 When log_selector = +received_sender was set, and the addition of the
3068 sender made the log line's construction buffer exactly full, or one byte
3069 less than full, an overflow happened when the terminating "\n" was
3072 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3073 when the result of a list match is failure because a DNS lookup failed.
3075 PH/40 RM_COMMAND is now used in the building process.
3077 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3078 the "build-* directories that it finds.
3080 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3081 address in a domain literal was a prefix of an interface address.
3083 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3084 when verifying a sender address, unless rewrite_headers is false.
3086 PH/44 Wrote a long comment about why errors_to addresses are verified as
3087 recipients, not senders.
3089 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3090 the ratelimit ACL was added.
3092 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3094 PH/46 Added extra information about PostgreSQL errors to the error string.
3096 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3097 avoiding the need to install special zones in a real server. This is
3098 backwards compatible; if it can't find the fake resolver, it drops back.
3099 Thus, both old and new test suites can be run.
3101 TF/02 Added util/ratelimit.pl
3103 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3104 clock is set back in time.
3106 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3107 Candler <B.Candler@pobox.com>.
3109 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3110 OK for addresses that are the result of redirections.
3112 PH/48 A number of further additions for the benefit of the new test suite,
3113 including a fake gethostbyname() that interfaces to the fake DNS resolver
3116 TF/06 The fix for widen_domains has also been applied to qualify_single and
3117 search_parents which are the other dnslookup options that can cause
3120 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3123 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3125 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3126 service" if it encountered an MX record with an empty target hostname.
3127 The message is now "an MX or SRV record indicated no SMTP service".
3129 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3130 if the DNS of the sending site is misconfigured. This is quite a
3131 common situation. This change restores the behaviour of treating a
3132 helo verification defer as a failure.
3134 PH/51 If self=fail was set on a router, the bounce message did not include the
3135 actual error message.
3141 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3143 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3144 command itself ended in error, the underlying message about the transport
3145 filter timeout was being overwritten with the pipe command error. Now the
3146 underlying error message should be appended to the second error message.
3148 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3150 PH/02 Reduce the amount of output that "make" produces by default. Full output
3151 can still be requested.
3153 PH/03 The warning log line about a condition test deferring for a "warn" verb
3154 was being output only once per connection, rather than after each
3155 occurrence (because it was using the same function as for successful
3156 "warn" verbs). This seems wrong, so I have changed it.
3158 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3159 it should not have, which might have caused a crash in the right
3160 circumstances, but probably never did.
3162 PH/04 Installed a modified version of Tony Finch's patch to make submission
3163 mode fix the return path as well as the Sender: header line, and to
3164 add a /name= option so that you can make the user's friendly name appear
3167 TF/03 Added the control = fakedefer ACL modifier.
3169 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3170 Mark Lowes for thorough testing.
3172 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3174 TK/03 Merged latest SRS patch from Miles Wilton.
3176 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3177 with the definition in sysexits.h (which is #included earlier).
3178 Fortunately, Exim does not actually use EX_OK. The code used to try to
3179 preserve the sysexits.h value, by assumimg that macro definitions were
3180 scanned for macro replacements. I have been disabused of this notion,
3181 so now the code just undefines EX_OK before #including unistd.h.
3183 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3184 in the smtp transport. When a block could not be written in a single
3185 write() function, the timeout was being re-applied to each part-write.
3186 This seems wrong - if the receiver was accepting one byte at a time it
3187 would take for ever. The timeout is now adjusted when this happens. It
3188 doesn't have to be particularly precise.
3190 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3191 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3193 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3194 without a domain if the check to <postmaster@domain> fails.
3196 SC/01 Eximstats: added -xls and the ability to specify output files
3197 (patch written by Frank Heydlauf).
3199 SC/02 Eximstats: use FileHandles for outputing results.
3201 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3203 SC/04 Eximstats: fixed display of large numbers with -nvr option
3205 SC/05 Eximstats: fixed merging of reports with empty tables.
3207 SC/06 Eximstats: added the -include_original_destination flag
3209 SC/07 Eximstats: removed tabs and trailing whitespace.
3211 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3213 TK/06 MBOX spool code: Add real "From " MBOX separator line
3214 so the .eml file is really in mbox format (even though
3215 most programs do not really care). Patch from Alex Miller.
3217 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3218 The latter is generated from $received_to and is only set if the
3219 message has one envelope recipient. SA can use these headers,
3220 obviously out-of-the-box. Patch from Alex Miller.
3222 PH/08 The ${def test on a variable was returning false if the variable's
3223 value was "0", contrary to what the specification has always said!
3224 The result should be true unless the variable is empty.
3226 PH/09 The syntax error of a character other than { following "${if
3227 def:variable_name" (after optional whitespace) was not being diagnosed.
3228 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3229 accidental colon was present, for example, could give incorrect results.
3231 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3232 result is used (not including appendfile, where other changes are about
3235 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3236 This involved changing a lot of size variables from int to off_t. It
3237 should work with maildirs and everything.
3239 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3240 spamd dying while we are connected to it.
3242 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3245 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3246 function for BDB 4.3.
3248 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3250 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3251 the sources for PCRE have been re-arranged and more files are now
3254 PH/15 The code I had for printing potentially long long variables in PH/11
3255 above was not the best (it lost precision). The length of off_t variables
3256 is now inspected at build time, and an appropriate printing format (%ld
3257 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3258 to be "long long int" or "long int". This is needed for the internal
3259 formatting function string_vformat().
3261 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3262 the configuration file to be ":syslog", then the script "guesses" where
3263 the logs files are, rather than using the compiled in default. In our
3264 case the guess is not the same as the compiled default, so the script
3265 suddenly stopped working when I started to use syslog. The patch checks
3266 to see if log_file_path is "". If so, it attempts to read it from exim
3267 with no configuration file to get the compiled in version, before it
3268 falls back to the previous guessing code."
3270 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3271 implementing BATV in an Exim configuration. See NewStuff for the gory
3274 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3275 Makefile that are specific to HP-UX.
3277 PH/18 If the "use_postmaster" option was set for a recipient callout together
3278 with the "random" option, the postmaster address was used as the MAIL
3279 FROM address for the random test, but not for the subsequent recipient
3280 test. It is now used for both.
3282 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3283 patch removes a few documentation additions to RFC 3028, because the
3284 latest draft now contains them. It adds the new en;ascii-case comparator
3285 and a new error check for 8bit text in MIME parts. Comparator and
3286 require names are now matched exactly. I enabled the subaddress
3287 extension, but it is not well tested yet (read: it works for me)."
3289 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3290 rework some of the code of TK/09 above to avoid the hardwired use of
3291 "%lld" and "long long". Replaced the call to snprintf() with a call to
3294 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3295 records point to non-existent hosts", "retry timeout exceeded", and
3296 "retry time not reached for any host after a long failure period".
3298 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3299 experimental DomainKeys support:
3301 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3302 (2) On an error such as an illegally used "control", the wrong name for
3303 the control was given.
3305 These problems did NOT occur unless DomainKeys support was compiled.
3307 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3309 PH/24 Added ${if match_ip condition.
3311 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3312 databases so that it will be absolutely obvious if a crash occurs in the
3313 DB library. This is a regular occurrence (often caused by mis-matched
3316 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3317 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3318 picked up on a user's system that detects such things. There doesn't seem
3319 to be a gcc warning option for this - only an attribute that has to be
3320 put on the function's prototype. It seems that in Fedora Core 4 they have
3321 set this on a number of new functions. No doubt there will be more in due
3324 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3325 specify a transport. However, if an address that was verified by such a
3326 router was the subject of a callout, Exim crashed because it tried to
3327 read the rcpt_include_affixes from the non-existent transport. Now it
3328 just assumes that the setting of that option is false. This bug was
3329 introduced by 4.51/PH/31.
3331 PH/28 Changed -d+all to exclude +memory, because that information is very
3332 rarely of interest, but it makes the output a lot bigger. People tend to
3333 do -d+all out of habit.
3335 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3336 code in os-type was giving problems when libc.so lives in lib64, like on
3339 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3340 aren't the modern standard, and it seems that some systems' include files
3341 don't always have them. Exim was already checking for some of the newer
3342 ones like T_AAAA, and defining it itself. I've added checks for all the
3343 record types that Exim uses.
3345 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3346 not automatically generating a new one, as it is supposed to. This
3347 prevented TLS from working. If the file did exist, but contained invalid
3348 data, a new version was generated, as expected. It was only the case of a
3349 non-existent file that was broken.
3351 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3352 with a change in libdomainkeys > 0.64.
3354 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3355 from DNS. If the selector record carries the flag, it now has
3356 precedence over the domain-wide flag.
3358 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3360 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3361 the use of a port name that isn't defined in /etc/services) occurred, the
3362 message was deferred as in a normal delivery, and thus remained on the
3363 spool, instead of being failed because of the mua_wrapper setting. This
3364 is now fixed, and I tidied up some of the mua_wrapper messages at the
3367 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3368 the messages in a hash of arrays rather than using individual hashes.
3369 This is a bit cleaner and results in dramatic memory savings, albeit
3370 at a slight CPU cost.
3372 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3373 as requested by Marc Sherman.
3375 SC/10 Eximstats: added histograms for user specified patterns as requested
3378 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3380 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3381 fopen() in the content-scanning modules that did not already have it.
3387 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3388 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3390 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3392 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3394 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3395 to test Sieve filters that use "vacation".
3397 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3398 that changes the way the GnuTLS parameters are stored in the cache file.
3399 The new format can be generated externally. For backward compatibility,
3400 if the data in the cache doesn't make sense, Exim assumes it has read an
3401 old-format file, and it generates new data and writes a new file. This
3402 means that you can't go back to an older release without removing the
3405 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3406 work if there are any delivery delays because "one_time" forces the
3407 parent to be marked "delivered", so its unseen clone is never tried
3408 again. For this reason, Exim now forbids the simultaneous setting of
3411 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3412 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3413 entry if you want to know the details. The fix, however, neglected to
3414 consider the case when local delivery batching is involved. The test for
3415 "previously delivered" was not happening when checking to see if an
3416 address could be batched with a previous (undelivered) one; under
3417 certain circumstances this could lead to multiple deliveries to the same
3420 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3421 in its include files, and this causes problems building Exim.
3423 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3424 header_syntax) but Exim was just ignoring anything given after a slash.
3425 In particular, this caused confusion with an attempt to use "verify =
3426 reverse_host_lookup/defer_ok". An error is now given when options are
3427 supplied for verify items that do not have them. (Maybe reverse_host_
3428 lookup should have a defer_ok option, but that's a different point.)
3430 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3431 defined by RFC 821) to 2048, because there were problems with some AUTH
3432 commands, and RFC 1869 says the size should be increased for extended
3433 SMTP commands that take arguments.
3435 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3438 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3439 "unknown" error; now it says that the functionality isn't in the binary.
3441 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3442 an address' error message when a string expansion fails (syntax or
3443 whatever). Otherwise the password may appear in the log. Following change
3444 PH/42 below, there is no longer a chance of it appearing in a bounce
3447 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3449 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3450 of its addresses was ever tried. (Bugzilla bug #2.)
3452 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3453 the result incorrectly in the debug output. (It correctly added a newline
3454 to what was transported.)
3456 TF/01 Added $received_time.
3458 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3459 commented out examples of how to interface to a virus scanner and to
3460 SpamAssassin. Also added commented examples of av_scanner and
3461 spamd_address settings.
3463 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3464 and controls are allowed in which ACLs. There were a couple of minor
3465 errors. Some of the entries in the conditions table (which is a table of
3466 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3467 negation of where the condition IS allowed.
3469 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3471 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3472 header file does not have a version number, so I've had to invent a new
3473 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3474 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3475 radiusclient), but it was contributed by a Radius user.
3477 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3478 files or directories, for OpenSSL.
3480 PH/19 When an Exim process that is running as root has to create an Exim log
3481 file, it does so in a subprocess that runs as exim:exim so as to get the
3482 ownership right at creation (otherwise, other Exim processes might see
3483 the file with the wrong ownership). There was no test for failure of this
3484 fork() call, which would lead to the process getting stuck as it waited
3485 for a non-existent subprocess. Forks do occasionally fail when resources
3486 run out. I reviewed all the other calls to fork(); they all seem to check
3489 PH/20 When checking for unexpected SMTP input at connect time (before writing
3490 the banner), Exim was not dealing correctly with a non-positive return
3491 from the read() function. If the client had disconnected by this time,
3492 the result was a log entry for a synchronization error with an empty
3493 string after "input=" when read() returned zero. If read() returned -1
3494 (an event I could not check), uninitialized data bytes were printed.
3495 There were reports of junk text (parts of files, etc) appearing after
3498 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3500 PH/22 Added support for macro redefinition, and (re)definition in between
3501 driver and ACL definitions.
3503 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3504 forgetting to use the resulting value; it was using the unexpanded value.
3506 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3507 hadn't been configured. The fix is from Juergen Kreileder, who
3508 understands it better than I do:
3510 "Here's what I see happening with three configured cyrus_sasl
3511 authenticators configured (plain, login, cram-md5):
3513 On startup auth_cyrus_sasl_init() gets called for each of these.
3514 This means three calls to sasl_listmech() without a specified mech_list.
3515 => SASL tests which mechs of all available mechs actually work
3516 => three warnings about OTP not working
3517 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3519 With the patch, sasl_listmech() also gets called three times. But now
3520 SASL's mech_list option is set to the server_mech specified in the the
3521 authenticator. Or in other words, the answer from sasl_listmech()
3522 gets limited to just the mech you're testing for (which is different
3524 => the return list contains just 'plain' or 'login', 'cram-md5' or
3525 nothing depending on the value of ob->server_mech.
3527 I've just tested the patch: Authentication still works fine,
3528 unavailable mechs specified in the exim configuration are still
3529 caught, and the auth.log warnings about OTP are gone."
3531 PH/25 When debugging is enabled, the contents of the command line are added
3532 to the debugging output, even when log_selector=+arguments is not
3535 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3536 answer is "GNU", and only if the return is "GNU/something" is the answer
3539 PH/27 $acl_verify_message is now set immediately after the failure of a
3540 verification in an ACL, and so is available in subsequent modifiers. In
3541 particular, the message can be preserved by coding like this:
3543 warn !verify = sender
3544 set acl_m0 = $acl_verify_message
3546 Previously, $acl_verify_message was set only while expanding "message"
3547 and "log_message" when a very denied access.
3549 PH/28 Modified OS/os.c-Linux with
3551 -#ifndef OS_LOAD_AVERAGE
3552 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3554 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3555 nomenclature these days.)
3557 PH/29 Installed patch from the Sieve maintainer that adds the options
3558 sieve_useraddress and sieve_subaddress to the redirect router.
3560 PH/30 In these circumstances:
3561 . Two addresses routed to the same list of hosts;
3562 . First host does not offer TLS;
3563 . First host accepts first address;
3564 . First host gives temporary error to second address;
3565 . Second host offers TLS and a TLS session is established;
3566 . Second host accepts second address.
3567 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3568 and peerdn, if requested) that were in fact used only for the second
3571 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3572 attention to any local part prefix or suffix that was matched by the
3573 router that accepted the address. It now behaves in the same way as it
3574 does for delivery: the affixes are removed from the local part unless
3575 rcpt_include_affixes is set on the transport.
3577 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3578 timeout during the DATA phase of an incoming message.
3580 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3581 applied a patch sanctioned by the Sieve maintainer.
3583 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3584 the uid or gid is negative. A case of a negative gid caused this to be
3585 noticed. The fix allows for either to be negative.
3587 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3588 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3591 PH/36 The change PH/12 above was broken. Fixed it.
3593 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3594 the grounds that routing the same address twice would always produce the
3595 same answer. This might have been true once, but it is certainly no
3596 longer true now. Routing a child address may depend on the previous
3597 routing that produced that child. Some complicated redirection strategies
3598 went wrong when messages had multiple recipients, and made Exim's
3599 behaviour dependent on the order in which the addresses were given.
3601 I have moved the duplicate checking until after the routing is complete.
3602 Exim scans the addresses that are assigned to local and remote
3603 transports, and removes any duplicates. This means that more work will be
3604 done, as duplicates will always all be routed, but duplicates are
3605 presumably rare, so I don't expect this is of any significance.
3607 For deliveries to pipes, files, and autoreplies, the duplicate checking
3608 still happens during the routing process, since they are not going to be
3611 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3612 It corrects a timeout issue with spamd. This is Ian's comment: "The
3613 background is that sometimes spamd either never reads data from a
3614 connection it has accepted, or it never writes response data. The exiscan
3615 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3616 blindly assumes that writes won't block so it may never time out."
3618 PH/39 Allow G after quota size as well as K and M.
3620 PH/40 The value set for $authenticated_id in an authenticator may not contain
3621 binary zeroes or newlines because the value is written to log lines and
3622 to spool files. There was no check on this. Now the value is run through
3623 the string_printing() function so that such characters are converted to
3624 printable escape sequences.
3626 PH/41 $message_linecount is a new variable that contains the total number of
3627 lines in the message. Compare $body_linecount, which is the count for the
3630 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3631 bounce and delay warning messages, except in certain special cases, which
3634 (a) An SMTP error message from a remote host;
3635 (b) A message specified in a :fail: redirection;
3636 (c) A message specified in a "fail" command in a system filter;
3637 (d) A message specified in a FAIL return from the queryprogram router;
3638 (e) A message specified by the cannot_route_message router option.
3640 In these cases only, Exim does include the error details in bounce and
3641 warning messages. There are also a few cases where bland messages such
3642 as "unrouteable address" or "local delivery error" are given.
3644 PH/43 $value is now also set for the "else" part of a ${run expansion.
3646 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3647 being worked on, but at least Exim now implements the latest version to
3650 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3651 process to complete was treated as a delivery failure, a timeout while
3652 writing the message to the pipe was logged, but erroneously treated as a
3653 successful delivery. Such timeouts include transport filter timeouts. For
3654 consistency with the overall process timeout, these timeouts are now
3655 treated as errors, giving rise to delivery failures by default. However,
3656 there is now a new Boolean option for the pipe transport called
3657 timeout_defer, which, if set TRUE, converts the failures into defers for
3658 both kinds of timeout. A transport filter timeout is now identified in
3661 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3662 systems where "make" and "gmake" are different, calling "gmake" at top
3663 level broke things. I've arranged for the value of $(MAKE) to be passed
3664 from the Makefile to this script so that it can call the same version of
3668 A note about Exim versions 4.44 and 4.50
3669 ----------------------------------------
3671 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3672 changes of various kinds. As a consequence, a big documentation update was
3673 needed. This delayed the release for rather longer than seemed good, especially
3674 in the light of a couple of (minor) security issues. Therefore, the changes
3675 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3676 release. So 4.44 and 4.50 are in effect two different branches that both start
3679 I have left the 4.50 change log unchanged; it contains all the changes since
3680 4.43. The change log for 4.44 is below; many of its items are identical to
3681 those for 4.50. This seems to be the most sensible way to preserve the
3682 historical information.
3688 1. Minor wording change to the doc/README.SIEVE file.
3690 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3691 computation of the current number of files was incorrect.
3693 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3694 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3697 4. Give more explanation in the error message when the command for a transport
3698 filter fails to execute.
3700 5. There are several places where Exim runs a non-Exim command in a
3701 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3702 was being done only for the command run by the queryprogram router. It is
3703 now done for all such subprocesses. The other cases are: ${run, transport
3704 filters, and the commands run by the lmtp and pipe transports.
3706 6. Added CONFIGURE_GROUP build-time option.
3708 7. Some older OS have a limit of 256 on the maximum number of file
3709 descriptors. Exim was using setrlimit() to set 1000 as a large value
3710 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3711 systems. I've change it so that if it can't get 1000, it tries for 256.
3713 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3714 was an oversight, and furthermore, ever since the addition of extra
3715 controls (e.g. 4.43/32), the checks on when to allow different forms of
3716 "control" were broken. There should now be diagnostics for all cases when a
3717 control that does not make sense is encountered.
3719 9. Added the /retain_sender option to "control=submission".
3721 10. $recipients is now available in the predata ACL (oversight).
3723 11. Tidy the search cache before the fork to do a delivery from a message
3724 received from the command line. Otherwise the child will trigger a lookup
3725 failure and thereby defer the delivery if it tries to use (for example) a
3726 cached ldap connection that the parent has called unbind on.
3728 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3729 of $address_data from the recipient verification was clobbered by the
3730 sender verification.
3732 13. The value of address_data from a sender verification is now available in
3733 $sender_address_data in subsequent conditions in the ACL statement.
3735 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3737 15. Added a new option "connect=<time>" to callout options, to set a different
3740 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3741 was its contents. (It was OK if the option was not defined at all.)
3743 17. A "Completed" log line is now written for messages that are removed from
3744 the spool by the -Mrm option.
3746 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3747 information about exactly what failed.
3749 19. Added -dd to debug only the daemon process.
3751 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3752 handles timeouts, both on the server side and network timeouts. Renamed the
3753 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3755 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3756 It is now set to "smtps".
3758 22. $host_address is now set to the target address during the checking of
3759 ignore_target_hosts.
3761 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3762 being passed; this would have caused $sender_host_name to have been used if
3763 matching the list had actually called for a host name (not very likely,
3764 since this list is usually IP addresses). A host name is now passed as
3767 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3768 code that specifies a non-restarting handler (typically sigaction() in
3769 modern systems) in an attempt to fix a rare and obscure crash bug.
3771 25. Narrowed the window for a race in the daemon that could cause it to ignore
3772 SIGCHLD signals. This is not a major problem, because they are used only to
3773 wake it up if nothing else does.
3775 26. A malformed maildirsize file could cause Exim to calculate negative values
3776 for the mailbox size or file count. Odd effects could occur as a result.
3777 The maildirsize information is now recalculated if the size or filecount
3780 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3781 support for a long time. Removed HAVE_SYS_VFS_H.
3783 28. Installed the latest version of exipick from John Jetmore.
3785 29. In an address list, if the pattern was not a regular expression, an empty
3786 subject address (from a bounce message) matched only if the pattern was an
3787 empty string. Non-empty patterns were not even tested. This was the wrong
3788 because it is perfectly reasonable to use an empty address as part of a
3789 database query. An empty address is now tested by patterns that are
3790 lookups. However, all the other forms of pattern expect the subject to
3791 contain a local part and a domain, and therefore, for them, an empty
3792 address still always fails if the pattern is not itself empty.
3794 30. Exim went into a mad DNS loop when attempting to do a callout where the
3795 host was specified on an smtp transport, and looking it up yielded more
3796 than one IP address.
3798 31. Re-factored the code for checking spool and log partition space into a
3799 function that finds that data and another that does the check. The former
3800 is then used to implement four new variables: $spool_space, $log_space,
3801 $spool_inodes, and $log_inodes.
3803 32. The RFC2047 encoding function was originally intended for short strings
3804 such as real names; it was not keeping to the 75-character limit for
3805 encoded words that the RFC imposes. It now respects the limit, and
3806 generates multiple encoded words if necessary. To be on the safe side, I
3807 have increased the buffer size for the ${rfc2047: expansion operator from
3810 33. It is now permitted to omit both strings after an "if" condition; if the
3811 condition is true, the result is "true". As before, when the second string
3812 is omitted, a false condition yields an empty string. This makes it less
3813 cumbersome to write custom ACL and router conditions.
3815 34. Failure to deliver a bounce message always caused it to be frozen, even if
3816 there was an errors_to setting on the router. The errors_to setting is now
3819 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3820 canonical form (fully expanded) before being placed in
3821 $sender_host_address.
3823 36. The table in the code that translates DNS record types into text (T_A to
3824 "A" for instance) was missing entries for NS and CNAME. It is just possible
3825 that this could have caused confusion if both these types were looked up
3826 for the same domain, because the text type is used as part of Exim's
3827 per-process caching. But the chance of anyone hitting this buglet seems
3830 37. The dnsdb lookup has been extended in a number of ways.
3832 (1) There is a new type, "zns", which walks up the domain tree until it
3833 finds some nameserver records. It should be used with care.
3835 (2) There is a new type, "mxh", which is like "mx" except that it returns
3836 just the host names, not the priorities.
3838 (3) It is now possible to give a list of domains (or IP addresses) to be
3839 looked up. The behaviour when one of the lookups defers can be
3840 controlled by a keyword.
3842 (4) It is now possible to specify the separator character for use when
3843 multiple records are returned.
3845 38. The dnslists ACL condition has been extended: it is now possible to supply
3846 a list of IP addresses and/or domains to be looked up in a particular DNS
3849 39. Added log_selector=+queue_time_overall.
3851 40. When running the queue in the test harness, wait just a tad after forking a
3852 delivery process, to get repeatability of debugging output.
3854 41. Include certificate and key file names in error message when GnuTLS fails
3855 to set them up, because the GnuTLS error message doesn't include the name
3856 of the failing file when there is a problem reading it.
3858 42. Allow both -bf and -bF in the same test run.
3860 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3862 44. The "Exiscan patch" is now merged into the mainline Exim source.
3864 45. Sometimes the final signoff response after QUIT could fail to get
3865 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3866 < 0 before doing a fflush(). This bug looks as though it goes back to the
3867 introduction of TLS in release 3.20, but "sometimes" must have been rare
3868 because the tests only now provoked it.
3870 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3871 (this can affect the format of dates).
3873 47. exim_tidydb, when checking for the continued existence of a message for
3874 which it has found a message-specific retry record, was not finding
3875 messages that were in split spool directories. Consequently, it was
3876 deleting retry records that should have stayed in existence.
3878 48. Steve fixed some bugs in eximstats.
3880 49. The SPA authentication driver was not abandoning authentication and moving
3881 on to the next authenticator when an expansion was forced to fail,
3882 contradicting the general specification for all authenticators. Instead it
3883 was generating a temporary error. It now behaves as specified.
3885 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3886 (the order specifies the preference for clients). The order is now AES256,
3887 AES128, 3DES, ARCFOUR128.
3889 51. Small patch to Sieve code - explicitly set From: when generating an
3892 52. Exim crashed if a remote delivery caused a very long error message to be
3893 recorded - for instance if somebody sent an entire SpamAssassin report back
3894 as a large number of 550 error lines. This bug was coincidentally fixed by
3895 increasing the size of one of Exim's internal buffers (big_buffer) that
3896 happened as part of the Exiscan merge. However, to be on the safe side, I
3897 have made the code more robust (and fixed the comments that describe what
3900 53. Now that there can be additional text after "Completed" in log lines (if
3901 the queue_time_overall log selector is set), a one-byte patch to exigrep
3902 was needed to allow it to recognize "Completed" as not the last thing in
3905 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3906 patch that reportedly fixes this has been added. I am not expert enough to
3907 create a test for it. This is what the patch creator wrote:
3909 "I found a little strange behaviour of ldap code when working with
3910 Windows 2003 AD Domain, where users was placed in more than one
3911 Organization Units. When I tried to give exim partial DN, the exit code
3912 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3913 But simultaneously result of request was absolutely normal ldap result,
3914 so I produce this patch..."
3916 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3917 REFERENCE, so I have modified the code to exclude the patch when that macro
3920 55. Some experimental protocols are using DNS PTR records for new purposes. The
3921 keys for these records are domain names, not reversed IP addresses. The
3922 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3923 leaves it alone. Component reversal etc. now happens only for IP addresses.
3926 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3928 57. Double the size of the debug message buffer (to 2048) so that more of very
3929 long debug lines gets shown.
3931 58. The exicyclog utility now does better if the number of log files to keep
3932 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3934 59. Two changes related to the smtp_active_hostname option:
3936 (1) $smtp_active_hostname is now available as a variable.
3937 (2) The default for smtp_banner uses $smtp_active_hostname instead
3938 of $primary_hostname.
3940 60. The host_aton() function is supposed to be passed a string that is known
3941 to be a valid IP address. However, in the case of IPv6 addresses, it was
3942 not checking this. This is a hostage to fortune. Exim now panics and dies
3943 if the condition is not met. A case was found where this could be provoked
3944 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3945 components; fortuitously, this particular loophole had already been fixed
3946 by change 4.50/55 above.
3948 If there are any other similar loopholes, the new check in host_aton()
3949 itself should stop them being exploited. The report I received stated that
3950 data on the command line could provoke the exploit when Exim was running as
3951 exim, but did not say which command line option was involved. All I could
3952 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3953 running as the user.
3956 61. There was a buffer overflow vulnerability in the SPA authentication code
3957 (which came originally from the Samba project). I have added a test to the
3958 spa_base64_to_bits() function which I hope fixes it.
3961 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3962 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3964 63. The daemon start-up calls getloadavg() while still root for those OS that
3965 need the first call to be done as root, but it missed one case: when
3966 deliver_queue_load_max is set with deliver_drop_privilege. This is
3967 necessary for the benefit of the queue runner, because there is no re-exec
3968 when deliver_drop_privilege is set.
3970 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3971 This has been fixed.
3973 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3974 was in use, was not putting the data itself into the right store pool;
3975 consequently, it could be overwritten for a subsequent message in the same
3976 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3979 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3981 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3982 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3983 the function were treating the return as a boolean value, which happened to
3984 work because 0=false and not-0=true, but is not correct code.
3986 68. The host_aton() function was not handling scoped IPv6 addresses (those
3987 with, for example, "%eth0" on the end) correctly.
3989 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3990 negated items (that is, ~something) in unsigned ints. Some compilers
3991 apparently mutter when there is no cast.
3993 70. If an address verification called from an ACL failed, and did not produce a
3994 user-specific message (i.e. there was only a "system" message), nothing was
3995 put in $acl_verify_message. In this situation, it now puts the system
3998 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3999 session; change 4.31/43 added the unwanted input to the log line - except
4000 that it did not do this in the start of session case. It now does.
4002 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4003 This is wrong and can cause the other end to generate a synchronization
4004 error if it is another Exim or anything else that does the synchronization
4005 check. A QUIT command is no longer sent after a timeout.
4007 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4008 during host lookups.
4010 74. The defer_ok option of callout verification was not working if it was used
4011 when verifying addresses in header lines, that is, for this case:
4013 verify = header_sender/callout=defer_ok
4015 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4016 those file descriptors could be used for SMTP connections. If anything
4017 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4018 could be sent to the SMTP client, causing chaos. The daemon now opens
4019 stdin, stdout, and stderr to /dev/null when it puts itself into the
4022 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4023 log by default. The user can override this with suitable Perl magic.
4025 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4026 the log message when discard triggers, was not working for the DATA ACL or
4027 for the non-SMTP ACL.
4029 78. Error message wording change in sieve.c.
4031 79. If smtp_accept_max_per_host was set, the number of connections could be
4032 restricted to fewer than expected, because the daemon was trying to set up
4033 a new connection before checking whether the processes handling previous
4034 connections had finished. The check for completed processes is now done
4035 earlier. On busy systems, this bug wouldn't be noticed because something
4036 else would have woken the daemon, and it would have reaped the completed
4039 80. If a message was submitted locally by a user whose login name contained one
4040 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4041 It caused a spool format error. I have fixed the spool reading code. A
4042 related problem was that the "from" clause in the Received: line became
4043 illegal because of the space(s). It is now covered by ${quote_local_part.
4045 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4048 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4050 83. Patch from Sieve maintainer to fix unterminated string problem in
4051 "vacation" handling.
4053 84. Some minor changes to the Linux configuration files to help with other
4054 OS variants using glibc.
4056 85. One more patch for Sieve to update vacation handling to latest spec.
4059 ----------------------------------------------------
4060 See the note above about the 4.44 and 4.50 releases.
4061 ----------------------------------------------------
4067 1. Change 4.43/35 introduced a bug that caused file counts to be
4068 incorrectly computed when quota_filecount was set in an appendfile
4071 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4072 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4075 3. Give more explanation in the error message when the command for a transport
4076 filter fails to execute.
4078 4. There are several places where Exim runs a non-Exim command in a
4079 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4080 was being done only for the command run by the queryprogram router. It is
4081 now done for all such subprocesses. The other cases are: ${run, transport
4082 filters, and the commands run by the lmtp and pipe transports.
4084 5. Some older OS have a limit of 256 on the maximum number of file
4085 descriptors. Exim was using setrlimit() to set 1000 as a large value
4086 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4087 systems. I've change it so that if it can't get 1000, it tries for 256.
4089 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4090 was an oversight, and furthermore, ever since the addition of extra
4091 controls (e.g. 4.43/32), the checks on when to allow different forms of
4092 "control" were broken. There should now be diagnostics for all cases when a
4093 control that does not make sense is encountered.
4095 7. $recipients is now available in the predata ACL (oversight).
4097 8. Tidy the search cache before the fork to do a delivery from a message
4098 received from the command line. Otherwise the child will trigger a lookup
4099 failure and thereby defer the delivery if it tries to use (for example) a
4100 cached ldap connection that the parent has called unbind on.
4102 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4103 of $address_data from the recipient verification was clobbered by the
4104 sender verification.
4106 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4107 was its contents. (It was OK if the option was not defined at all.)
4109 11. A "Completed" log line is now written for messages that are removed from
4110 the spool by the -Mrm option.
4112 12. $host_address is now set to the target address during the checking of
4113 ignore_target_hosts.
4115 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4116 being passed; this would have caused $sender_host_name to have been used if
4117 matching the list had actually called for a host name (not very likely,
4118 since this list is usually IP addresses). A host name is now passed as
4121 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4122 code that specifies a non-restarting handler (typically sigaction() in
4123 modern systems) in an attempt to fix a rare and obscure crash bug.
4125 15. Narrowed the window for a race in the daemon that could cause it to ignore
4126 SIGCHLD signals. This is not a major problem, because they are used only to
4127 wake it up if nothing else does.
4129 16. A malformed maildirsize file could cause Exim to calculate negative values
4130 for the mailbox size or file count. Odd effects could occur as a result.
4131 The maildirsize information is now recalculated if the size or filecount
4134 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4135 support for a long time. Removed HAVE_SYS_VFS_H.
4137 18. Updated exipick to current release from John Jetmore.
4139 19. Allow an empty sender to be matched against a lookup in an address list.
4140 Previously the only cases considered were a regular expression, or an
4143 20. Exim went into a mad DNS lookup loop when doing a callout where the
4144 host was specified on the transport, if the DNS lookup yielded more than
4147 21. The RFC2047 encoding function was originally intended for short strings
4148 such as real names; it was not keeping to the 75-character limit for
4149 encoded words that the RFC imposes. It now respects the limit, and
4150 generates multiple encoded words if necessary. To be on the safe side, I
4151 have increased the buffer size for the ${rfc2047: expansion operator from
4154 22. Failure to deliver a bounce message always caused it to be frozen, even if
4155 there was an errors_to setting on the router. The errors_to setting is now
4158 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4159 canonical form (fully expanded) before being placed in
4160 $sender_host_address.
4162 24. Updated eximstats to version 1.33
4164 25. Include certificate and key file names in error message when GnuTLS fails
4165 to set them up, because the GnuTLS error message doesn't include the name
4166 of the failing file when there is a problem reading it.
4168 26. Expand error message when OpenSSL has problems setting up cert/key files.
4171 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4172 (this can affect the format of dates).
4174 28. exim_tidydb, when checking for the continued existence of a message for
4175 which it has found a message-specific retry record, was not finding
4176 messages that were in split spool directories. Consequently, it was
4177 deleting retry records that should have stayed in existence.
4179 29. eximstats updated to version 1.35
4180 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4181 1.35 - bugfix such that pie charts by volume are generated correctly
4183 30. The SPA authentication driver was not abandoning authentication and moving
4184 on to the next authenticator when an expansion was forced to fail,
4185 contradicting the general specification for all authenticators. Instead it
4186 was generating a temporary error. It now behaves as specified.
4188 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4189 (the order specifies the preference for clients). The order is now AES256,
4190 AES128, 3DES, ARCFOUR128.
4192 31. Small patch to Sieve code - explicitly set From: when generating an
4195 32. Exim crashed if a remote delivery caused a very long error message to be
4196 recorded - for instance if somebody sent an entire SpamAssassin report back
4197 as a large number of 550 error lines. This bug was coincidentally fixed by
4198 increasing the size of one of Exim's internal buffers (big_buffer) that
4199 happened as part of the Exiscan merge. However, to be on the safe side, I
4200 have made the code more robust (and fixed the comments that describe what
4203 33. Some experimental protocols are using DNS PTR records for new purposes. The
4204 keys for these records are domain names, not reversed IP addresses. The
4205 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4206 leaves it alone. Component reversal etc. now happens only for IP addresses.
4209 34. The host_aton() function is supposed to be passed a string that is known
4210 to be a valid IP address. However, in the case of IPv6 addresses, it was
4211 not checking this. This is a hostage to fortune. Exim now panics and dies
4212 if the condition is not met. A case was found where this could be provoked
4213 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4214 components; fortuitously, this particular loophole had already been fixed
4215 by change 4.50/55 or 4.44/33 above.
4217 If there are any other similar loopholes, the new check in host_aton()
4218 itself should stop them being exploited. The report I received stated that
4219 data on the command line could provoke the exploit when Exim was running as
4220 exim, but did not say which command line option was involved. All I could
4221 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4222 running as the user.
4225 35. There was a buffer overflow vulnerability in the SPA authentication code
4226 (which came originally from the Samba project). I have added a test to the
4227 spa_base64_to_bits() function which I hope fixes it.
4230 36. The daemon start-up calls getloadavg() while still root for those OS that
4231 need the first call to be done as root, but it missed one case: when
4232 deliver_queue_load_max is set with deliver_drop_privilege. This is
4233 necessary for the benefit of the queue runner, because there is no re-exec
4234 when deliver_drop_privilege is set.
4236 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4237 was in use, was not putting the data itself into the right store pool;
4238 consequently, it could be overwritten for a subsequent message in the same
4239 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4242 38. Sometimes the final signoff response after QUIT could fail to get
4243 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4244 < 0 before doing a fflush(). This bug looks as though it goes back to the
4245 introduction of TLS in release 3.20, but "sometimes" must have been rare
4246 because the tests only now provoked it.
4252 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4253 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4254 Now it returns an integer. A number of calls were still expecting a T/F
4255 return. Fortuitously, in all cases, the tests worked in OK situations,
4256 which is the norm. However, things would have gone wrong on any write
4257 failures on the smtp file descriptor. This function is used when sending
4258 messages over SMTP and also when doing verify callouts.
4260 2. When Exim is called to do synchronous delivery of a locally submitted
4261 message (the -odf or -odi options), it no longer closes stderr before doing
4264 3. Implemented the mua_wrapper option.
4266 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4268 5. Implemented the functions header_remove(), header_testname(),
4269 header_add_at_position(), and receive_remove_recipient(), and exported them
4272 6. If an ACL "warn" statement specified the addition of headers, Exim already
4273 inserted X-ACL-Warn: at the start if there was no header name. However, it
4274 was not making this test for the second and subsequent header lines if
4275 there were newlines in the string. This meant that an invalid header could
4276 be inserted if Exim was badly configured.
4278 7. Allow an ACL "warn" statement to add header lines at the start or after all
4279 the Received: headers, as well as at the end.
4281 8. Added the rcpt_4xx retry error code.
4283 9. Added postmaster_mailfrom=xxx to callout verification option.
4285 10. Added mailfrom=xxxx to the callout verification option, for verify=
4288 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4289 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4291 12. Inserted some casts to stop certain compilers warning when using pointer
4292 differences as field lengths or precisions in printf-type calls (mostly
4293 affecting debugging statements).
4295 13. Added optional readline() support for -be (dynamically loaded).
4297 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4298 same clock tick as a message's arrival, so that its received time was the
4299 same as the "first fail" time on the retry record, and that message
4300 remained on the queue past the ultimate address timeout, every queue runner
4301 would try a delivery (because it was past the ultimate address timeout) but
4302 after another failure, the ultimate address timeout, which should have then
4303 bounced the address, did not kick in. This was a "< instead of <=" error;
4304 in most cases the first failure would have been in the next clock tick
4305 after the received time, and all would be well.
4307 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4308 being recognized when the domain list was tested by the match_domain
4309 condition in an expansion string.
4311 16. Added the ${str2b64: operator.
4313 17. Exim was always calling setrlimit() to set a large limit for the number of
4314 processes, without checking whether the existing limit was already
4315 adequate. (It did check for the limit on file descriptors.) Furthermore,
4316 errors from getrlimit() and setrlimit() were being ignored. Now they are
4317 logged to the main and panic logs, but Exim does carry on, to try to do its
4318 job under whatever limits there are.
4320 18. Imported PCRE 5.0.
4322 19. Trivial typo in log message " temporarily refused connection" (the leading
4325 20. If the log selector return_path_on_delivery was set and an address was
4326 redirected to /dev/null, the delivery process crashed because it assumed
4327 that a return path would always be set for a "successful" delivery. In this
4328 case, the whole delivery is bypassed as an optimization, and therefore no
4331 21. Internal re-arrangement: the function for sending a challenge and reading
4332 a response while authentication was assuming a zero-terminated challenge
4333 string. It's now changed to take a pointer and a length, to allow for
4334 binary data in such strings.
4336 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4338 23. Exim was not respecting finduser_retries when seeking the login of the
4339 uid under which it was called; it was always trying 10 times. (The default
4340 setting of finduser_retries is zero.) Also, it was sleeping after the final
4341 failure, which is pointless.
4343 24. Implemented tls_on_connect_ports.
4345 25. Implemented acl_smtp_predata.
4347 26. If the domain in control=submission is set empty, Exim assumes that the
4348 authenticated id is a complete email address when it generates From: or
4349 Sender: header lines.
4351 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4352 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4353 chown and chgrp in /bin and hostname in /usr/bin.
4355 28. Exim was keeping the "process log" file open after each use, just as it
4356 does for the main log. This opens the possibility of it remaining open for
4357 long periods when the USR1 signal hits a daemon. Occasional processlog
4358 errors were reported, that could have been caused by this. Anyway, it seems
4359 much more sensible not to leave this file open at all, so that is what now
4362 29. The long-running daemon process does not normally write to the log once it
4363 has entered its main loop, and it closes the log before doing so. This is
4364 so that log files can straightforwardly be renamed and moved. However,
4365 there are a couple of unusual error situations where the daemon does write
4366 log entries, and I had neglected to close the log afterwards.
4368 30. The text of an SMTP error response that was received during a remote
4369 delivery was being truncated at 512 bytes. This is too short for some of
4370 the long messages that one sometimes sees. I've increased the limit to
4373 31. It is now possible to make retry rules that apply only when a message has a
4374 specific sender, in particular, an empty sender.
4376 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4377 it possible to be selective about when SMTP synchronization is enforced.
4379 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4381 32. Implemented hosts_connection_nolog.
4383 33. Added an ACL for QUIT.
4385 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4388 35. Added mailbox_size and mailbox_filecount to appendfile.
4390 36. Added control = no_multiline_responses to ACLs.
4392 37. There was a bug in the logic of the code that waits for the clock to tick
4393 in the case where the clock went backwards by a substantial amount such
4394 that the microsecond fraction of "now" was more than the microsecond
4395 fraction of "then" (but the whole seconds number was less).
4397 38. Added support for the libradius Radius client library this is found on
4398 FreeBSD (previously only the radiusclient library was supported).
4404 1. When certain lookups returned multiple values in the form name=value, the
4405 quoting of the values was not always being done properly. Specifically:
4406 (a) If the value started with a double quote, but contained no whitespace,
4408 (b) If the value contained whitespace other than a space character (i.e.
4409 tabs or newlines or carriage returns) it was not quoted.
4410 This fix has been applied to the mysql and pgsql lookups by writing a
4411 separate quoting function and calling it from the lookup code. The fix
4412 should probably also be applied to nisplus, ibase and oracle lookups, but
4413 since I cannot test any of those, I have not disturbed their existing code.
4415 2. A hit in the callout cache for a specific address caused a log line with no
4416 reason for rejecting RCPT. Now it says "Previous (cached) callout
4417 verification failure".
4419 3. There was an off-by-one bug in the queryprogram router. An over-long
4420 return line was truncated at 256 instead of 255 characters, thereby
4421 overflowing its buffer with the terminating zero. As well as fixing this, I
4422 have increased the buffer size to 1024 (and made a note to document this).
4424 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4425 when Exim is waiting for an SMTP response from a remote server, Exim
4426 restarts its select() call on the socket, thereby resetting its timeout.
4427 This is not a problem when such interrupts are rare. Somebody set up a cron
4428 job to run exiwhat every 2 minutes, which is less than the normal select()
4429 timeout (5 or 10 minutes). This meant that the select() timeout never
4430 kicked in because it was always reset. I have fixed this by comparing the
4431 time when an interrupt arrives with the time at the start of the first call
4432 to select(). If more time than the timeout has elapsed, the interrupt is
4433 treated as a timeout.
4435 5. Some internal re-factoring in preparation for the addition of Sieve
4436 extensions (by MH). In particular, the "personal" test is moved to a
4437 separate function, and given an option for scanning Cc: and Bcc: (which is
4438 not set for Exim filters).
4440 6. When Exim created an email address using the login of the caller as the
4441 local part (e.g. when creating a From: or Sender: header line), it was not
4442 quoting the local part when it contained special characters such as @.
4444 7. Installed new OpenBSD configuration files.
4446 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4447 try to make them clearer.
4449 9. Callout options, other than the timeout value, were being ignored when
4450 verifying sender addresses in header lines. For example, when using
4452 verify = header_sender/callout=no_cache
4454 the cache was (incorrectly) being used.
4456 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4457 only the Cygwin environment.
4459 11. When return_path_on_delivery was set as a log selector, if different remote
4460 addresses in the same message used different return paths and parallel
4461 remote delivery occurred, the wrong values would sometimes be logged.
4462 (Whenever a remote delivery process finished, the return path value from
4463 the most recently started remote delivery process was logged.)
4465 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4466 lines when AUTH and/or TLS are in use. This is the "received protocol"
4467 field. Exim used to use "asmtp" for authenticated SMTP, without any
4468 indication (in the protocol name) for TLS use. Now it follows the RFC and
4469 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4470 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4471 names appear in log lines as well as in Received: header lines.
4473 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4474 extensions, and comparison tests, and to fix some bugs.
4476 14. Changes to the "personal" filter test:
4478 (1) The test was buggy in that it was just doing the equivalent of
4479 "contains" tests on header lines. For example, if a user's address was
4480 anne@some.where, the "personal" test would incorrectly be true for
4482 To: susanne@some.where
4484 This test is now done by extracting each address from the header in turn,
4485 and checking the entire address. Other tests that are part of "personal"
4486 are now done using regular expressions (for example, to check local parts
4487 of addresses in From: header lines).
4489 (2) The list of non-personal local parts in From: addresses has been
4490 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4491 taken from the Sieve specification recommendations.
4493 (3) If the message contains any header line starting with "List-" it is
4494 treated as non-personal.
4496 (4) The test for "circular" in the Subject: header line has been removed
4497 because it now seems ill-conceived.
4499 15. Minor typos in src/EDITME comments corrected.
4501 16. Installed latest exipick from John Jetmore.
4503 17. If headers_add on a router specified a text string that was too long for
4504 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4505 of string_sprintf() is now avoided.
4507 18. $message_body_size was not set (it was always zero) when running the DATA
4508 ACL and the local_scan() function.
4510 19. For the "mail" command in an Exim filter, no default was being set for
4511 the once_repeat time, causing a random time value to be used if "once" was
4512 specified. (If the value happened to be <= 0, no repeat happened.) The
4513 default is now 0s, meaning "never repeat". The "vacation" command was OK
4514 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4515 (I found it when inspecting the code).
4517 20. There is now an overall timeout for performing a callout verification. It
4518 defaults to 4 times the callout timeout, which applies to individual SMTP
4519 commands during the callout. The overall timeout applies when there is more
4520 than one host that can be tried. The timeout is checked before trying the
4521 next host. This prevents very long delays if there are a large number of
4522 hosts and all are timing out (e.g. when the network connections are timing
4523 out). The value of the overall timeout can be changed by specifying an
4524 additional sub-option for "callout", called "maxwait". For example:
4526 verify = sender/callout=5s,maxwait=20s
4528 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4529 to the end before writing, but this should make it even safer).
4531 22. Exim was forgetting that it had advertised PIPELINING for the second and
4532 subsequent messages on an SMTP connection. It was also not resetting its
4533 memory on STARTTLS and an internal HELO.
4535 23. When Exim logs an SMTP synchronization error within a session, it now
4536 records whether PIPELINING has been advertised or not.
4538 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4539 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4540 rather than long int.
4542 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4544 26. Added the never_mail option to autoreply.
4550 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4551 crash if the getsockname() call failed; this can happen if a connection is
4552 closed very soon after it is established. The problem was simply in the
4553 order in which certain operations were done, causing Exim to try to write
4554 to the SMTP stream before it had set up the file descriptor. The bug has
4555 been fixed by making things happen in the correct order.
4561 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4562 before the connection was closed, thus losing the rejection response.
4564 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4565 some early Solaris releases, but causes trouble in current releases where
4566 socklen_t is defined.
4568 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4571 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4574 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4575 that releasing the top of it at the end releases what was used for sub-
4576 expansions (unless the block got too big). However, discard this block if
4577 the first thing is a variable or header, so that we can use its block when
4578 it is dynamic (useful for very large $message_headers, for example).
4580 6. Lookups now cache *every* query, not just the most recent. A new, separate
4581 store pool is used for this. It can be recovered when all lookup caches are
4582 flushed. Lookups now release memory at the end of their result strings.
4583 This has involved some general refactoring of the lookup sources.
4585 7. Some code has been added to the store_xxx() functions to reduce the amount
4586 of flapping under certain conditions.
4588 8. log_incoming_interface used to affect only the <= reception log lines. Now
4589 it causes the local interface and port to be added to several more SMTP log
4590 lines, for example "SMTP connection from", and rejection lines.
4592 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4594 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4596 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4597 could be overwritten at the end of the current message (or the start of a
4598 new message if it was set in a HELO ACL). The value is now preserved for
4599 the duration of the SMTP connection.
4601 12. If a transport had a headers_rewrite setting, and a matching header line
4602 contained an unqualified address, that address was qualified, even if it
4603 did not match any rewriting rules. The underlying bug was that the values
4604 of the flags that permit the existence of unqualified sender and recipient
4605 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4606 non-local messages, and by -bnq for local messages) were not being
4607 preserved with the message after it was received.
4609 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4610 "next input=" as part of the text comprising the host identity instead of
4611 the correct text. The code was using the same buffer for two different
4612 strings. However, depending on which order the printing function evaluated
4613 its arguments, the bug did not always show up. Under Linux, for example, my
4614 test suite worked just fine.
4616 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4617 doesn't work with some older versions of Perl. It has been changed to "my",
4618 which in any case is probably the better facility to use.
4620 15. A really picky compiler found some instances of statements for creating
4621 error messages that either had too many or two few arguments for the format
4624 16. The size of the buffer for calls to the DNS resolver has been increased
4625 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4626 for addresses that have a lot of PTR records. This alleviates a problem; it
4627 does not fully solve it.
4629 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4630 buffer now truncates the list and logs the incident, which is the same
4631 action as happens when Exim is looking up a host name and its aliases.
4632 Previously in this situation something unpredictable would happen;
4633 sometimes it was "internal error: store_reset failed".
4635 18. If a server dropped the connection unexpectedly when an Exim client was
4636 using GnuTLS and trying to read a response, the client delivery process
4637 crashed while trying to generate an error log message.
4639 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4640 string, for example:
4642 warn message = H1: something\nH2: something
4644 the text was added as a single header line from Exim's point of view
4645 though it ended up OK in the delivered message. However, searching for the
4646 second and subsequent header lines using $h_h2: did not work. This has been
4647 fixed. Similarly, if a system filter added multiple headers in this way,
4648 the routers could not see them.
4650 20. Expanded the error message when iplsearch is called with an invalid key to
4651 suggest using net-iplsearch in a host list.
4653 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4654 ACLs are no longer actually imposed (and a message to that effect is
4657 22. If a "gecos" field in a passwd entry contained escaped characters, in
4658 particular, if it contained a \" sequence, Exim got it wrong when building
4659 a From: or a Sender: header from that name. A second bug also caused
4660 incorrect handling when an unquoted " was present following a character
4661 that needed quoting.
4663 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4664 was not being matched caselessly.
4666 24. Arranged for all hyphens in the exim.8 source to be escaped with
4669 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4670 properly. Recipient callout cache records were still being keyed to include
4671 the sender, even when use_sender was set false. This led to far more
4672 callouts that were necessary. The sender is no longer included in the key
4673 when use_sender is false.
4675 26. Added "control = submission" modifier to ACLs.
4677 27. Added the ${base62d: operator to decode base 62 numbers.
4679 28. dnsdb lookups can now access SRV records.
4681 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4682 the configuration file.
4684 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4685 (-v) mode. This makes the output for a verbose queue run more intelligible.
4687 31. Added a use_postmaster feature to recipient callouts.
4689 32. Added the $body_zerocount variable, containing the number of binary zero
4690 bytes in the message body.
4692 33. The time of last modification of the "new" subdirectory is now used as the
4693 "mailbox time last read" when there is a quota error for a maildir
4696 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4698 35. Added +ignore_unknown as a special item in host lists.
4700 36. Code for decoding IPv6 addresses in host lists is now included, even if
4701 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4702 address was recognized as an IP address, but was then not correctly decoded
4703 into binary, causing unexpected and incorrect effects when compared with
4710 1. Very minor rewording of debugging text in manualroute to say "list of
4711 hosts" instead of "hostlist".
4713 2. If verify=header_syntax was set, and a header line with an unqualified
4714 address (no domain) and a large number of spaces between the end of the
4715 name and the colon was received, the reception process suffered a buffer
4716 overflow, and (when I tested it) crashed. This was caused by some obsolete
4717 code that should have been removed. The fix is to remove it!
4719 3. When running in the test harness, delay a bit after writing a bounce
4720 message to get a bit more predictability in the log output.
4722 4. Added a call to search_tidyup() just before forking a reception process. In
4723 theory, someone could use a lookup in the expansion of smtp_accept_max_
4724 per_host which, without the tidyup, could leave open a database connection.
4726 5. Added the variables $recipient_data and $sender_data which get set from a
4727 lookup success in an ACL "recipients" or "senders" condition, or a router
4728 "senders" option, similar to $domain_data and $local_part_data.
4730 6. Moved the writing of debug_print from before to after the "senders" test
4733 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4734 problems for message scanning, either using a data ACL, or using
4735 local_scan() because the Received: header was not generated till after they
4736 were called (in order to set the time as the time of reception completion).
4737 I have revised the way this works. The header is now generated after the
4738 body is received, but before the ACL or local_scan() are called. After they
4739 are run, the timestamp in the header is updated.
4745 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4746 before starting a queue runner without re-exec. This happened only when
4747 deliver_drop_privilege was set or when the Exim user was set to root. The
4748 effect of the bug was that timeouts during subsequent deliveries caused
4749 crashes instead of being properly handled. The handler is now left at its
4750 default (and expected) setting.
4752 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4753 message, again when deliver_drop_privilege is set or Exim is run as root.
4754 The bug described in (1) was not present in this case, but the tidying up
4755 of the other signals was missing. I have made the two cases consistent.
4757 3. The ignore_target_hosts setting on a manualroute router was being ignored
4758 for hosts that were looked up using the /MX notation.
4760 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4763 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4764 operated on the sender address. After changing the $sender_address to <>
4765 for the sender address verify, Exim was re-instated it as the original
4766 (before rewriting) address, but remembering that it had rewritten it, so it
4767 wasn't rewriting it again. This bug also had the effect of breaking the
4768 sender address verification caching when the sender address was rewritten.
4770 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4771 This has been changed so that if the ip literal address matches
4772 ignore_target_hosts, the router declines.
4774 7. Added expansion conditions match_domain, match_address, and match_local_
4775 part (NOT match_host).
4777 8. The placeholder for the Received: header didn't have a length field set.
4779 9. Added code to Exim itself and to exim_lock to test for a specific race
4780 condition that could lead to file corruption when using MBX delivery. The
4781 issue is with the lockfile that is created in /tmp. If this file is removed
4782 after a process has opened it but before that process has acquired a lock,
4783 there is the potential for a second process to recreate the file and also
4784 acquire a lock. This could lead to two Exim processes writing to the file
4785 at the same time. The added code performs the same test as UW imapd; it
4786 checks after acquiring the lock that its file descriptor still refers to
4787 the same named file.
4789 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4790 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4793 11. Added the smtp_active_hostname option. If used, this will typically be made
4794 to depend on the incoming interface address. Because $interface_address is
4795 not set up until the daemon has forked a reception process, error responses
4796 that can happen earlier (such as "too many connections") no longer contain
4799 12. If an expansion in a condition on a "warn" statement fails because a lookup
4800 defers, the "warn" statement is abandoned, and the next ACL statement is
4801 processed. Previously this caused the whole ACL to be aborted.
4803 13. Added the iplsearch lookup type.
4805 14. Added ident_timeout as a log selector.
4807 15. Added tls_certificate_verified as a log selector.
4809 16. Added a global option tls_require_ciphers (compare the smtp transport
4810 option of the same name). This controls incoming TLS connections.
4812 17. I finally figured out how to make tls_require_ciphers do a similar thing
4813 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4814 before starting the TLS session.
4816 18. Tabs are now shown as \t in -bP output.
4818 19. If the log selector return_path_on_delivery was set, Exim crashed when
4819 bouncing a message because it had too many Received: header lines.
4821 20. If two routers both had headers_remove settings, and the first one included
4822 a superfluous trailing colon, the final name in the first list and the
4823 first name in the second list were incorrectly joined into one item (with a
4824 colon in the middle).
4830 1. Added -C and -D options to the exinext utility, mainly to make it easier
4831 to include in the automated testing, but these could be helpful when
4832 multiple configurations are in use.
4834 2. The exinext utility was not formatting the output nicely when there was
4835 an alternate port involved in the retry record key, nor when there was a
4836 message id as well (for retries that were specific to a specific message
4837 and a specific host). It was also confused by IPv6 addresses, because of
4838 the additional colons they contain. I have fixed the IPv4 problem, and
4839 patched it up to do a reasonable job for IPv6.
4841 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4842 delivery, the log line now contains "pipelined" if PIPELINING was used.
4844 4. An SMTP transport process used to panic and die if the bind() call to set
4845 an explicit outgoing interface failed. This has been changed; it is now
4846 treated in the same way as a connect() failure.
4848 5. A reference to $sender_host_name in the part of a conditional expansion
4849 that was being skipped was still causing a DNS lookup. This no longer
4852 6. The def: expansion condition was not recognizing references to header lines
4853 that used bh_ and bheader_.
4855 7. Added the _cache feature to named lists.
4857 8. The code for checking quota_filecount in the appendfile transport was
4858 allowing one more file than it should have been.
4860 9. For compatibility with Sendmail, the command line option
4868 and sets the incoming protocol and host name (for trusted callers). The
4869 host name and its colon can be omitted when only the protocol is to be set.
4870 Note the Exim already has two private options, -pd and -ps, that refer to
4871 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4872 "s", but I don't think that's a major issue.
4874 10. A number of refactoring changes to the code, none of which should affect
4877 (a) The number of logging options was getting close to filling up the
4878 32-bit word that was used as a bit map. I have split them into two classes:
4879 those that are passed in the argument to log_write(), and those that are
4880 only ever tested independently outside of that function. These are now in
4881 separate 32-bit words, so there is plenty of room for expansion again.
4882 There is no change in the user interface or the logging behaviour.
4884 (b) When building, for example, log lines, the code previously used a
4885 macro that called string_cat() twice, in order to add two strings. This is
4886 not really sufficiently general. Furthermore, there was one instance where
4887 it was actually wrong because one of the argument was used twice, and in
4888 one call a function was used. (As it happened, calling the function twice
4889 did not affect the overall behaviour.) The macro has been replaced by a
4890 function that can join an arbitrary number of extra strings onto a growing
4893 (c) The code for expansion conditions now uses a table and a binary chop
4894 instead of a serial search (which was left over from when there were very
4895 few conditions). Also, it now recognizes conditions like "pam" even when
4896 the relevant support is not compiled in: a suitably worded error message is
4897 given if an attempt is made to use such a condition.
4899 11. Added ${time_interval:xxxxx}.
4901 12. A bug was causing one of the ddress fields not to be passed back correctly
4902 from remote delivery subprocesses. The field in question was not being
4903 subsequently used, so this caused to problems in practice.
4905 13. Added new log selectors queue_time and deliver_time.
4907 14. Might have fixed a bug in maildirsizefile handling that threw up
4908 "unexpected character" debug warnings, and recalculated the data
4909 unnecessarily. In any case, I expanded the warning message to give more
4912 15. Added the message "Restricted characters in address" to the statements in
4913 the default ACL that block characters like @ and % in local parts.
4915 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4916 Three changes have been made:
4918 (a) There was a serious bug; a negative response to MAIL caused the whole
4919 recipient domain to be cached as invalid, thereby blocking all messages
4920 to all local parts at the same domain, from all senders. This bug has
4921 been fixed. The domain is no longer cached after a negative response to
4922 MAIL if the sender used is not empty.
4924 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4927 (c) A new callout option, "use_sender" has been added for people who want
4928 the modified behaviour.
4934 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4937 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4938 indeed breaks things for older releases.
4940 3. Added additional logging to the case where there is a problem reading data
4941 from a filter that is running in a subprocess using a pipe, in order to
4942 try to track down a specific problem.
4944 4. Testing facility fudge: when running in the test harness and attempting
4945 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4946 getting "No route to host". Convert this to a timeout.
4948 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4951 6. Some OS don't have socklen_t but use size_t instead. This affects the
4952 fifth argument of getsockopt() amongst other things. This is now
4953 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4954 can be set for individual OS. I have set it for SunOS5, OSF1, and
4955 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4956 some earlier ones do not.
4958 7. Change 4.30/15 was not doing the test caselessly.
4960 8. The standard form for an IPv6 address literal was being rejected by address
4961 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4962 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4963 this, as well as the form without the "IPv6" on the front (but only when
4964 address literals are enabled, of course).
4966 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4968 10. Exim crashed if a message with an empty sender address specified by -f
4969 encountered a router with an errors_to setting. This could be provoked only
4970 by a command such as
4974 where an empty string was supplied; "<>" did not hit this bug.
4976 11. Installed PCRE release 4.5.
4978 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4979 remained set. It is now erased.
4981 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4982 times from message ids (which are base 36 rather than the normal 62).
4984 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4985 were being counted as actual protocol errors, and logged if the log
4986 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4987 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4988 and DATA following a set of rejected RCPTs do not count as protocol errors.
4989 In other words, Exim assumes they were pipelined, though this may not
4990 actually be the case. Of course, in all cases the client gets an
4991 appropriate error code.
4993 15. If a lookup fails in an ACL condition, a message about the failure may
4994 be available; it is used if testing the ACL cannot continue, because most
4995 such messages specify what the cause of the deferral is. However, some
4996 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4997 that caused an old message to be retained and used if a later statement
4998 caused a defer, replacing the real cause of the deferral.
5000 16. If an IP address had so many PTR records that the DNS lookup buffer
5001 was not large enough to hold them, Exim could crash while trying to process
5002 the truncated data. It now detects and logs this case.
5004 17. Further to 4.21/58, another change has been made: if (and only if) the
5005 first line of a message (the first header line) ends with CRLF, a bare LF
5006 in a subsequent header line has a space inserted after it, so as not to
5007 terminate the header.
5009 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5010 unnecessarily, used atoi() instead of strtol(), and didn't check the
5011 termination when getting file sizes from file names by regex.
5013 19. Completely re-implemented the support for maildirsize files, in the light
5014 of a number of problems with the previous contributed implementation
5015 (4.30/29). In particular:
5017 . If the quota is zero, the maildirsize file is maintained, but no quota is
5020 . If the maildir directory does not exist, it is created before any attempt
5021 to write a maildirsize file.
5023 . The quota value in the file is just a cache; if the quota is changed in
5024 the transport, the new value overrides.
5026 . A regular expression is available for excluding directories from the
5029 20. The autoreply transport checks the characters in options that define the
5030 message's headers; it allows continued headers, but it was checking with
5031 isspace() after an embedded newline instead of explicitly looking for a
5034 21. If all the "regular" hosts to which an address was routed had passed their
5035 expiry times, and had not reached their retry times, the address was
5036 bounced, even if fallback hosts were defined. Now Exim should go on to try
5039 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5040 equivalent code in the SMTP transport. Some hosts send humungous responses
5041 to HELO/EHLO, more than 1024 it seems.
5043 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5044 gives compiler warnings in some environments. I've now done it "properly",
5047 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5048 (because of problems with the built-in one) was declared to return uschar *
5049 instead of char *, causing compiler failure.
5051 25. Fixed a file descriptor leak when processing alias/forward files.
5053 26. Fixed a minor format string issue in dbfn.c.
5055 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5057 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5058 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5061 29. When a sender address is verified, it is cached, to save repeating the test
5062 when there is more than one recipient in a message. However, when the
5063 verification involves a callout, it is possible for different callout
5064 options to be set for different recipients. It is too complicated to keep
5065 track of this in the cache, so now Exim always runs a verification when a
5066 callout is required, relying on the callout cache for the optimization.
5067 The overhead is duplication of the address routing, but this should not be
5070 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5071 to be verified with callout=postmaster, and the main callout worked but the
5072 postmaster check failed, the verification correctly failed. However, if a
5073 subsequent RCPT command asked for sender verification *without* the
5074 postmaster check, incorrect caching caused this verification also to fail,
5077 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5078 it was not caching the DNS options (qualify_single, search_parents) that
5079 were used when the lookup failed. A subsequent lookup with different
5080 options therefore always gave the same answer, though there were cases
5081 where it should not have. (Example: a "domains = !$mx_any" option on a
5082 dnslookup router: the "domains" option is always processed without any
5083 widening, but the router might have qualify_single set.) Now Exim uses the
5084 cached value only when the same options are set.
5086 32. Added John Jetmore's "exipick" utility to the distribution.
5088 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5089 than a timeout (e.g. a certificate is required, and is not provided), an
5090 Exim server now closes the connection immediately. Previously it waited for
5091 the client to close - but if the client is SSL, it seems that they each
5092 wait for each other, leading to a delay before one of them times out.
5094 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5095 maintained 0.8.x compatibility because I don't think many are using it, and
5096 it is clearly obsolete.
5098 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5101 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5102 client certificate was expired. A simple patch fixes this, though I don't
5103 understand the full logic of why the verify callback is called multiple
5106 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5107 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5108 which causes problems with some clients (such as the Certicom SSL Plus
5109 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5110 disables the coutermeasure allowing Eudora to connect."
5112 38. Exim was not checking that a write() to a log file succeeded. This could
5113 lead to Bad Things if a log got too big, in particular if it hit a file
5114 size limit. Exim now panics and dies if it cannot write to a log file, just
5115 as it does if it cannot open a log file.
5117 39. Modified OS/Makefile-Linux so that it now contains
5119 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5121 The two -D definitions ensure that Exim is compiled with large file
5122 support, which makes it possible to handle log files that are bigger than
5125 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5126 instance) a domain was checked against a named list that involved a lookup,
5127 causing $domain_data to be set, then another domain was checked against the
5128 same list, then the first domain was re-checked, the value of $domain_data
5129 after the final check could be wrong. In particular, if the second check
5130 failed, it could be set empty. This bug probably also applied to
5133 41. The strip_trailing_dot option was not being applied to the address given
5134 with the -f command-line option.
5136 42. The code for reading a message's header from the spool was incrementing
5137 $received_count, but never initializing it. This meant that the value was
5138 incorrect (doubled) while delivering a message in the same process in which
5139 it was received. In the most common configuration of Exim, this never
5140 happens - a fresh exec is done - but it can happen when
5141 deliver_drop_privilege is set.
5143 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5144 it now includes up to 150 characters of the unexpected data in the log
5147 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5148 and building data strings. The size of both of these buffers was 10 000
5149 bytes - far larger than anybody would *ever* want, thought I. Needless to
5150 say, somebody hit the limit. I have increased the maximum line length to
5151 20 000 and the maximum data length of concatenated lines to 100 000. I have
5152 also fixed two bugs, because there was no checking on these buffers. Tsk,
5153 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5154 buffer is too small.
5156 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5157 lsearch lookups. Now it does.
5159 46. When parsing a route_list item in a manualroute router, a fixed-length
5160 buffer was used for the list of hosts. I made this 1024 bytes long,
5161 thinking that nobody would ever have a list of hosts that long. Wrong.
5162 Somebody had a whole pile of complicated expansion conditions, and the
5163 string was silently truncated, leading to an expansion error. It turns out
5164 that it is easier to change to an unlimited length (owing to other changes
5165 that have happened since this code was originally written) than to build
5166 structure for giving a limitation error. The length of the item that
5167 expands into the list of hosts is now unlimited.
5169 47. The lsearch lookup could not handle data where the length of text line was
5170 more than 4095 characters. Such lines were truncated, leading to shortened
5171 data being returned. It should now handle lines of any length.
5173 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5174 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5177 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5178 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5179 no longer happens in comment lines. A list of replacements is now placed
5180 at the head of all of the source files, except those whose only change is
5181 to replace PERL_COMMAND in the very first #! line.
5183 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5184 messages on the queue, with a bottom-up merge sort, using code contributed
5185 by Michael Haardt. This should make operations like -bp somewhat faster on
5186 large queues. It won't affect queue runners, except when queue_run_in_order
5189 51. Installed eximstats 1.31 in the distribution.
5191 52. Added support for SRV lookups to the dnslookup router.
5193 53. If an ACL referred to $message_body or $message_body_end, the value was not
5194 reset for any messages that followed in the same SMTP session.
5196 54. The store-handling optimization for building very long strings was not
5197 differentiating between the different store pools. I don't think this
5198 actually made any difference in practice, but I've tidied it.
5200 55. While running the routers to verify a sender address, $sender_address
5201 was still set to the sender address. This is wrong, because when routing to
5202 send a bounce to the sender, it would be empty. Therefore, I have changed
5203 it so that, while verifying a sender address, $sender_address is set to <>.
5204 (There is no change to what happens when verifying a recipient address.)
5206 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5207 target A or AAAA records (if not already returned) without resetting the
5208 qualify_single or search_parents options of the DNS resolver. These are
5209 inappropriate in this case because the targets of MX and SRV records must
5210 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5211 target that, when qualified, matched something in the local domain. These
5212 two options are now turned off when doing these lookups.
5214 57. It seems that at least some releases of Reiserfs (which does not have the
5215 concept of a fixed number of inodes) returns zero and not -1 for the
5216 number of available inodes. This interacted badly with check_spool_inodes,
5217 which assumed that -1 was the "no such thing" setting. What I have done is
5218 to check that the total number of inodes is greater than zero before doing
5219 the test of how many are available.
5221 58. When a "warn" ACL statement has a log_message modifier, the message is
5222 remembered, and not repeated. This is to avoid a lot of repetition when a
5223 message has many recipients that cause the same warning to be written.
5224 Howewer, Exim was preserving the list of already written lines for an
5225 entire SMTP session, which doesn't seem right. The memory is now reset if a
5226 new message is started.
5228 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5229 debugging output unless log_rewrite was also set.
5231 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5232 of (int)(handle) when we know that handle contains (void *)(-1).
5234 61. The Exim daemon panic-logs an error return when it closes the incoming
5235 connection. However "connection reset by peer" seems to be common, and
5236 isn't really an error worthy of noting specially, so that particular error
5239 62. When Exim is trying to find all the local interfaces, it used to panic and
5240 die if the ioctl to get the interface flags failed. However, it seems that
5241 on at least one OS (Solaris 9) it is possible to have an interface that is
5242 included in the list of interfaces, but for which you get a failure error
5243 for this call. This happens when the interface is not "plumbed" into a
5244 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5245 failure of the "get flags" call assumes that the interface is down.
5247 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5248 makes life easier for people who are doing arithmetic on fields extracted
5249 from dates, where you often get leading zeros that should not be
5250 interpreted as octal.
5252 64. Added qualify_domain to the redirect router, to override the global
5255 65. If a pathologically long header line contained very many addresses (the
5256 report of this problem mentioned 10 000) and each of them was rewritten,
5257 Exim could use up a very large amount of memory. (It kept on making new
5258 copies of the header line as it rewrote, and never released the old ones.)
5259 At the expense of a bit more processing, the header rewriting function has
5260 been changed so that it no longer eats memory in this way.
5262 66. The generation of the Received: header has been moved from the time that a
5263 message starts to be received, to the time that it finishes. The timestamp
5264 in the Received: header should now be very close to that of the <= log
5265 line. There are two side-effects of this change:
5267 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5268 logged header lines no longer include the local Received: line, because
5269 it has not yet been created. The same applies to a copy of the message
5270 that is returned to a non-SMTP sender when a message is rejected.
5272 (b) When a filter file is tested using -bf, no additional Received: header
5273 is added to the test message. After some thought, I decided that this
5276 This change does not affect the value of $received_for. It is still set
5277 after address rewriting, but before local_scan() is called.
5279 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5281 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5282 gave an unhelpful panic error message, and a defer error. I have managed to
5283 change this behaviour so that it now rejects any supplied certificate,
5284 which seems right, as the list of acceptable certificates is empty.
5286 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5287 gave an unhelpful defer error. I have not managed to make this reject any
5288 supplied certificates, but the error message it gives is "no certificate
5289 supplied", which is not helpful.
5291 70. exigrep's output now also includes lines that are not associated with any
5292 message, but which match the given pattern. Implemented by a patch from
5293 Martin Sluka, which also tidied up the Perl a bit.
5295 71. Recipient callout verification, like sender verification, was using <> in
5296 the MAIL FROM command. This isn't really the right thing, since the actual
5297 sender may affect whether the remote host accepts the recipient or not. I
5298 have changed it to use the actual sender in the callout; this means that
5299 the cache record is now keyed on a recipient/sender pair, not just the
5300 recipient address. There doesn't seem to be a real danger of callout loops,
5301 since a callout by the remote host to check the sender would use <>.
5302 [SEE ABOVE: changed after hitting problems.]
5304 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5305 temporary errors. However, in the case of such a code being given after
5306 the end of a data transmission (i.e. after ".") Exim was failing to write
5307 a retry record for the message. (Yes, there was some broken host that was
5308 actually sending 8xx at this point.)
5310 73. An unknown lookup type in a host list could cause Exim to panic-die when
5311 the list was checked. (An example that provoked this was putting <; in the
5312 middle of a list instead of at the start.) If this happened during a DATA
5313 ACL check, a -D file could be left lying around. This kind of configuration
5314 error no longer causes Exim to die; instead it causes a defer errror. The
5315 incident is still logged to the main and panic logs.
5317 74. Buglet left over from Exim 3 conversion. The message "too many messages
5318 in one connection" was written to the rejectlog but not the mainlog, except
5319 when address rewriting (yes!) was being logged.
5321 75. Added write_rejectlog option.
5323 76. When a system filter was run not as root (that is, when system_filter_user
5324 was set), the values of the $n variables were not being returned to the
5325 main process; thus, they were not subsequently available in the $sn
5328 77. Added +return_path_on_delivery log selector.
5330 78. A connection timeout was being treated differently from recipients deferred
5331 when testing hosts_max_try with a message that was older than the host's
5332 retry timeout. (The host should not be counted, thus allowing all hosts to
5333 be tried at least once before bouncing.) This may have been the cause of an
5334 occasionally reported bug whereby a message would remain on the queue
5335 longer than the retry timeout, but would be bounced if a delivery was
5336 forced. I say "may" because I never totally pinned down the problem;
5337 setting up timeout/retry tests is difficult. See also the next item.
5339 79. The ultimate address timeout was not being applied to errors that involved
5340 a combination of host plus message (for example, a timeout on a MAIL
5341 command). When an address resolved to a number of possible hosts, and they
5342 were not all tried for each delivery (e.g. because of hosts_max_try), a
5343 message could remain on the queue longer than the retry timeout.
5345 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5348 81. Fixed an obscure SMTP outgoing bug which required at least the following
5349 conditions: (a) there was another message waiting for the same server;
5350 (b) the server returned 5xx to all RCPT commands in the first message so
5351 that the message was not completed; (c) the server dropped the connection
5352 or gave a negative response to the RSET that Exim sends to abort the
5353 transaction. The observed case was a dropped connection after DATA that had
5354 been sent in pipelining mode. That is, the server had advertised PIPELINING
5355 but was not implementing it correctly. The effect of the bug was incorrect
5356 behaviour, such as trying another host, and this could lead to a crash.
5362 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5363 and daemon.c were passed as pointers to ints; they should have been
5364 pointers to socklen_t variables (which are typically unsigned ints).
5366 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5369 3. Fixed a really odd bug that affected only the testing scheme; patching a
5370 certain fixed string in the binary changed the value of another string that
5371 happened to be identical to the end of the original first string.
5373 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5374 name", it returns that address as the IP address. On some operating
5375 systems (e.g. Solaris), it also passes back the IP address string as the
5376 "host name". However, on others (e.g. Linux), it passes back an empty
5377 string. Exim wasn't checking for this, and was changing the host name to an
5378 empty string, assuming it had been canonicized.
5380 5. Although rare, it is permitted to have more than one PTR record for a given
5381 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5382 all the names associated with an address, because they do in Solaris.
5383 However, it seems that they do not in Linux for data that comes from the
5384 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5385 I found this out when I moved to a new Linux workstation and tried to run
5386 the Exim test suite.
5388 To get round this problem I have changed the code so that it now does its
5389 own call to the DNS to look up PTR records when searching for a host name.
5390 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5391 addresses that are only in /etc/hosts are still found.
5393 This behaviour is, however, controlled by an option called host_lookup_
5394 order, which defaults to "bydns:byaddr". If people want to use the other
5395 order, or indeed, just use one or the other means of lookup, they can
5396 specify it in this variable.
5398 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5399 some operating systems, this comes back from gethostbyaddr() as an empty
5400 string, and this is what Exim used to test for. However, it seems that in
5401 other systems, "." is yielded. Exim now tests for this case too.
5403 7. The values of check_spool_space and check_log_space are now held internally
5404 as a number of kilobytes instead of an absolute number of bytes. If a
5405 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5406 kilobyte. This means that much larger values can be stored.
5408 8. Exim monitor: an attempt to get the action menu when not actually pointing
5409 at a message produces an empty menu entitled "No message selected". This
5410 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5411 no entries in it ("Shell widget menu has zero width and/or height"). So I
5412 have added a single, blank menu entry in this case.
5414 9. Added ${quote_local_part.
5416 10. MIME decoding is now applied to the contents of Subject: header lines when
5419 11. Now that a reference to $sender_host_address automatically causes a reverse
5420 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5421 host lookup before query-style lookups in lists that might use this
5422 variable. This has therefore been abolished, and the "net-" prefix is no
5423 longer necessary for query-style lookups.
5425 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5426 have been a typo for LFLAGS, so it has been changed.
5428 13. The install script calls Exim with "-C /dev/null" in order to find the
5429 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5430 to be output. Howeve, since Exim outputs its version number before the
5431 error, it didn't break the script. It just looked ugly. I fixed this by
5432 always allowing "-C /dev/null" if the caller is root.
5434 14. Ignore overlarge ACL variable number when reading spool file - insurance
5435 against a later release with more variables having written the file.
5437 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5438 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5439 this, as well as the form without the "IPv6" on the front.
5441 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5442 OS/Makefile-Darwin file.
5444 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5445 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5447 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5448 variable. However, some OS use ints for inodes. I've added cast to long int
5449 to get rid of the compiler warning.
5451 19. I had forgotten to lock out "/../" in configuration file names when
5452 ALT_CONFIG_PREFIX was set.
5454 20. Routers used for verification do not need to specify transports. However,
5455 if such a router generated a host list, and callout was configured, Exim
5456 crashed, because it could not find a port number from the (non-existent)
5457 transport. It now assumes port 25 in this circumstance.
5459 21. Added the -t option to exigrep.
5461 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5462 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5463 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5464 them is set, LOOKUP_LSEARCH is forced.
5466 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5468 24. Added sender and host information to the "rejected by local_scan()" log
5469 line; previously there was no indication of these.
5471 25. Added .include_if_exists.
5473 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5474 Linux. It turns out that not all file systems support this. Apparently some
5475 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5476 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5477 not supported on the file descriptor, is now ignored when Exim is trying to
5478 sync a directory. This applies only to Linux.
5480 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5482 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5483 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5486 29. Added support for maildirsize files from supplied patch (modified a bit).
5488 30. The use of :fail: followed by an empty string could lead Exim to respond to
5489 sender verification failures with (e.g.):
5491 550 Verification failed for <xxx>
5492 550 Sender verify failed
5494 where the first response line was missing the '-' that indicates it is not
5495 the final line of the response.
5497 31. The loop for finding the name of the user that called Exim had a hardwired
5498 limit of 10; it now uses the value of finduser_retries, which is used for
5499 all other user lookups.
5501 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5504 33. Exim was neglecting to zero errno before one call of strtol() when
5505 expanding a string and expecting an integer value. On some systems this
5506 resulted in spurious "integer overflow" errors. Also, it was casting the
5507 result into an int without checking.
5509 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5510 did not work. The code looks as if it has *never* worked, though it appears
5511 to have been documented since at least releast 1.62. I have made it work.
5513 35. The "timeout_DNS" error in retry rules, also documented since at least
5514 1.62, also never worked. As it isn't clear exactly what this means, and
5515 clearly it isn't a major issue, I have abolished the feature by treating it
5516 as "timeout", and writing a warning to the main and panic logs.
5518 36. The display of retry rules for -brt wasn't always showing the error code
5521 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5522 timeout_connect_A, timeout_connect_MX.
5524 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5525 to the empty sender.
5527 39. The daemon was not analysing the content of -oX till after it had closed
5528 stderr and disconnected from the controlling terminal. This meant that any
5529 syntax errors were only noted on the panic log, and the return code from
5530 the command was 0. By re-arranging the code a little, I've made the
5531 decoding happen first, so such errors now appear on stderr, and the return
5532 code is 1. However, the actual setting up of the sockets still happens in
5533 the disconnected process, so errors there are still only recorded on the
5536 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5537 connections (as happens on some IP stacks) was logged at start up time as
5538 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5539 it from "IPv6 and IPv4", which means that two separate sockets are being
5542 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5543 says whether AF_INET or AF_INET6 was passed as an argument.
5545 42. Exiwhat output was messed up when time zones were included in log
5548 43. Exiwhat now gives more information about the daemon's listening ports,
5549 and whether -tls-on-connect was used.
5551 44. The "port" option of the smtp transport is now expanded.
5553 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5554 silently ignored. Now an error message is written to the main and panic
5557 46. There's a new ACL modifier called "logwrite" which writes to a log file
5558 as soon as it is encountered.
5560 47. Added $local_user_uid and $local_user_gid at routing time.
5562 48. Exim crashed when trying to verify a sender address that was being
5565 49. Exim was recognizing only a space character after ".include". It now also
5566 recognizes a tab character.
5568 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5569 extracting the relevant information from the specification. The man page no
5570 longer contains scrambled data for the -d option, and I've added a section
5571 at the front about calling Exim under different names.
5573 51. Added "extra_headers" argument to the "mail" command in filter files.
5575 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5578 53. Installed eximstats 1.29.
5580 54. Added transport_filter_timeout as a generic transport option.
5582 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5583 Cc: header lines. This was required by RFC 822, but it not required by RFC
5586 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5587 incoming messages that did not have them. Now it does so only if the
5588 message originates locally, that is, if there is no associated remote host
5589 address. When Resent- header lines are present, this applies to the Resent-
5590 lines rather than the non-Resent- lines.
5592 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5593 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5595 58. Messages for configuration errors now include the name of the main
5596 configuration file - useful now that there may be more than one file in a
5597 list (.included file names were always shown).
5599 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5600 for those rare installations that do not start the daemon as root or run it
5601 setuid root. I've cut out the call to initgroups() if the daemon is not
5604 60. The Exim user and group can now be bound into the binary as text strings
5605 that are looked up at the start of Exim's processing.
5607 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5609 62. Added $mailstore_basename variable.
5611 63. Installed patch to sieve.c from Michael Haardt.
5613 64. When Exim failed to open the panic log after failing to open the main log,
5614 the original message it was trying to log was written to stderr and debug
5615 output, but if they were not available (the usual case in production), it
5616 was lost. Now it is written to syslog before the two lines that record the
5617 failures to open the logs.
5619 65. Users' Exim filters run in subprocesses under the user's uid. It is
5620 possible for a "deliver" command or an alias in a "personal" command to
5621 provoke an address rewrite. If logging of address rewriting is configured,
5622 this fails because the process is not running as root or exim. There may be
5623 a better way of dealing with this, but for the moment (because 4.30 needs
5624 to be released), I have disabled address rewrite logging when running a
5625 filter in a non-root, non-exim process.
5631 1. The buildconfig auxiliary program wasn't quoting the value set for
5632 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5633 not defined. This bug was masked in 4.22 by the effect that was fixed in
5636 2. Some messages that were rejected after a message id was allocated were
5637 shown as "incomplete" by exigrep. It no longer does this for messages that
5638 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5640 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5641 have allow_domain_literals set, the ID did not get logged in the <= line.
5642 Domain literals are now always recognized in Message-ID: header lines.
5644 4. The first argument for a ${extract expansion item is the key name or field
5645 number. Leading and trailing spaces in this item were not being ignored,
5646 causing some misleading effects.
5648 5. When deliver_drop_privilege was set, single queue runner processes started
5649 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5650 same command in the process it spins off) were not dropping privilege.
5652 6. When the daemon running as "exim" started a queue runner, it always
5653 re-executed Exim in the spun-off process. This is a waste of effort when
5654 deliver_drop_privilege is set. The new process now just calls the
5655 queue-runner function directly.
5661 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5664 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5665 ignored. Though the use of -oP was forcing the writing of a pid file, it
5666 was always written to the default place.
5668 3. If the message "no IP address found for host xxxx" is generated during
5669 incoming verification, it is now followed by identification of the incoming
5670 connection (so you can more easily find what provoked it).
5672 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5674 5. Added some features to "harden" Exim a bit more against certain attacks:
5676 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5677 be put in Local/Makefile. This is like the never_users runtime option,
5678 but it cannot be overridden. The default setting is "root".
5680 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5681 prefix string with which any file named in a -C command line option
5684 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5685 is retained for -C and -D only if the caller of Exim is root. Without
5686 it, the exim user may also use -C and -D and retain privilege.
5688 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5689 command line option is disabled.
5691 6. Macro names set by the -D option must start with an upper case letter, just
5692 like macro names defined in the configuration file.
5694 7. Added "dereference=" facility to LDAP.
5696 8. Two instances of the typo "uknown" in the source files are fixed.
5698 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5699 the Configure-Makefile script screwed up while processing it.
5701 10. Incorporated PCRE 4.4.
5703 11. The SMTP synchronization check was not operating right at the start of an
5704 SMTP session. For example, it could not catch a HELO sent before the client
5705 waited for the greeting. There is now a check for outstanding input at the
5706 point when the greeting is written. Because of the duplex, asynchronous
5707 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5708 way, but not yet received, when the check is performed.
5710 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5711 on TCP/IP sockets, because this apparently causes some broken clients to
5714 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5715 unchanged) from the Cygwin maintainer.
5717 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5718 was supported instead of testing for mbx. Effectively a typo.
5720 15. The spa authenticator server code was not checking that the input it
5721 received was valid base64.
5723 16. The debug output line for the "set" modifier in ACLs was not showing the
5724 name of the variable that was being set.
5726 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5728 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5729 lookup on its own. Something else was needed to trigger the lookup. For
5730 example, a match in host_lookup or the need for a host name in a host list.
5731 Now, if $sender_host_name is referenced and the host name has not yet been
5732 looked up, a lookup is performed. If the lookup fails, the variable remains
5733 empty, and $host_lookup_failed is set to "1".
5735 19. Added "eqi" as a case-independent comparison operator.
5737 20. The saslauthd authentication condition could segfault if neither service
5738 nor realm was specified.
5740 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5741 error message that was logged was misleading, and incoming SMTP
5742 connections were dropped. The message is now more accurate, and temporary
5743 errors are given to SMTP connections.
5745 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5746 (except RSET) with a 421 error, until QUIT is received. However, it was
5747 failing to send a response to QUIT.
5749 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5750 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5751 ACL is now run after the helo_try_verify_hosts code.
5753 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5754 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5755 case-independent, so other case variants are also recognized). Apparently
5756 some systems use these upper case variants.
5758 25. If more than two messages were waiting for the same host, and a transport
5759 filter was specified for the transport, Exim sent two messages over the
5760 same TCP/IP connection, and then failed with "socket operation on non-
5761 socket" when it tried to send the third.
5763 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5765 27. The extern definition of crypt16() in expand.c was not being excluded when
5766 the OS had its own crypt16() function.
5768 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5769 as a preferred synonym for return_size_limit, both as an option and as an
5772 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5774 30. Changed the default configuration ACL to relax the local part checking rule
5775 for addresses that are not in any local domains. For these addresses,
5776 slashes and pipe symbols are allowed within local parts, but the sequence
5777 /../ is explicitly forbidden.
5779 31. SPA server authentication was not clearing the challenge buffer before
5782 32. log_message in a "warn" ACL statement was writing to the reject log as
5783 well as to the main log, which contradicts the documentation and doesn't
5784 seem right (because no rejection is happening). So I have stopped it.
5786 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5787 However, I am unable to do any testing of this.
5789 34. Fixed an infelicity in the appendfile transport. When checking directories
5790 for a mailbox, to see if any needed to be created, it was accidentally
5791 using path names with one or more superfluous leading slashes; tracing
5792 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5794 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5795 contents are added to the log line that is written for every discarded
5796 recipient. (Previously a log_message setting was ignored.)
5798 36. The ${quote: operator now quotes the string if it is empty.
5800 37. The install script runs exim in order to find its version number. If for
5801 some reason other than non-existence or emptiness, which it checks, it
5802 could not run './exim', it was installing it with an empty version number,
5803 i.e. as "exim-". This error state is now caught, and the installation is
5806 38. An argument was missing from the function that creates an error message
5807 when Exim fails to connect to the socket for saslauthd authentication.
5808 This could cause Exim to crash, or give a corrupted message.
5810 39. Added isip, isip4, and isip6 to ${if conditions.
5812 40. The ACL variables $acl_xx are now saved with the message, and can be
5813 accessed later in routers, transports, and filters.
5815 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5816 strings in the file are not string-expanded.
5818 42. If a MAIL command specified a SIZE value that was too large to fit into an
5819 int variable, the check against message_size_limit failed. Such values are
5820 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5821 day this will have to be increased, but I don't think I want to be around
5822 when emails are that large.
5829 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5830 iconv() is not standard in FreeBSD.
5832 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5833 IPv6 enabled. The observed symptom was a segmentation fault on return from
5834 the function os_common_find_running_interfaces() in src/os.c.
5836 3. In the check_special_case() function in daemon.c I had used "errno" as an
5837 argument name, which causes warnings on some systems. This was basically a
5838 typo, since it was named "eno" in the comments!
5840 4. The code that waits for the clock to tick (at a resolution of some fraction
5841 of a second) so as to ensure message-id uniqueness was always waiting for
5842 at least one whole tick, when it could have waited for less. [This is
5843 almost certainly not relevant at current processor speeds, where it is
5844 unlikely to ever wait at all. But we try to future-proof.]
5846 5. The function that sleeps for a time interval that includes fractions of a
5847 second contained a race. It did not block SIGALRM between setting the
5848 timer, and suspending (a couple of lines later). If the interval was short
5849 and the sigsuspend() was delayed until after it had expired, the suspension
5850 never ended. On busy systems this could lead to processes getting stuck for
5853 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5854 process, before it forks any delivery processes. The open lookup caching
5855 mechanism meant that the open file or database connection was passed into
5856 the delivery process. The problem was that delivery processes always tidy
5857 up cached lookup data. This could cause a problem for the next delivery
5858 process started by the queue runner, because the external queue runner
5859 process does not know about the closure. So the next delivery process
5860 still has data in the lookup cache. In the case of a file lookup, there was
5861 no problem because closing a file descriptor in a subprocess doesn't affect
5862 the parent. However, if the lookup was caching a connection to a database,
5863 the connection was closed, and the second delivery process was likely to
5864 see errors such as "PGSQL: query failed: server closed the connection
5865 unexpectedly". The problem has been fixed by closing all cached lookups
5866 in a queue runner before running a delivery process.
5868 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5869 seem to have the "const" qualifier which it has on other OS. I've
5872 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5873 *of the same type* (client or server) with the same public name that an
5874 error should be diagnosed.
5876 9. When Exim looked up a host name for an IP address, but failed to find the
5877 original IP address when looking up the host name (a safety check), it
5878 output the message "<ip address> does not match any IP for NULL", which was
5879 confusing, to say the least. The bug was that the host name should have
5880 appeared instead of "NULL".
5882 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5883 user that is built into the binary, and the -C or -D options is used, root
5884 privilege is dropped before the configuration file is read. In addition,
5885 logging is switched to stderr instead of the normal log files. If the
5886 configuration then re-defines the Exim user, the unprivileged environment
5887 is probably not what is expected, so Exim logs a panic warning message (but
5890 However, if deliver_drop_privilege is set, the unprivileged state may well
5891 be exactly what is intended, so the warning has been cut out in that case,
5892 and Exim is allowed to try to write to its normal log files.
5898 1. smtp_return_error_details was not giving details for temporary sender
5899 or receiver verification errors.
5901 2. Diagnose a configuration error if two authenticators have the same public
5904 3. Exim used not to create the message log file for a message until the first
5905 delivery attempt. This could be confusing when incoming messages were held
5906 for policy or load reasons. The message log file is now created at the time
5907 the message is received, and an initial "Received" line is written to it.
5909 4. The automatically generated man page for command line options had a minor
5910 bug that caused no ill effects; however, a more serious problem was that
5911 the procedure for building the man page automatically didn't always
5912 operate. Consequently, release 4.20 contains an out-of-date version. This
5913 shouldn't happen again.
5915 5. When building Exim with embedded Perl support, the script that builds the
5916 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5917 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5919 6. The freeze_tell option was not being used for messages that were frozen on
5920 arrival, either by an ACL or by local_scan().
5922 7. Added the smtp_incomplete_transaction log selector.
5924 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5925 was accepting AUTH without a new EHLO.
5927 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5928 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5931 10. Logging of TCP/IP connections (when configured) now happens in the main
5932 daemon process instead of the child process, so that the TCP/IP connection
5933 count is more accurate (but it can never be perfect).
5935 11. The use of "drop" in a nested ACL was not being handled correctly in the
5936 outer ACL. Now, if condition failure induced by the nested "drop" causes
5937 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5938 or "require"), the connection is dropped.
5940 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5941 that yield "discard" can now be used with an "accept" or a "discard" verb,
5942 but an error is generated for any others (because I can't see a useful way
5943 to define what should happen).
5945 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5946 are now processed in the same way as lines in the Exim configuration file.
5947 In particular, continuation lines are supported.
5949 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5951 15. Added -ti meaning -t -i.
5953 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5954 domains, and warn by logging if others are found.
5956 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5957 returned from GIFCONF when Exim is trying to find the list of interfaces on
5958 a host. The code in os.c has been modified to copy each ifreq to an aligned
5959 structure in all cases.
5961 Also, in some cases, the returned ifreq's were being copied to a 'struct
5962 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5963 means the last couple of bytes of an IPv6 address could be chopped if the
5964 ifreq contained only a normal sockaddr (14 bytes storage).
5966 18. Named domain lists were not supported in the hosts_treat_as_local option.
5967 An entry such as +xxxx was not recognized, and was treated as a literal
5970 19. Ensure that header lines added by a DATA ACL are included in the reject log
5971 if the ACL subsequently rejects the message.
5973 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5974 MD5 (which is deprecated).
5976 21. When testing a filter file using -bf, Exim was writing a message when it
5977 took the sender from a "From " line in the message, but it was not doing so
5978 when it took $return_path from a Return-Path: header line. It now does.
5980 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5981 with a valid header line field name (a series of printing characters
5982 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5984 23. Changed "disc" in the source to "disk" to conform to the documentation and
5985 the book and for uniformity.
5987 24. Ignore Sendmail's -Ooption=value command line item.
5989 25. When execve() failed while trying to run a command in a pipe transport,
5990 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5991 could be confused with a return value of 69 from the command itself. This
5992 has been changed to 127, the value the shell returns if it is asked to run
5993 a non-existent command. The wording for the related log line suggests a
5994 non-existent command as the problem.
5996 26. If received_header_text expands to an empty string, do not add a Received:
5997 header line to the message. (Well, it adds a token one on the spool, but
5998 marks it "old" so that it doesn't get used or transmitted.)
6000 27. Installed eximstats 1.28 (addition of -nt option).
6002 28. There was no check for failure on the call to getsockname() in the daemon
6003 code. This can fail if there is a shortage of resources on the system, with
6004 ENOMEM, for example. A temporary error is now given on failure.
6006 29. Contrary to the C standard, it seems that in some environments, the
6007 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6008 program. Exim now does this explicitly; it affects the formatting of
6009 timestamps using strftime().
6011 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6012 complaints. I've now initialized all the variables, to avoid this.
6014 32. Header lines added by a system filter were not being "seen" during
6015 transport-time rewrites.
6017 33. The info_callback() function passed to OpenSSL is set up with type void
6018 (*)(SSL *, int, int), as described somewhere. However, when calling the
6019 function (actually a macro) that sets it up, the type void(*)() is
6020 expected. I've put in a cast to prevent warnings from picky compilers.
6022 34. If a DNS black list lookup found a CNAME record, but there were no A
6023 records associated with the domain it pointed at, Exim crashed.
6025 35. If a DNS black list lookup returned more than one A record, Exim ignored
6026 all but the first. It now scans all returned addresses if a particular IP
6027 value is being sought. In this situation, the contents of the
6028 $dnslist_value variable are a list of all the addresses, separated by a
6031 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6032 to accept a host name and all its aliases if the forward lookup for any of
6033 them yielded the IP address of the incoming connection. Now it accepts only
6034 those names whose forward lookup yields the correct IP address. Any other
6035 names are discarded. This closes a loophole whereby a rogue DNS
6036 administrator could create reverse DNS records to break through a
6037 wildcarded host restriction in an ACL.
6039 37. If a user filter or a system filter that ran in a subprocess used any of
6040 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6041 the wrong values were passed to the pipe command ($thisaddress had the
6042 value of $0, $0 had the value of $1, etc). This bug was introduced by
6043 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6045 38. Improved the line breaking for long SMTP error messages from ACLs.
6046 Previously, if there was no break point between 40 and 75 characters, Exim
6047 left the rest of the message alone. Two changes have been made: (a) I've
6048 reduced the minimum length to 35 characters; (b) if it can't find a break
6049 point between 35 and 75 characters, it looks ahead and uses the first one
6050 that it finds. This may give the occasional overlong line, but at least the
6051 remaining text gets split now.
6053 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6054 file descriptors might be low, and that setting 1000 would always raise it.
6055 It turns out that in some environments, the limit is already over 1000 and
6056 that lowering it causes trouble. So now Exim takes care not to decrease it.
6058 40. When delivering a message, the value of $return_path is set to $sender_
6059 address at the start of routing (routers may change the value). By an
6060 oversight, this default was not being set up when an address was tested by
6061 -bt or -bv, which affected the outcome if any router or filter referred to
6064 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6065 log only when "message" or "log_message" are set. However, if one of the
6066 conditions was an address verification, or a call to a nested ACL, the
6067 messages generated by the underlying test were being passed through. This
6068 no longer happens. The underlying message is available in $acl_verify_
6069 message for both "message" and "log_message" expansions, so it can be
6070 passed through if needed.
6072 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6073 new expansion $bh_ to give the encoded byte string without charset
6074 translation. Translation happens only if iconv() is available; HAVE_ICONV
6075 indicates this at build time. HEADERS_CHARSET gives the charset to
6076 translate to; headers_charset can change it in the configuration, and
6077 "headers charset" can change it in an individual filter file.
6079 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6080 that creates RFC 2047 encoded "words" labels them as that charset instead
6081 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6082 expansion operator; (ii) when Exim creates a From: line for a local
6083 message; (iii) when a header line is rewritten to include a "phrase" part.
6085 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6086 buggy, causing it to skip the first lines of messages whose message ID
6087 ended in 'D'. This would not have bitten before Exim release 4.14, because
6088 message IDs were unlikely to end in 'D' before then. The effect was to have
6089 incorrect size information for certain domains.
6091 45. #include "config.h" was missing at the start of the crypt16.c module. This
6092 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6095 46. If there was a timeout during a "random" callout check, Exim treated it as
6096 a failure of the random address, and carried on sending RSET and the real
6097 address. If the delay was just some slowness somewhere, the response to the
6098 original RCPT would be taken as a response to RSET and so on, causing
6099 mayhem of various kinds.
6101 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6102 when I implemented it. It didn't allow for the fact that some option values
6103 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6104 the right test for positive values.
6106 48. Domain names in DNS records are case-independent. Exim always looks them up
6107 in lower case. Some resolvers return domain names in exactly the case they
6108 appear in the zone file, that is, they may contain uppercase letters. Not
6109 all resolvers do this - some return always lower case. Exim was treating a
6110 change of case by a resolver as a change of domain, similar to a widening
6111 of a domain abbreviation. This triggered its re-routing code and so it was
6112 trying to route what was effectively the same domain again. This normally
6113 caused routing to fail (because the router wouldn't handle the domain
6114 twice). Now Exim checks for this case specially, and just changes the
6115 casing of the domain that it ultimately uses when it transmits the message
6118 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6121 50. If a filter generated a file delivery with a non-absolute name (possible if
6122 no home directory exists for the router), the forbid_file option was not
6125 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6126 the existing equality matching.
6128 52. Exim was using ints instead of ino_t variables in some places where it was
6129 dealing with inode numbers.
6131 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6132 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6133 called TMPDIR, and if it finds it is different, it changes its value.
6135 54. The smtp_printf() function is now made available to local_scan() so
6136 additional output lines can be written before returning. There is also an
6137 smtp_fflush() function to enable the detection of a dropped connection.
6138 The variables smtp_input and smtp_batched_input are exported to
6141 55. Changed the default runtime configuration: the message "Unknown user"
6142 has been removed from the ACL, and instead placed on the localuser router,
6143 using the cannot_route_message feature. This means that any verification
6144 failures that generate their own messages won't get overridden. Similarly,
6145 the "Unrouteable address" message that was in the ACL for unverifiable
6146 relay addresses has also been removed.
6148 56. Added hosts_avoid_esmtp to the smtp transport.
6150 57. The exicyclog script was not checking for the esoteric option
6151 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6152 will work only if exicyclog is run under the appropriate euid.
6154 58. Following a discussion on the list, the rules by which Exim recognises line
6155 endings on incoming messages have been changed. The -dropcr and drop_cr
6156 options are now no-ops, retained only for backwards compatibility. The
6157 following line terminators are recognized: LF CRLF CR. However, special
6158 processing applies to CR:
6160 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6161 nor a local message in the state where . is a terminator.
6163 (ii) If a bare CR is encountered in a header line, an extra space is added
6164 after the line terminator so as not to end the header. The reasoning
6165 behind this is that bare CRs in header lines are most likely either
6166 to be mistakes, or people trying to play silly games.
6168 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6169 was being incorrectly given as 18 bytes larger than it should have been.
6170 This is a VOB (very old bug).
6172 60. This may never have affected anything current, but just in case it has:
6173 When the local host is found other than at the start of a list of hosts,
6174 the local host, those with the same MX, and any that follow, are discarded.
6175 When the list in question was part of a longer list of hosts, the following
6176 hosts (not currently being processed) were also being discarded. This no
6177 longer happens. I'm not sure if this situation could ever has previously
6180 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6183 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6184 header. This is something that is recommended in a new Internet Draft, and
6185 is something that is documented as being done by Sendmail. There are two
6186 possible values. For messages generated by the autoreply transport, Exim
6189 Auto-Submitted: auto-replied
6191 whereas for all other generated messages (e.g. bounces) it adds
6193 Auto-Submitted: auto-generated
6195 63. The "personal" condition in filters now includes a test for the
6196 Auto-Submitted: header. If it contains the string "auto-" the message it
6197 not considered personal.
6199 64. Added rcpt_include_affixes as a generic transport option.
6201 65. Added queue_only_override (default true).
6203 66. Added the syslog_duplication option.
6205 67. If what should have been the first header line of a message consisted of
6206 a space followed by a colon, Exim was mis-interpreting it as a header line.
6207 It isn't of course - it is syntactically invalid and should therefore be
6208 treated as the start of the message body. The misbehaviour could have
6209 caused a number of strange effects, including loss of data in subsequent
6210 header lines, and spool format errors.
6212 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6213 client host had authenticated. This control can now be exercised by an ACL
6214 for more flexibility.
6216 69. By default, callouts do not happen when testing with -bh. There is now a
6217 variant, -bhc, which does actually run the callout code, including
6218 consulting and updating the callout cache.
6220 70. Added support for saslauthd authentication, courtesy of Alexander
6223 71. If statvfs() failed on the spool or log directories while checking their
6224 size for availability, Exim confusingly gave the error "space shortage".
6225 Furthermore, in debugging mode it crashed with a floating point exception.
6226 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6227 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6228 serious problem, Exim now writes to the main and panic logs when this
6229 happens, with details of the failure. It then refuses to accept the
6230 incoming message, giving the message "spool directory problem" or "log
6231 directory problem" with a 421 code for SMTP messages.
6233 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6234 0, 1, and 2 exist, because some OS complain for execs without them (see
6235 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6236 descriptors. However, the code omitted to check that the open succeeded,
6237 causing mysterious errors if for some reason the permissions on /dev/null
6238 got screwed. Now Exim writes a message to the main and panic logs, and
6239 bombs out if it can't open /dev/null.
6241 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6242 interact so that it is all more flexible. It is supposed to remain
6243 backwards compatible. Also added extra_local_interfaces.
6245 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6246 to bomb out with an assertion failure - to the client this appears as a
6247 connection drop. This problem occurs in the part of the code that was taken
6248 from the Samba project. Fortunately, the assertion is in a very simple
6249 function, so I have fixed this by reproducing the function inline in the
6250 one place where it is called, and arranging for authentication to fail
6251 instead of killing the process with assert().
6253 75. The SPA client code was not working when the server requested OEM rather
6254 than Unicode encoding.
6256 76. Added code to make require_files with a specific uid setting more usable in
6257 the case where statting the file as root fails - usually a non-root-mounted
6258 NFS file system. When this happens and the failure is EACCES, Exim now
6259 forks a subprocess and does the per-uid checking as the relevant uid.
6261 77. Added process_log_path.
6263 78. If log_file_path was not explicitly set, a setting of check_log_space or
6264 check_log_inodes was ignored.
6266 79. If a space check for the spool or log partitions fails, the incident is now
6267 logged. Of course, in the latter case the data may get lost...
6269 80. Added the %p formatting code to string_format() so that it can be used to
6270 print addresses in debug_print(). Adjusted all the address printing in the
6271 debugging in store.c to use %p rather than %d.
6273 81. There was a concern that a line of code in smtp_in.c could overflow a
6274 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6275 initially expressed, the concern was not well-founded, because trailing
6276 spaces are removed early. However, if the trailing spaces were followed by
6277 a NULL, they did not get removed, so the overflow was possible. Two fixes
6280 (a) I re-wrote the offending code in a cleaner fashion.
6281 (b) If an incoming SMTP command contains a NULL character, it is rejected
6284 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6285 runs initgroups(), so that if the Exim user is in any additional groups,
6286 they will be used during message reception.
6292 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob